End-of-life care in the critical care setting:

Nurses’ practices and factors affecting these practices

Kristen Louise Ranse BNurs (Hons) MCritCare Nursing

Submitted in fulfilment of the requirements of the degree of

Doctor of Philosophy (PhD)

School of Nursing and Midwifery

Institute of Health and Biomedical Innovation

Queensland University of Technology

February 2013

ii

Keywords

End of life, practices, critical care, intensive care, nurse, PARIHS framework,

context, evidence, facilitation, survey, mixed methods, descriptive statistics,

exploratory factor analysis, general linear model, palliative values, patient and family

centred care, preparedness

iii

Abstract

Background

Critical care units are designed and resourced to save lives, yet the provision of end-

of-life care is a significant component of nursing work in these settings. Limited

research has investigated the actual practices of critical care nurses in the provision

of end-of-life care, or the factors influencing these practices. To improve the care

that patients at the end of life and their families receive, and to support nurses in the

provision of this care, further research is needed. The purpose of this study was to

identify critical care nurses’ end-of-life care practices, the factors influencing the

provision of end-of-life care and the factors associated with specific end-of-life care

practices.

Methods

A three-phase exploratory sequential mixed-methods design was utilised. Phase one

used a qualitative approach involving interviews with a convenience sample of five

intensive care nurses to identify their end-of-life care experiences and practices. In

phase two, an online survey instrument was developed, based on a review of the

literature and the findings of phase one. The survey instrument was reviewed by six

content experts and pilot tested with a convenience sample of 28 critical care nurses

(response rate 45%) enrolled in a postgraduate critical care nursing subject. The

refined survey instrument was used in phase three of this study to conduct a national

survey of critical care nurses. Descriptive analyses, exploratory factor analysis and

univariate general linear modelling was undertaken on completed survey responses

from 392 critical care nurses (response rate 25%).

Results

Six end-of-life care practice areas were identified in this study: information sharing,

environmental modification, emotional support, patient and family-centred decision

making, symptom management and spiritual support. The items most frequently

identified as always undertaken by critical care nurses in the provision of end-of-life

care were from the information sharing and environmental modification practice

iv

areas. Items least frequently identified as always undertaken included items from the

emotional support practice area.

Eight factors influencing the provision of end-of-life care were identified: palliative

values, patient and family preferences, knowledge, preparedness, organisational

culture, resources, care planning, and emotional support for nurses. Strong agreement

was noted with items reflecting values consistent with a palliative approach and

inclusion of patient and family preferences. Variation was noted in agreement for

items regarding opportunities for knowledge acquisition in the workplace and formal

education, yet most respondents agreed that they felt adequately prepared. A context

of nurse-led practice was identified, with variation in access to resources noted.

Collegial support networks were identified as a source of emotional support for

critical care nurses.

Critical care nurses reporting values consistent with a palliative approach and/or

those who scored higher on support for patient and family preferences were more

likely to be engaged in end-of-life care practice areas identified in this study. Nurses

who reported higher levels of preparedness and access to opportunities for

knowledge acquisition were more likely to report engaging in interpersonal practices

that supported patient and family centred decision making and emotional support of

patients and their families. A negative relationship was identified between the

explanatory variables of emotional support for nurses and death anxiety, and the

patient and family centred decision making practice area. Contextual factors had a

limited influence as explanatory variables of specific end-of-life care practice areas.

Gender was identified as a significant explanatory variable in the emotional and

spiritual support practice areas, with male gender associated with lower summated

scores on these practice scales.

Conclusions

Critical care nurses engage in practices to share control with and support inclusion of

families experiencing death and dying. The most frequently identified end-of-life

care practices were those that are easily implemented, practical strategies aimed at

supporting the patient at the end of life and the patient’s family. These practices

arguably require less emotional engagement by the nurse. Critical care nurses’

v

responses reflected values consistent with a palliative approach and a strong

commitment to the inclusion of families in end-of-life care, and these factors were

associated with engagement in all end-of-life care practice areas. Perceived

preparedness or confidence with the provision of end-of-life care was associated with

engagement in interpersonal caring practices. Critical care nurses autonomously

engage in the provision of end-of-life care within the constraints of an environment

designed for curative care and rely on their colleagues for emotional support. Critical

care nurses must be adequately prepared and supported to provide comprehensive

care in all areas of end-of-life care practice. The findings of this study raise important

implications, and informed recommendations for practice, education and further

research.

vi

Contents

Keywords .................................................................................................................... ii Abstract ...................................................................................................................... iii Contents ..................................................................................................................... vi List of Tables .............................................................................................................. x List of Figures .......................................................................................................... xiii List of Abbreviations .............................................................................................. xiv Statement of Original Authorship .......................................................................... xv Grants and Publications ......................................................................................... xvi Acknowledgements ................................................................................................. xvii 1 INTRODUCTION ............................................................................................ 1

1.1 Purpose ............................................................................................................... 5 1.2 Structure of this thesis ........................................................................................ 6

2 LITERATURE REVIEW ................................................................................ 8 2.1 Introduction ........................................................................................................ 8 2.2 Critical care nurses’ end-of-life care practices ................................................... 8

2.2.1 Patient and family centred decision making ............................................... 9 2.2.2 Communication and continuity of care .................................................... 10 2.2.3 Emotional and practical support .............................................................. 11 2.2.4 Symptom management and comfort care .................................................. 12 2.2.5 Spiritual support ....................................................................................... 12

2.3 Factors influencing the provision of end-of-life care ....................................... 13 2.4 End-of-life care in the critical care setting—Context ....................................... 15

2.4.1 Organisational culture ............................................................................. 15 2.4.2 The physical environment ......................................................................... 17 2.4.3 Staffing ...................................................................................................... 18

2.5 End-of-life care in the critical care setting—Evidence .................................... 19 2.5.1 Research ................................................................................................... 19 2.5.2 Nurse practice knowledge ........................................................................ 20 2.5.3 Patient and family preferences ................................................................. 22

2.6 End-of-life care in the critical care setting—Facilitation ................................. 23 2.6.1 Emotional support for nurses ................................................................... 24 2.6.2 Nurses’ end-of-life care values and beliefs .............................................. 25 2.6.3 Coordination of care delivery ................................................................... 25

2.7 Summary........................................................................................................... 27 3 METHODOLOGY ......................................................................................... 28

3.1 Introduction ...................................................................................................... 28 3.2 Mixed methods ................................................................................................. 28 3.3 Ethical considerations ....................................................................................... 32 3.4 Summary........................................................................................................... 33

4 PHASE ONE METHODS .............................................................................. 34 4.1 Introduction ...................................................................................................... 34 4.2 The setting ........................................................................................................ 34 4.3 Participants ....................................................................................................... 35 4.4 Method of data collection ................................................................................. 35 4.5 Procedure .......................................................................................................... 36

vii

4.6 Data management and analysis ........................................................................ 37 4.7 Summary........................................................................................................... 39

5 PHASE ONE FINDINGS............................................................................... 40 5.1 Introduction ...................................................................................................... 40 5.2 Participant characteristics ................................................................................. 40 5.3 Category: Beliefs about end-of-life care .......................................................... 41

5.3.1 Value of participating in end-of-life care ................................................. 42 5.3.2 Complexity of end-of-life care .................................................................. 43

5.4 Category: End-of-life care in the intensive care context .................................. 44 5.4.1 Organisational support ............................................................................. 45 5.4.2 Modifying the bedside environment .......................................................... 46

5.5 Category: Facilitating end-of-life care ............................................................. 49 5.5.1 Uncertainty and ambiguity ....................................................................... 49 5.5.2 Emotional intensity ................................................................................... 50 5.5.3 Comfort care ............................................................................................. 54 5.5.4 Humanising the patient ............................................................................. 55 5.5.5 Facilitating the family’s experience ......................................................... 57

5.6 Discussion......................................................................................................... 65 5.7 Summary........................................................................................................... 66

6 PHASE TWO METHODS............................................................................. 67 6.1 Introduction ...................................................................................................... 67 6.2 Development of the survey instrument............................................................. 67

6.2.1 Critical care nurses’ end-of-life care practices ....................................... 68 6.2.2 Factors influencing the provision of end-of-life care ............................... 70 6.2.3 Demographic information ........................................................................ 72

6.3 Expert panel ...................................................................................................... 72 6.3.1 Participants .............................................................................................. 72 6.3.2 Procedure ................................................................................................. 73 6.3.3 Data management and analysis ................................................................ 74

6.4 Pilot test ............................................................................................................ 74 6.4.1 The setting and sample ............................................................................. 75 6.4.2 Procedure ................................................................................................. 75 6.4.3 Data management and analysis ................................................................ 76

6.5 Summary........................................................................................................... 76 7 PHASE TWO RESULTS ............................................................................... 77

7.1 Introduction ...................................................................................................... 77 7.2 Expert panel ...................................................................................................... 77 7.3 Pilot test ............................................................................................................ 84

7.3.1 Demographic information ........................................................................ 88 7.3.2 Critical care nurses’ end-of-life care practices ....................................... 91 7.3.3 Factors influencing the provision of end-of-life care ............................... 98

7.4 Summary......................................................................................................... 108 8 PHASE THREE METHODS ...................................................................... 110

8.1 Introduction .................................................................................................... 110 8.2 The setting and sample ................................................................................... 110 8.3 Procedure ........................................................................................................ 111 8.4 Data management and analysis ...................................................................... 112

8.4.1 Descriptive statistics ............................................................................... 113 8.4.2 Exploratory factor analysis .................................................................... 114 8.4.3 Bivariate statistical analysis ................................................................... 118

viii

8.4.4 Multivariable statistical analysis ............................................................ 119 8.5 Summary......................................................................................................... 121

9 PHASE THREE RESULTS ........................................................................ 122 9.1 Introduction .................................................................................................... 122 9.2 Phase three participants .................................................................................. 122 9.3 Demographic characteristics .......................................................................... 123 9.4 Descriptive statistics ....................................................................................... 127 9.5 Exploratory factor analysis ............................................................................. 129

9.5.1 Exploratory factor analysis: Critical care nurses’ end-of-life care practices .................................................................................................. 129

9.5.2 Exploratory factor analysis: Factors influencing the provision of end-of-life care ............................................................................................... 136

9.6 Bivariate analysis............................................................................................ 145 9.6.1 Bivariate analysis of critical care nurses’ end-of-life care practices

summated scales and demographic variables ........................................ 147 9.6.2 Bivariate analysis of critical care nurses’ end-of-life care practices

summated scales and factors influencing the provision of end-of-life care summated scales ............................................................................. 153

9.7 Multivariable analysis .................................................................................... 156 9.7.1 Information sharing practices model ..................................................... 156 9.7.2 Environmental modification practices model ......................................... 157 9.7.3 Emotional support practices model ........................................................ 158 9.7.4 Patient and family centred decision making practices model ................ 159 9.7.5 Symptom management practices model .................................................. 160 9.7.6 Spiritual support practices model ........................................................... 162

9.8 Assumption testing ......................................................................................... 166 9.9 Summary......................................................................................................... 167

10 DISCUSSION ................................................................................................ 168 10.1 Introduction ............................................................................................ 168 10.2 What are the most frequently reported end-of-life care practices of

Australian critical care nurses? ....................................................................... 168 10.3 What are the factors influencing the provision of end-of-life care as

perceived by Australian critical care nurses? ................................................. 174 10.3.1 Context ............................................................................................... 174 10.3.2 Evidence ............................................................................................. 177 10.3.3 Facilitation ......................................................................................... 181

10.4 What factors are associated with Australian critical care nurses’ end-of-life care practices? ..................................................................................... 186

10.4.1 Facilitation ......................................................................................... 186 10.4.2 Evidence ............................................................................................. 188 10.4.3 Context ............................................................................................... 190 10.4.4 Demographic explanatory variables .................................................. 191

10.5 Strengths and limitations ........................................................................ 194 10.6 Summary ................................................................................................. 201

11 RECOMMENDATIONS AND CONCLUSIONS ..................................... 203 11.1 Introduction ............................................................................................ 203 11.2 Recommendations for practice ............................................................... 203 11.3 Recommendations for education ............................................................ 206 11.4 Recommendations for research .............................................................. 208 11.5 Conclusions ............................................................................................ 210

ix

12 REFERENCES ............................................................................................. 213 13 APPENDICES............................................................................................... 232

13.1 QUT research ethics approval—Phase one ............................................ 232 13.2 QUT research ethics approval—Phases two and three ........................... 234 13.3 ACT health ethics approval .................................................................... 236 13.4 Email indicating support for phase three of the study from the

Australian College of Critical Care Nurses .................................................... 238 13.5 Phase one bulletin board advertisement ................................................. 239 13.6 Phase one information sheet and consent form ...................................... 240 13.7 Email invitation to experts to participate in the survey review panel .... 242 13.8 Information sheet for expert panel .......................................................... 243 13.9 Phase two pilot test email invitation, email reminder and email retest

invitation ......................................................................................................... 245 13.10 Phase three email invitation and email reminder .................................... 247 13.11 Information sheet attached to email invitation to participate (phase

two pilot test and phase three) ........................................................................ 249 13.12 Final version of survey instrument (Word document version) ............... 251 13.13 Participant responses as a percentage of the sample, means and

standard deviations for each item in the Critical care nurses’ end-of-life care practices section of the survey instrument (n = 392) ............................. 262

13.14 Participant responses as a percentage of the sample, means and standard deviations for each item in the Factors influencing the provision of end-of-life care section of the survey instrument (n = 392) ....................... 265

13.15 Structure matrix for exploratory factor analysis with oblique rotation—Critical care nurses’ end-of-life care practices ............................. 269

13.16 Structure matrix for exploratory factor analysis with oblique rotation—Factors influencing the provision of end-of-life care .................... 272

x

List of Tables

Table 5-1: Phase one participant characteristics ........................................................ 40

Table 5-2: Categories and subcategories developed from phase one interview

data ............................................................................................................ 41

Table 6-1: Comparison of Clarke et al.’s (2003) domains of end-of-life care,

subcategories developed from phase one interview data and initial a

priori scales in the Critical care nurses’ end-of-life care practices

section of the survey instrument ............................................................... 69

Table 6-2: Domains and associated a priori scales in the Factors influencing the

provision of end-of-life care section of the survey instrument ................. 70

Table 7-1: Expert panel members .............................................................................. 78

Table 7-2: Content validity index for each item (I-CVI) in the Critical care

nurses’ end-of-life care practices section of the survey instrument ......... 78

Table 7-3: Content validity index for each item (I-CVI) in the Factors

influencing the provision of end-of-life care section of the survey

instrument .................................................................................................. 80

Table 7-4: Items deleted from the survey instrument following review by the

expert panel ............................................................................................... 83

Table 7-5: Additional items included in the Critical care nurses’ end-of-life care

practices section of the survey instrument ................................................ 85

Table 7-6: Additional items included in the Factors influencing the provision of

end-of-life care section of the survey instrument ...................................... 86

Table 7-7: Demographic characteristics of pilot test participants (n = 28) ................ 89

Table 7-8: Pilot test participants’ current workplace and participation in nursing

work (n = 28) ............................................................................................. 90

Table 7-9: Items in the patient and family centred decision making scale ................ 92

Table 7-10: Items in the explaining and providing information scale ....................... 92

Table 7-11: Items in the continuity of care scale ....................................................... 94

Table 7-12: Items in the emotional and practical support for patients and families

scale ........................................................................................................... 94

Table 7-13: Items in the comfort care—environmental factors scale ........................ 95

xi

Table 7-14: Items in the symptom management and comfort care scale ................... 96

Table 7-15: Items in the spiritual support scale ......................................................... 97

Table 7-16: Internal consistency and test-retest reliability of the a priori scales in

the Critical care nurses’ end-of-life care practices section of the

survey instrument ...................................................................................... 98

Table 7-17: Items in the organisational culture scale ................................................. 99

Table 7-18: Items in the resources scale .................................................................. 100

Table 7-19: Items in the patient and family preferences scale ................................. 102

Table 7-20: Items in the nursing education and experience scale ............................ 103

Table 7-21: Items in the nurse practice knowledge scale ........................................ 104

Table 7-22: Items in the emotional support for nurses scale ................................... 105

Table 7-23: Items in the nurses’ end-of-life care values and beliefs scale .............. 106

Table 7-24: Items in the coordination of care delivery scale ................................... 107

Table 7-25: Internal consistency and test-retest reliability of the a priori scales in

the Factors influencing the provision of end-of-life care section of the

survey instrument .................................................................................... 108

Table 8-1: Items reverse scored prior to factor analysis .......................................... 113

Table 9-1: Demographic characteristics of phase three participants (n = 392) ....... 124

Table 9-2: Phase three participants’ location, workplace and participation in

work (n = 392) ......................................................................................... 125

Table 9-3: The mean and standard deviation for the 10 items in the Critical care

nurses’ end-of-life care practices section most frequently undertaken

based on their mean score (n = 392) ....................................................... 127

Table 9-4: The mean and standard deviation for the 10 most strongly agreed

items in the Factors influencing the provision of end-of-life care

section of the survey instrument based on their mean score (n = 392) ... 128

Table 9-5: Factor loadings for exploratory factor analysis with oblique rotation

of Critical care nurses’ end-of-life care practices section of the survey 132

Table 9-6: Factor correlation matrix of Critical care nurses’ end-of-life care

practices section of the survey instrument .............................................. 135

Table 9-7: Final factors, labels and Cronbach’s alphas for Critical care nurses’

end-of-life care practices section of the survey instrument .................... 136

xii

Table 9-8: Factor loadings for exploratory factor analysis with oblique rotation

of Factors influencing the provision of end-of-life care section of the

survey instrument .................................................................................... 139

Table 9-9: Factor correlation matrix for the Factors influencing the provision of

end-of-life care section of the survey instrument .................................... 143

Table 9-10: Final factors, labels and Cronbach’s alpha for the Factors

influencing the provision of end-of-life care section of the survey

instrument ................................................................................................ 144

Table 9-11: Psychometric properties of the summated scales from the Critical

care nurses’ end-of-life care practices section of the survey

instrument (n = 392) ................................................................................ 145

Table 9-12: Psychometric properties of the summated scales from the Factors

influencing the provision of end-of-life care section of the survey

instrument ................................................................................................ 146

Table 9-13: Spearman’s rho correlation coefficients between outcome variables

and explanatory factors ........................................................................... 154

Table 9-14: Parameter estimates for information sharing model ............................. 156

Table 9-15: Parameter estimates for environmental modification model ................ 157

Table 9-16: Parameter estimates for emotional support model ............................... 158

Table 9-17: Parameter estimates for patient and family centred decision making

model ....................................................................................................... 160

Table 9-18: Parameter estimates for symptom management model ........................ 161

Table 9-19: Parameter estimates for spiritual support model .................................. 162

Table 9-20: Comparison of practice models and contributing parameters .............. 164

xiii

List of Figures

Figure 3-1: Three-phase exploratory sequential mixed methods design (adapted

from Creswell & Plano Clark, 2007) ........................................................ 31

Figure 9-1: Flow chart of participants in phase three .............................................. 123

Figure 9-2: Scree plot of eigen values for factors in the Critical care nurses’ end-

of-life care practices section of the survey instrument ........................... 130

Figure 9-3: Scree plot of eigenvalues for factors in the Factors influencing the

provision of end-of-life care section of the survey instrument ............... 137

Figure 9-4: Histogram of information sharing summated scale............................... 146

Figure 9-5: Boxplot of years of experience in nursing and information sharing

practices .................................................................................................. 148

Figure 9-6: Boxplot of years of experience in nursing and emotional support ........ 149

Figure 9-7: Boxplot of years of experience in nursing and symptom management 151

Figure 9-8: Residual plot of emotional support practices model ............................. 166

xiv

List of Abbreviations

ACCCN Australian College of Critical Care Nurses

AACN American Association of Colleges of Nursing

ANZICS Australian and New Zealand Intensive Care Society

CCU Coronary care unit

CFA Confirmatory factor analysis

CVI Content validity index

DAS Death Anxiety Scale

DoHA Department of Health and Ageing

EFA Exploratory factor analysis

EOL End of life

EOLC End-of-life care

ESHF End stage heart failure

FTE Full-time equivalent

HDU High dependency unit

ICU Intensive care unit

I-CVI Item content validity index

KMO Kaiser-Meyer-Oklin

MSA Measure of sampling adequacy

NHMRC National Health and Medical Research Council

NiPCAS Neonatal Palliative Care Attitude Scale

PAF Principal axis factoring

PARIHS Promoting Action on Research Implementation in Health Services

PCC4U Palliative Care Curriculum for Undergraduates

PCQN Palliative Care Quiz for Nursing

PEPA Program of Experience in the Palliative Approach

S-CVI Scale content validity index

QUT Queensland University of Technology

URL Uniform resource locator

xv

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet

requirements for an award at this or any other higher education institution. To the

best of my knowledge and belief, the thesis contains no material previously

published or written by another person except where due reference is made.

Kristen Ranse

February 2013

xvi

Grants and Publications

ACT Nursing & Midwifery Board 2008, Novice Researcher Grant, AU$3000.

Ranse, K. (2009, October). Intensive care nurses’ end-of-life care practices and

beliefs. Paper presented at the 34th Australian and New Zealand Annual Scientific

Meeting on Intensive Care, Perth, WA.

Ranse, K., Yates, P., & Coyer, F. (2012). End-of-life care in the intensive care

setting: A descriptive exploratory qualitative study of nurses’ beliefs and practices.

Australian Critical Care, 25(1), 4–12.

Ranse, K. (2012, October). Factors influencing the provision of end-of-life care in

critical care settings. Paper presented at the 37th Australian and New Zealand

Annual Scientific Meeting on Intensive Care, Adelaide, SA.

xvii

Acknowledgements

This thesis was edited by Elite Editing, and editorial intervention was restricted to Standards

D and E of the Australian Standards for Editing Practice.

This PhD journey would not have been possible without the assistance and support I

have received from many people. First, I would like to acknowledge my principal

supervisor, Professor Patsy Yates, and my associate supervisor, Dr Fiona Coyer, for

their expert advice, continuing support and encouragement. I greatly appreciate and

value the time and effort that they have contributed to this PhD research project and

my research training during candidature.

I would like to thank the critical care nurses that contributed to each phase of this

study. Without their enthusiasm for this study and their willingness to share their

personal experiences of the provision of end-of-life care in critical care settings, this

research would not have been possible.

Thank you to Dr Laurie Grealish, Dr Jan Taylor and Professor Paul Morrison, who

encouraged me to undertake a PhD and continue to contribute to my development as

a nursing academic. Thanks also to Professor Mary Cruickshank and my colleagues

in the Disciplines of Nursing & Midwifery at the University of Canberra for your

interest in my work and ongoing support.

This PhD would not be possible without the love and support of my husband, Jamie,

and parents, Stuart and Marion Storey. Thanks for all of the emotional and practical

support over the last six years. Finally, I would like to acknowledge my children,

Charlotte, Sadie and Lucas, who have shared their lives so far with Mummy’s PhD. I

hope that I can inspire and share a love of learning with you.

1

1 INTRODUCTION

The term ‘critical care’ encompasses a range of clinical units, including general and

specialty intensive care units (ICUs), combined intensive care/coronary care units

(CCUs), paediatric ICUs, emergency departments and aeromedical retrieval units

(Drennan, Hicks, & Hart, 2010). These units have developed in response to advances

in medicine and technology that have enhanced the ability to save lives and delay

death. Patients are admitted to critical care units with traumatic injury, critical illness

or an acute exacerbation of a chronic illness. Increasingly, patients have multiple

comorbidities creating prognostic difficulties. The potential benefits of critical care

for the patient and their likelihood of survival to discharge are just some of the

admission criteria considered by the medical team in the allocation of critical care

services (Judson & Fisher, 2006). The critical care setting is dominated by a culture

of cure, and critically ill patients are monitored and treated using advanced

technology and specialised care aimed at treating critical illness and saving lives.

Despite this curative focus, death remains an inevitable part of life. A review of the

Australian and New Zealand Intensive Care Society (ANZICS) adult patient database

revealed average overall Australian ICU mortality rates of approximately 10%

(Moran, Bristow, Solomon, George, & Hart, 2008). Internationally, ICU mortality

rates of between 15% and 35% have been reported (Seymour, 2001). The death of a

patient in a critical care unit can occur suddenly during the provision of active

treatment or following the withdrawal or withholding of life-sustaining treatment. An

incidence of death following the withdrawal of life-sustaining treatment of between

49% and 90% has been reported in the literature (M. J. Bloomer, Triuvoipati,

Tsiripillis, & Botha, 2010; Le Conte et al., 2010; Luce, 1997). Thus, despite

difficulties in determining patient prognosis in the critical care patient population,

more than half of all patient deaths in critical care are expected following a decision

to withdraw or withhold life-sustaining treatment. Although critical care is

dominated by a curative focus, the mortality rate of patients results in the provision

of end-of-life care comprising a significant component of nursing work.

2

The National Palliative Care Strategy 2010 (Australian Government Department of

Health and Ageing [DoHA], 2010) acknowledges that definitions of end of life vary

from the final days to the last two years of life (p. 19). In contrast, the definition

proposed by Palliative Care Australia (2008) is not time bound. Instead, Palliative

Care Australia defines end of life as ‘that part of life where a person is living with,

and impaired by, an eventually fatal condition, even if the prognosis is ambiguous or

unknown’ (p. 8). In the critical care environment, in which goals of care are

predominantly focused on cure, limited attention has been given to defining what is

meant by end of life. In practice, however, end of life is often considered the period

of life following a decision to withdraw or withhold life-sustaining treatment, as

death of the patient is an expected outcome. It is this phase of care that is the focus of

the present study.

Previous authors have acknowledged the philosophical tensions that exist between

palliative care and critical care (Davidson et al., 2002; Davidson et al., 2003; Wotton,

Borbasi, & Redden, 2005). In particular, the holistic approach of palliative care can

be at odds with the reductionist curative focus of critical care. Critical care units are

designed and resourced to save lives, and the imposing technological imperative

creates an environment that could disrupt the provision of a natural death

incorporating patient and family centred holistic care. In fact, in this context, death

can be viewed as a professional failure (Beckstrand, Callister, & Kirchhoff, 2006;

Cartwright, Steinberg, Williams, Najman, & Williams, 1997; Halcomb, Daly,

Jackson, & Davidson, 2004; McKeown, Cairns, Cornbleet, & Longmate, 2010). The

provision of end-of-life care to patients and their families in the curative context of

critical care thus presents unique challenges and important opportunities for critical

care nurses engaged in this work. For example, patients are admitted to critical care

units during an acute phase of illness or following traumatic injury, limiting the time

for the critical care nurse to establish a relationship with the patient and the patient’s

family (Beckstrand, Smith, Heaston, & Bond, 2008; LeConte et al., 2010; Nelson &

Meier, 1999). This is further complicated by the fact that up to 95% of critically ill

patients are cognitively impaired as a result of their illness, injury and/or medical

management (Truog et al., 2008). In addition, the length of time between the

withdrawal of life-sustaining treatment and subsequent death of the patient is

relatively short (Efstathiou & Clifford, 2011). In a chart audit undertaken over a 12-

3

month period in an Australian metropolitan ICU, the majority of patients (44%) died

between one and six hours following withdrawal of life-sustaining treatment (M. J.

Bloomer et al., 2010). This is consistent with figures reported in epidemiological

studies in ICUs internationally (Hall, Rocker, & Murray, 2004; Wunsch, Harrison,

Harvey, & Rowan, 2005). Thus, critical care nurses have limited time to establish

relationships and provide comprehensive care to the patient at the end of life and to

the patient’s family. This presents a unique challenge and stressor for critical care

nurses engaged in this work (Efstathiou & Clifford, 2011).

Despite the limited time frame available for the delivery of end-of-life care, critical

care nurses are in a unique position to control the timing and experience of death and

dying for patients and their families (Seymour, 2001). The timing of removal of

treatment and the dying trajectory can be controlled to create a natural death,

whereby the demise of the patient appears unrelated to the removal of treatment

(Seymour, 2001). Thus, death occurs naturally rather than as a result of the actions or

inactions by medical professionals directly involved (Seymour, 2001; Timmermans,

2005). Previous research also suggests that implementation of a decision to withdraw

life-sustaining treatment and the subsequent death of the patient can be delayed to

enable family to accept the situation and/or relatives to gather at the patient’s bedside

(Calvin, Lindy, & Clingon, 2009; Fridh, Forsberg, & Bergbom, 2009; Heland, 2006).

However, these practices can subject the patient to prolonged and unnecessary

suffering and place burden on finite critical care resources (Burns, Jacobs, & Jacobs,

2011; Crump, Schaffer, & Schulte, 2010).

Nurses in critical care tend to work alongside their medical colleagues (Seymour,

2001). They practice with a degree of autonomy and work in a context that supports

high nurse-to-patient ratios. This context presents important opportunities for nurses

to deliver skilled, comprehensive care to patients at the end of life and their families.

Despite the frequency with which nurses provide end-of-life care in critical care units

and the opportunities provided by the critical care practice context, evidence

indicates nurses are not adequately prepared to provide care to patients at the end of

life and their families. In a recent survey completed by 475 critical care nurses in the

United States, only 15.5% of participants reported that their nursing training included

compulsory content pertaining to withdrawal of life-sustaining treatment and 44% of

4

participants indicated that they had attended in-service about withdrawal of life-

sustaining treatment (Kirchhoff & Kowalkowski, 2010). Limited coverage of end-of-

life content in critical care texts has also been reported (Kirchhoff, Beckstrand, &

Anumandla, 2003).

The ageing population combined with the institutionalisation of death in hospitals,

rather than natural death at home, has contributed to a lack of exposure to death and

dying in society. New nurses may have no prior experience with death and dying

until confronted in their workplace with the expectation to care for a patient at the

end of life. Conversely, experienced critical care nurses can develop a blunted

emotional response to death and dying in their work whereby death becomes

normalised through frequent exposure (Badger, 2005b; Shorter & Stayt, 2010).

Inadequate preparation and support of nurses providing end-of-life care, in addition

to the stress associated with the provision of this care, can compromise the quality of

care afforded to patients and their families and contribute to poor retention of critical

care nurses. Identification and development of resources and strategies to prepare

and support nurses in the provision of end-of-life care are needed.

Limited research has been conducted in Australia on end-of-life care in critical care

settings from a nursing perspective. Previous research has explored intensive care

nurses’ perceptions of medical futility and their role in decision making and

withdrawal of treatment (Heland, 2006) and the lived experience of intensive care

nurses caring for patients having treatment withheld or withdrawn (Halcomb et al.,

2004; Jones & FitzGerald, 1998). Each of these studies was undertaken from a

qualitative approach, using small non-probability samples. These studies provide

some insight into the issues but the ability to generalise the findings beyond the

settings of the studies is limited. More recently, the findings of audits of medical

records of patients who had died in an Australian ICU have been published (M.

Bloomer, Lee, & O’Connor, 2010; M. J. Bloomer et al., 2010). The audits described

the frequency of not-for-resuscitation orders, attendance at family meetings,

treatment withdrawal and the specific interventions withdrawn. While these studies

provide important information regarding end-of-life care practices in Australian

ICUs, the use of chart audit as the primary source of data is limited because it

assumes that such practices are consistently documented by care providers. Indeed,

5

the authors noted the absence of documentation regarding communication between

the nurse and family in their setting (M. Bloomer et al., 2010). Similarly, a

documentation audit undertaken in the United States found limited presence of

critical care nurses at the end of life, but noted that it was likely due to a lack of

nursing documentation rather than a lack of nursing care (Kirchhoff, Anumandla,

Foth, Lues, & Gilbertson-White, 2004). Further research to identify the actual end-

of-life care practices of Australian critical care nurses using alternative methods is

needed.

Internationally, qualitative research exploring various aspects of critical care nurses’

experiences of end-of-life care has been reported (Calvin et al., 2009; Espinosa,

Young, Symes, Haile, & Walsh, 2010; Fridh et al., 2009; McKeown et al., 2010;

McMillen, 2008; Vanderspank-Wright, Fothergill-Bourbonnais, Brajtman, &

Gagnon, 2011). In addition, research conducted in the United States using both

qualitative and quantitative approaches provides insight into critical care nurses’

perceptions of barriers to and facilitators of end-of-life care (Beckstrand et al., 2006;

Beckstrand & Kirchhoff, 2005; Kirchhoff & Beckstrand, 2000; Nelson et al., 2006).

However, several differences exist between the critical care environments of the

United States and Australia, including in nursing roles, staffing structure and patient

acuity (Ogle, Copley, Bethune, & Parkin, 2004).

To evaluate and improve the care that dying patients and their families receive and to

support nurses in the provision of optimal end-of-life care, a better understanding of

the end-of-life care practices of critical care nurses and the factors influencing these

practices is required.

1.1 Purpose

The purpose of this three-phase exploratory sequential mixed methods study was to

identify the end-of-life care practices of Australian critical care nurses, the factors

influencing the provision of end-of-life care and the factors associated with specific

end-of-life care practices. Phase one involved a qualitative exploration of the end-of-

6

life care experiences and practices of critical care nurses through semi-structured

interviews at an Australian tertiary teaching hospital. The findings of phase one,

combined with a review of the literature, contributed to the development of a survey

instrument in phase two. The survey instrument was reviewed by an expert panel and

pilot tested with a convenience sample of critical care nurses to obtain preliminary

evidence of the reliability and validity of the tool. The quantitative phase of this

study, phase three, utilised the survey instrument to survey a national sample of

critical care nurses to identify the frequency of select end-of-life care practices and

the factors influencing these practices.

Specifically, the key research questions in this study are:

1. What are the most frequently reported end-of-life care practices of Australian

critical care nurses?

2. What are the factors influencing the provision of end-of-life care as perceived

by Australian critical care nurses?

3. What factors are associated with Australian critical care nurses’ end-of-life

care practices?

1.2 Structure of this thesis

This chapter has provided a brief introduction to the key issues surrounding end-of-

life care in the critical care context and significance of this study to nursing practice.

The purpose of this study was identified and three key research questions presented.

Chapter Two provides a critical appraisal of nursing and sociological literature

related to end-of-life care in critical care contexts. Clarke et al.’s (2003) domains of

end-of-life care and the Promoting Action on Research Implementation in Health

Services (PARIHS) framework (Kitson, Harvey, & McCormack, 1998; Rycroft-

Malone et al., 2002) are discussed and then used as conceptual frameworks to

organise the literature reviewed and presented in Chapter Two.

7

Mixed methods is introduced in Chapter Three as the methodological approach used

in this study. The key strengths and criticisms of mixed methods are discussed and

the three-phase exploratory sequential mixed methods design utilised in this study is

described. Chapter Three concludes following a discussion of the ethical

considerations pertinent to the study.

Chapter Four describes the methods used in phase one of this study to interview a

small sample of critical care nurses. The findings of analysis of phase one interviews

are presented in Chapter Five.

Chapter Six describes the methods used in phase two of the study to develop and test

a survey instrument. Chapter Seven documents the results of phase two that provide

preliminary evidence of the validity and reliability of the survey instrument.

Phase three consisted of a national survey of critical care nurses’ end-of-life care

practices and the factors influencing the provision of end-of-life care using the

survey instrument developed in phase two. Chapter Eight describes phase three

methods, including the sample and setting, procedure, data management and

analysis. The results of descriptive statistics, exploratory factor analysis (EFA) and

multivariable analysis of data obtained from responses to the national survey are

presented in Chapter Nine.

Chapter Ten provides a discussion of the key findings of this study pertaining to each

research question. This chapter concludes following a discussion of the strengths and

limitations of this study. The findings of this study raise a number of important

implications. Recommendations for practice, education and research are outlined and

concluding remarks are presented in Chapter Eleven.

8

2 LITERATURE REVIEW

2.1 Introduction

This chapter provides a review of nursing, critical care and sociological literature

related to end-of-life issues in critical care contexts to provide a contextual base for

this study and situate the contribution of this study within existing nursing

knowledge and practice. The chapter will first provide a critical appraisal of the

literature pertaining to the end-of-life care experiences and practices of critical care

nurses. Clarke et al.’s (2003) six domains of end-of-life care will be introduced and

then used as a guiding framework to present empirical evidence related to critical

care nurses’ end-of-life care practices. The PARIHS framework (Kitson et al., 1998;

Rycroft-Malone et al., 2002) will then be described, and its role as a conceptual

framework to identify the factors that influence the provision of end-of-life care in

critical care settings will be discussed. Each of the key elements of the PARIHS

framework—context, evidence and facilitation—will then be defined and a critical

review of the literature for each element as it relates to the provision of end-of-life

care in critical care settings will be presented.

2.2 Critical care nurses’ end-of-life care practices

Few studies have explored actual nursing practices undertaken in the provision of

end-of-life care in the critical care setting. Clarke et al. (2003), using a modified

Delphi approach involving 36 expert clinicians in the United States, identified seven

domains of end-of-life care and their associated quality indicators for the intensive

care setting. The seven domains identified were patient and family centred decision

making, communication, continuity of care, emotional and practical support,

symptom management and comfort care, spiritual support, and emotional and

organisational support for intensive care clinicians. The domains and quality

indicators were then reviewed by 15 nurse-physician teams to identify interventions

and behaviours that could be implemented to improve these indicators (Clarke et al.,

2003). Although Clarke et al.’s findings provide some insight into the end-of-life

care practices of critical care nurses, the sample was limited to participants

9

considered clinical experts, and empirical testing and validation of behaviours

identified with larger samples was not undertaken. The practice of expert nurses

differs to those of nurses with less experience in and knowledge of end-of-life care.

In an Australian study that compared the practice behaviours of specialist intensive

care nurses with those of generalist nurses working in intensive care, a qualitative

difference was noted in the knowledge and practice of these nurses (Fairweather &

Gardner, 2000).

The domains of end-of-life care identified by Clarke et al. (2003) broadly identify

relevant areas of critical care nurses’ end-of-life care practices. As such, the domains

were selected as a framework to present the literature review related to end-of-life

care practices for this thesis and inform the development of the survey instrument.

Specifically, a critical review of the literature related to critical care nurses’ end-of-

life care experiences and practices was undertaken and is presented as it relates to the

domains of patient and family centred decision making, communication, continuity

of care, emotional and practical support, symptom management and comfort care,

and spiritual support. The domain emotional and organisational support for intensive

care clinicians will not be addressed in this section as it does not relate directly to the

care of the patient; however, this domain will be addressed later in this chapter when

the factors influencing the provision of end-of-life care are considered.

2.2.1 Patient and family centred decision making

Although most people state that their final wish is to die at home, the reality is that

most die in a hospital or other health care institution (Kellehear, 2008). Critically ill

and dying patients are admitted to acute care hospitals, where a cascade of

intervention can occur in an attempt to treat illness, postpone death and prolong life.

The institutionalisation of death and dying through admission of critically ill and

dying patients to hospital affords health care professionals increased social control

over these processes (Hart, Sainsbury, & Short, 1998). Practices related to patient

and family centred decision making can assist critical care nurses to share control of

death and dying with the patient and family.

10

Position statements published by the ANZICS (2003) and the American College of

Critical Care Medicine (Truog et al., 2008) provide support for the participation of

the patient and family in decision making. The ANZICS position statement

acknowledges that, when considering decisions to withhold or withdraw treatment,

the preference of the patient (if known) or their likely preference as expressed by

their next of kin should be considered. The ANZICS position statement does not

specifically address the role of the nurse or the care to be provided following the

withholding or withdrawal of treatment except to state that an alternative, comfort-

focused care plan should be implemented. The statement published by the American

College of Critical Care Medicine provides recommendations for the withholding

and withdrawal of treatment, the provision of end-of-life care and the continued care

of the family following the death of the patient (Truog et al., 2008). This statement

(Truog et al., 2008) supports the participation of the patient and family in decision

making regarding withholding and withdrawing treatment and the subsequent care

provided, highlighting family-centred care as a ‘comprehensive ideal for managing

end-of-life care’ (p. 953). The inclusion of the patient and family in decision making

regarding withholding or withdrawing life-sustaining treatment and the provision of

end-of-life care, as well as their involvement in the provision of this care for the

patient at the end of life is important (Kirchhoff et al., 2000) and can influence the

experience of end-of-life care for the patient and family. Patient and family centred

decision making is thus accepted as an important component of care.

2.2.2 Communication and continuity of care

Communication between health care providers and patients at the end of life and

their families and continuity of care are important practices in the provision of end-

of-life care. Providing and explaining information to patients and their families have

been identified in previous research as practices undertaken by critical care nurses

(Borbasi, Wotton, Redden, & Chapman, 2005; Fridh et al., 2009; McClement &

Degner, 1995). However, the actual content of this information can vary. Fridh et al.

(2009) found that critical care nurses communicated with families about the

withdrawal of treatment and dying process, yet communication with patients was on

11

a more superficial level, such as who was in the room and the physical nursing care

they were receiving. Although continuity of care has been cited as a component of

good end-of-life care (Hov, Hedelin, & Athlin, 2007; Kirchhoff et al., 2000) and lack

of continuity as a barrier to the provision of end-of-life care (Kirchhoff et al., 2000;

Vanderspank-Wright et al., 2011), little discussion of the practices that comprise and

support continuity of care for patients at the end of life and their families in the

critical care context have been provided.

2.2.3 Emotional and practical support

The provision of emotional and practical support to family members has been

recognised as important to quality end-of-life care (Halcomb et al., 2004; Kirchhoff

& Beckstrand, 2000; Puntillo et al., 2001). Halcomb et al. (2004) found that during

the provision of end-of-life care, the presence of family shifted the nurse’s attention

from the patient to the family. Nurses provide emotional and practical support to

patients and their families just by being there, providing a physical presence (Hodde,

Engelberg, Treece, Steinberg, & Curtis, 2004; Thompson, McClement, & Daeninck,

2006) and providing support and time for families to be alone with the patient after

death (Beckstrand & Kirchhoff, 2005; Bloomer, Morphet, O’Connor, Lee, &

Griffiths, in press). However, these practices could be constrained by the pre-existing

relationship between the nurse and family, and the nurse’s personal comfort with the

provision of end-of-life care.

In addition to physical presence, emotional and practical support can be provided

through interventions designed to create positive experiences and memories for the

families of a patient at the end of life. Although reflective accounts have been

published in the literature documenting nurses’ engagement in memory making,

positive experiences for patients at the end of life and their families (Faas, 2004), few

empirical studies have reported this area of practice (Coombs, 2010; Hansen,

Goodell, DeHaven, & Smith, 2009; Vanderspank-Wright et al., 2011). Recently,

Vanderspank-Wright et al. (2011) identified a theme providing memories in which

12

nurses organise or create positive and memorable experiences for the patient and the

patient’s family.

2.2.4 Symptom management and comfort care

Clarke et al.’s (2003) symptom management and comfort care domain incorporates

the physical care of the patient at the end of life. Maintaining patient dignity and

managing patient pain are two factors that have been identified as important to end-

of-life care in previous studies (Beckstrand et al., 2006; Halcomb et al., 2004;

Kirchhoff & Beckstrand, 2000; Kirchhoff et al., 2000). In a national survey of a

stratified random sample of 906 American critical care nurses’ (response rate 30%)

knowledge, beliefs and practices related to end-of-life care, 78% of respondents

believed that patients sometimes or frequently received inadequate pain management

(Puntillo et al., 2001). Similarly, an Australian study of 17 registered nurses’

perceptions of providing end-of-life care to patients with end stage heart failure

(ESHF) in acute care and community settings, difficulty in managing pain and

breathlessness was identified (Wotton et al., 2005).

Other interventions such as positioning, pressure area care and hygiene practices

including mouth care and eye care are important practices in the provision of comfort

care to the patient at the end of life (Latour, Fulbrook, & Albarran, 2009; Puntillo et

al., 2001). Moreover, the continued use of interventions that do not provide comfort

such as cardiac monitoring must be reviewed (Mulligan, 2005; Truog et al., 2008). In

the generally short period between withdrawal of treatment and death of the patient,

critical care nurses must provide comprehensive symptom management and comfort

care to meet the physical care needs of the patient at the end of life.

2.2.5 Spiritual support

Spiritual support is widely recognised as a component of holistic care and is included

as a domain of care in Clarke et al.’s (2003) framework. Exley (2004) states that the

13

provision of spiritual support is ‘one of the most difficult “needs” to assess and meet’

(p. 117), with today’s multicultural society and the large number of religions

contributing to the challenge of providing spiritual support. The practices undertaken

by critical care nurses in the provision of spiritual care to patients at the end of life

and their families were not identified in the literature reviewed.

2.3 Factors influencing the provision of end-of-life care

In addition to identifying the current end-of-life care practices of critical care nurses,

this study aimed to identify the factors influencing the provision of end-of-life care,

and the relationships between these factors and specific end-of-life care practices.

Due to the limited research in this field and the broad range of social and personal

factors that could potentially influence end-of-life care practices, an overarching

framework was used to guide and organise the literature review to identify the

important variables to be examined to achieve this aim. Specifically, the PARIHS

framework was used to capture the complexity of context, evidence and facilitation

factors that can influence nursing practice.

The PARIHS framework describes the factors influencing the uptake of evidence in

clinical practice (Kitson et al., 1998). The PARIHS framework consists of three key

elements: context, evidence and facilitation (Rycroft-Malone et al., 2002). In

comparison to linear practice models in which component factors are considered in

isolation, the authors of this framework suggest that each element should be

considered simultaneously, acknowledging the interdependence of the elements and

capturing the complexity and reality of clinical practice (Kitson et al., 1998). Each

element is measured on a high–low continuum. Most successful implementation of

uptake of evidence or a specified change in practice occurs when each element is

rated as high. When an element is rated as low, the influence of this element may be

overcome if other elements are rated highly (Kitson et al., 1998).

An assumption implicit in this conceptual framework is that implementation of a

specified change in practice—the outcome variable—will improve the quality of

14

patient care. In the context of the current study, the outcome variable is critical care

nurses’ end-of-life care practices as described by nurses themselves. This study is

limited, therefore, to the perspective of the critical care nurses providing this care. It

does not consider the perspectives of the recipients of this care, or attempt to define

good end-of-life care.

A further assumption of the framework is that each of the elements and subelements

defined in the framework are discrete yet equally important variables influencing

practice (Kitson et al., 1998; Kitson et al., 2008). Since the framework was originally

developed, conceptual analyses based on a critical review of the literature has been

undertaken on each element of the framework (Harvey et al., 2002; McCormack et

al., 2002; Rycroft-Malone et al., 2002). The findings of these analyses and a critical

synthesis of the published literature pertaining to the PARIHS framework (Helfrich

et al., 2010) indicate that some elements and subelements within the framework lack

conceptual clarity and that overlap between subelements exists. Although the

framework’s original authors have suggested that the key elements of the framework

are interdependent (Kitson et al., 1998), schematic representation of the elements on

a dichotomous continuum alludes to linearity, rather than multidimensionality

(Helfrich et al., 2010). Reported use of the PARIHS framework in empirical research

mainly includes the use of the key elements of the framework as a guide to change

practice or evaluate changes in practice post hoc (Helfrich et al., 2010; Rycroft-

Malone et al., 2002). Further empirical research is needed to test this conceptual

framework (Rycroft-Malone, Seers, et al., 2004). Specifically, testing of the

relationships between the elements of the framework is required to understand the

importance of the individual elements and the relationships between these elements

(Rycroft-Malone et al., 2002). In the current study, each of the elements and

subelements (that is, factors influencing the provision of end-of-life care) are

examined as discrete concepts. However, the relationships between these factors and

the relative importance of these factors from the perspective of critical care nurse

participants are examined through multivariable analyses.

The key elements of the PARIHS framework—context, evidence and facilitation—

thus provide an organising framework to capture the complexity of factors that can

influence nursing practice. In the following section, a definition for each of these

15

elements and the relevance of each element to the study of factors influencing the

provision of end-of-life care in critical care settings is examined.

2.4 End-of-life care in the critical care setting—Context

In the PARIHS framework, the element context includes all aspects of the setting

where ‘practice takes place’ (McCormack et al., 2002, p. 94). Conceptual analysis of

the element context identified culture, the physical environment, leadership and

evaluation of performance as factors contributing to the context of the practice

setting (McCormack et al., 2002; Rycroft-Malone et al., 2002). The critical care

setting is a practice context that has a number of unique features. Three specific

aspects of the context that could influence the provision of end-of-life care by nurses

in the critical care setting are discussed below: organisational culture, the physical

environment and staffing.

2.4.1 Organisational culture

The organisational culture of critical care is dominated by a curative focus, yet

practice is surrounded in uncertainty and ambiguity related to patient prognosis and

treatment decisions. Patients can be admitted to critical care units for a trial of life-

sustaining treatment, sparking a cascade of interventions implemented through

decisions by the medical team. The decision to withdraw these interventions appears

more difficult, requiring consensus among the health care team and between the

health care team and the patient’s family (Timmermans, 2005).

Critical care nurses can experience stress and dissatisfaction related to the curative

treatment of a patient that they perceive will not recover. Heland (2006) interviewed

seven intensive care nurses to obtain their perceptions and experiences of medical

futility. The findings of the study identified that medical futility was not a concept

that could easily be defined and nurses participating in the provision of treatment

16

deemed futile may experience moral distress and frustration (Heland, 2006).

Similarly, in a survey of problems with decision making by critical care nurses, the

continuation of curative treatment when the patient is dying (poor prognosis) was

identified as a major source of dissatisfaction (Bucknall & Thomas, 1997). For some,

the provision of futile treatment is considered inflicting harm on the patient

(Halcomb et al., 2004). Other studies have also captured nurses’ expressions of the

provision of futile treatment through use of emotive terms by participants such as

torturing and flogging the patient (Badger, 2005a; Halcomb et al., 2004; Kirchhoff et

al., 2000; McKeown et al, 2010).

Despite this curative focus, death remains an inevitable part of life. The death of a

patient can be viewed as a professional failure by medical staff in a culture focused

on cure (Beckstrand et al., 2006; Cartwright et al., 1997). Halcomb et al. (2004) also

identified the perception of death as failure by intensive care nurses in the findings of

their study. Perceived professional failure following the death of a patient and the

provision of futile care have been identified as possibly contributing to stress and

burnout among critical care nurses (Badger, 2005a; Borbasi et al., 2005; Cartwright

et al., 1997; Davidson et al., 2002; Halcomb et al., 2004; Heland, 2006).

Qualitative studies undertaken in Australia have reported the exclusion of nurses

from decision making related to withdrawal of life-sustaining treatment (Halcomb et

al., 2004) and in some units, unilateral decision making by medical staff (Heland,

2006). Nurses who are experienced and able to clearly articulate their thoughts are

more likely to contribute to the decision-making process (Heland, 2006). Despite a

general lack of participation in the decision-making process, when a decision is made

to withdraw or withhold life-sustaining treatment, critical care nurses report that they

are left to manage the withdrawal of treatment and care of the patient at the end of

life (Halcomb et al., 2004; Heland, 2006). These perceptions are supported by the

findings of a survey of critical care nurses in the United States in which 45% of

participants identified that doctors were rarely or never present at the patient’s

bedside during treatment withdrawal and 30.3% of participants identified that doctors

were sometimes present (Kirchhoff & Kowalkski, 2010).

17

2.4.2 The physical environment

The physical environment of the practice setting is important to the delivery of

nursing care. The provision of end-of-life care in non-palliative care settings such as

acute medical and cardiac wards has been identified as being difficult (Nordgren &

Olsson, 2004; Wotton et al., 2005), as these settings are designed to support the care

of acutely ill rather than dying patients. Generally, the physical environment of

critical care units is designed to be open plan to permit the observation of critically ill

patients and improve staff access (Bloomer et al., in press). Critical care units are

noisy environments with constant activity. The health care team and ancillary staff,

the provision of patient care and technological interventions utilised in monitoring

and treating critically ill patients are some of the factors that contribute to noise

(Christenson, 2007; Monsen & Edell-Gustafsson, 2005) and activity levels.

Several studies have identified lack of space and privacy as barriers to end-of-life

care by nurses in acute and critical care settings (Bloomer et al., in press; Borbasi et

al., 2005; Calvin et al., 2009; Fridh et al., 2009; Jackson, Purkis, Burnham, Hundt, &

Blaxter, 2010; McKeown et al., 2010; Thompson et al., 2006). The care of a dying

patient in a single room has been identified as important in maintaining patient

dignity, providing privacy for the patient and the patient’s family and allowing

unrestricted visiting. However, placing patients in a single room can isolate them

from nursing staff and pose a barrier to nurses being with dying patients (Davidson et

al., 2003). The availability of facilities for use by the family, such as a quiet room to

congregate, has also been identified as important (Kirchhoff et al., 2000), yet these

facilities are often lacking in critical care units (Bloomer et al., in press). In addition,

previous Australian research has identified that non-palliative care areas lack

appropriate physical resources such as specialised mattresses and pillows, to provide

optimal care for dying patients (Davidson et al., 2003).

18

2.4.3 Staffing

Staffing is another aspect of the practice context that is unique in the critical care

setting. High nurse-to-patient ratios exist in critical care settings in Australia. These

patient-care ratios are supported by position statements from both the Australian

College of Critical Care Nurses (ACCCN, 2003) and the Joint Faculty of Intensive

Care Medicine (JFICM, 2003) that outline that ventilated patients require a nurse-to-

patient ratio of 1:1 and patients classified as high dependency are cared for in a 1:2

nurse-to-patient ratio. These high nurse-to-patient ratios provide a staffing

environment in critical care in which patients at the end of life and their families can

receive comprehensive care and support (Campbell, 2006; McKeown et al., 2010).

Lack of staff and time constraints have been identified as barriers to end-of-life care

in acute and critical care settings (Beckstrand et al., 2006; Thompson et al., 2006). In

acute medical wards, lack of staff can compromise nurse-patient ratios if patients that

are dying are not afforded the usual staffing ratio (Thompson et al., 2006). Ciccarello

(2003) recommends that quality care of a dying patient requires a full-time nurse and

warns against nurse managers allocating care of the dying patient last, based on the

belief that they require little care. Other authors support this view, highlighting that

palliative care is not implemented when ‘nothing more can be done’ (Thelan, 2005,

p. 34), but requires extensive nursing care focused on the holistic needs of patients

and their families (Davidson et al., 2002; Rushton, Williams, & Sabatier, 2002;

Thelan, 2005).

Although staffing levels in critical care are supported by position statements, critical

care units in Australia are experiencing nursing staff shortages and poor retention of

experienced nurses (Ogle et al., 2004; Morrison, Beckmann, Durie, Carless, &

Gillies, 2001). The staffing skill mix can result in the allocation of casual or junior

nurses to care for the dying patient while experienced nurses are allocated care of

critically ill patients (Beckstrand et al., 2006). This allocation could be influenced by

the curative culture of critical care with care of the patient at the end of life provided

by a junior nurse to enable experienced critical care nurses to focus on the care of

patients whose lives could be saved.

19

2.5 End-of-life care in the critical care setting—Evidence

Although some nurses equate evidence with research (Rycroft-Malone, Harvey, et

al., 2004), within the PARIHS framework, the evidence element is broadly defined to

include all types of evidence that inform clinical practice, including research

evidence derived from scientific studies, clinicians’ practice knowledge and the

patient’s experience and preferences (Rycroft-Malone, Seers, et al., 2004). Evidence

drawn from these different sources can be conflicting and it is important that

evidence be critically appraised prior to changing or implementing the evidence in

practice (Rycroft-Malone, Seers, et al., 2004). Research evidence, nurse practice

knowledge and patient and family preferences are sources of evidence that inform

critical care nurses’ end-of-life care practices.

2.5.1 Research

Guidelines and policies developed from research evidence or expert opinion are

important in maintaining minimum standards and providing guidance for clinical

practice. The ANZICS (2003) position statement on withholding and withdrawing

treatment, the NSW Department of Health’s (2005) Guidelines for End-of-Life Care

and Decision Making and Palliative Care Australia’s (2005) Standards for Providing

Quality Palliative Care for all Australians are three documents that provide guidance

for critical care nurses regarding end-of-life decision making and the provision of

end-of-life care. Although the ANZICS position statement is specific to the intensive

care setting, it only provides broad statements around the withdrawal and

withholding of treatment and does not specifically address the role of the critical care

nurse or provide guidance for the management of a patient at the end of life. The

NSW Department of Health Guidelines and the Palliative Care Australia standards

are broad high-level standards developed to guide end-of-life decision making and

end-of-life care across community and acute settings. In addition, the Australian

Government National Health and Hospitals Reform Commission (2009) provides

broad recommendations regarding end-of-life care, supporting the provision of end-

of-life care, including specialist palliative care across all care settings. Thus, end-of-

20

life care needs to be available where people need it. Critical care is a unique practice

context and further specific guidance, including best practice guidelines and policies,

to support the provision of end-of-life care in critical care settings is required

(Davidson et al., 2002; Halcomb et al., 2004).

From an organisational perspective, the development of policies and guidelines that

are contextually relevant at a local level are also important in maintaining minimum

standards and improving the quality of care provided. Policies and guidelines can

assist clinicians by providing prompts or cues for their practice. In two separate

interventional studies that utilised a standardised order form to guide end-of-life care,

nurses reported satisfaction with the form in guiding their practice, although no

change in the quality of end-of-life care was identified in the findings (Hall et al.,

2004; Treece et al., 2004).

Previous research has identified lack of guidelines as a barrier to providing end-of-

life care (Nordgren & Olsson, 2004). In a study undertaken in 15 ICUs in North

America, all policies, protocols, guidelines, standing orders and forms for

documenting end-of-life care were analysed to identify content and how they could

be altered to improve the quality of end-of-life care (Clarke et al., 2004). Clarke et al.

(2004) found that the content of most documents was related to symptom

management and comfort care and few documents considered patient and family

decision making, communication, emotional support or spiritual support. In a recent

survey of a random sample of critical care nurses undertaken in the United States, no

use of clinical practice guidelines was reported by participants (Kirchhoff &

Kowalkowski, 2010).

2.5.2 Nurse practice knowledge

The Australian Council of Healthcare Standards Guidelines for Intensive Care Units

(ACHS, 1997) state that a minimum of 50% of registered nurses employed in an ICU

21

should have postgraduate critical care qualifications. The ACCCN (2003) advocate

that 75% of nurses with qualifications is ideal. Although the majority of nurses

employed in critical care units hold postgraduate qualifications, knowledge of end-

of-life care can be limited. Campbell (2002) argues that, despite the frequency with

which nurses in critical care units provide care for patients at the end of life, critical

care courses maintain a curative focus and end-of-life content, including comfort

care of the dying patient, is largely absent. This view is supported by a content

analysis of graduate nursing education curricula in the United States that identified a

lack of end-of-life care and/or palliative care content in graduate programmes (Paice

et al., 2006). Paice et al. (2006) suggest that lack of content of end-of-life care in

graduate programmes affects the quality of end-of-life care. Although specific

research has not been undertaken in Australia in regard to end-of-life care content in

postgraduate critical care curricula, a study of critical care courses provided in higher

education institutions in Australia identified inconsistencies in course content (both

practical and theoretical) (Aitken, Currey, Marshall, & Elliott, 2006).

In addition to postgraduate education, continuing professional development and

critical care textbooks are key resources used by critical care nurses to enhance their

knowledge base (Kirchhoff et al., 2003; Paice et al., 2006). Continuing professional

development opportunities in critical care settings focus on curative interventions

and lack end-of-life content. Kirchhoff and Kowalkowski (2010) reported that only

33.8% of critical care nurses that responded to their survey identified content related

to withdrawal of life-sustaining treatment in their critical care unit orientation

program and 44% of respondents reported attendance at in-service related to

withdrawal of life-sustaining treatment.

A content analysis of critical care nursing texts using the American Association of

Colleges of Nursing (AACN) end-of-life content areas as a framework revealed that

end-of-life care content was also lacking in texts (Kirchhoff et al., 2003). Ethical and

legal issues, brain death and organ donation were content most commonly covered in

texts; however, less than half of the 14 texts reviewed covered these topics. Three of

the texts reviewed did not include any content related to end-of-life care. Similarly,

in the ACCCN’s Critical Care Nursing textbook (Elliott, Aitken, & Chaboyer, 2012)

apart from an organ donation chapter, end-of-life care content is limited to just a few

22

pages that provide a broad overview of key issues in end-of-life care for the patient,

family and nurse and a discussion of end-of-life decision making. The lack of content

in regard to end-of-life care reflects the dominance of the medical discourse and

curative focus of critical care and can influence critical care nurses’ end-of-life care

knowledge and associated practices.

Nurses can also draw on their personal knowledge and experience to inform their

practice. Several studies have identified that nurses learn through personal

experience in the provision of end-of-life care (Fridh et al., 2009; Zomorodi & Lynn,

2010b). Beckstrand and Kirchhoff (2005) reported that nurses draw on their own

personal experiences of death and dying of a family member or friend to inform their

practice. Others studies have identified a lack of experience in caring for patients at

the end of life and a lack of knowledge as barriers to end-of-life care (Davidson et

al., 2003; Nordgren & Olsson, 2004; Rodriguez, Barnato, & Arnold, 2007;

Thompson et al., 2006).

2.5.3 Patient and family preferences

Patient (and family) preferences should be considered when reviewing evidence to

inform practice in the clinical setting (Kitson et al., 1998). Although the patient’s

preference is considered important evidence informing practice, it can be difficult to

ascertain the preferences of critically ill patients due to cognitive impairment as a

result of illness, injury or medical management. In these circumstances, the health

care team seeks an understanding of the patient’s potential view, as expressed by the

patient’s next of kin or significant other (ANZICS, 2003). Once consensus is reached

among the medical team to withhold or withdraw treatment, agreement of the next of

kin or legal authority is sought (ANZICS, 2003). Although the burden of end-of-life

decision making should not rest with the family, some participants in Halcomb et

al.’s (2004) study perceived that families are sometimes over involved in decision

making and feel responsible for any decisions made.

23

Several studies have identified that families misunderstand the meaning of the

patient’s illness (Badger, 2005b) and life-support/life-saving measures (Beckstrand

& Kirchhoff, 2005; Kirchhoff & Beckstrand, 2000). In addition, lack of acceptance

of a patient’s prognosis by family members and family requesting additional

treatment despite the patient’s wishes pose barriers to end-of-life care. It has been

suggested that keeping families informed of the patient’s condition, treatment plan

and possible outcomes of treatment options facilitates transition to end-of-life care

(Badger, 2005b; Borbasi et al., 2005; Norton, Tilden, Tolle, Nelson, & Eggman,

2003).

The concept of timing in relation to withdrawal of life-sustaining treatment and the

family’s preferences and experience has also been noted in the literature (Calvin et

al., 2009; Fridh et al., 2009; Heland, 2006; Jones & Fitzgerald, 1998; Seymour,

2001) Previous research suggests that implementation of withdrawal of life-

sustaining treatment and the subsequent death of the patient can be delayed to enable

family to accept the situation and/or relatives to gather at the patient’s bedside

(Calvin et al., 2009; Fridh et al., 2009; Heland, 2006). However, this practice can

subject the patient to prolonged and unnecessary suffering and place burden on finite

critical care resources (Burns et al., 2011; Crump et al., 2010).

2.6 End-of-life care in the critical care setting—Facilitation

Facilitation within the PARIHS framework is narrowly defined as ‘a technique by

which one person makes things easier for others’ (Rycroft-Malone et al., 2002, p.

177). This definition focuses on the role of one person, an appointed facilitator, in

supporting others to change practice through processes of ‘enabling and helping

rather than telling or persuading’ (Harvey et al., 2002, p. 585). Helfrich et al. (2010)

identifies that the limitation of this definition of facilitation is that it does not capture

other possible facilitators of practice change, including interventions that may

support implementation of a specified practice. In a conceptual analysis of the

element facilitation, Harvey et al. (2002) found varied descriptions and uses of the

term facilitation in the existing health literature. For the purpose of the current study,

24

facilitation was more broadly defined as the process of making things easier. The

transition from curative to comfort-only care and the provision of end-of-life care in

the critical care setting may be facilitated by the emotional support provided to the

critical care nurse, the beliefs and values of the nurse and factors influencing the

coordination of care delivery.

2.6.1 Emotional support for nurses

Nurses in critical care units are frequently exposed to human suffering and death not

experienced in other workplaces. Previous research has identified coping strategies

used by critical care nurses (Badger, 2005a; Espinosa et al., 2010). In a recent study

involving interviews with eight critical care nurses, Shorter and Stayt (2010)

identified that critical care nurses were complacent towards death and they suggest

that this may be a coping strategy for these nurses. Alternatively, nurses who are not

comfortable with the provision of end-of-life care can cope by focusing on technical

aspects of care, avoiding interpersonal interaction and compromising the care

provided to the patients and their families (Badger, 2005a; McClement & Degner,

1995; Shorter & Stayt, 2010). For a nurse to be there for patients at the end of life

and their families, they need to feel comfortable with being there themselves

(Thompson et al., 2006).

The provision of end-of-life care in critical care units can contribute to stress and

burnout among critical care nurses (Shorter & Stayt, 2010). In addition to personal

coping strategies, critical care nurses can seek emotional support from informal and

formal support mechanisms in the workplace. Several studies have identified that

critical care nurses draw on informal support provided by their colleagues (Halcomb

et al., 2004; Jones & Fitzgerald, 1998; Kirchhoff et al., 2000; McClement & Degner,

1995; Shorter & Stayt, 2010), rather than formal support such as debriefing

(Halcomb et al., 2004; Jones & Fitzgerald, 1998). A national survey of critical care

nurses undertaken in the United States found that counselling and debriefing rarely

(38%) or never (49%) occurred (Puntillo et al., 2001). Identification and

implementation of strategies to support nurses are required to maintain quality care

25

for patients and their families and to minimise stress and eliminate burnout among

critical care nurses (Davidson et al., 2002).

2.6.2 Nurses’ end-of-life care values and beliefs

Nelson (2006) identified that personal attitudes and behaviours are the most

important barriers to end-of-life care and also the most difficult to change. Nurses’

values and beliefs could be influenced by the dominant medical discourse and

curative culture of critical care, as well as their personal knowledge and past

experiences (Nelson, 2006). The values and beliefs of critical care nurses influence

the facilitation of end-of-life care and nurses’ engagement in select practices. Two

alternate views of end-of-life care in critical care settings have been identified in the

current literature. Some critical care nurses share the perception of death as failure

and this perception could pose as a barrier to end-of-life care (Beckstrand et al.,

2006). Previous research has also identified that some critical care nurses view the

provision of care to patients at the end of life and their families as a privilege

(Calvin, Kite-Powell, & Hickey, 2007; Fridh et al., 2009; Vanderspank-Wright et al.,

2011). Strategies targeted at changing personal attitudes rather than the broader

organisational culture may be more effective in reducing barriers and improving the

provision of end-of-life care in critical care settings (Nelson, 2006).

2.6.3 Coordination of care delivery

Issues related to the coordination of care delivery are pertinent to the care of patients

at the end of life and their families in critical care. Previous research has identified

lack of communication in relation to prognosis, lack of information and conflicting

information from different health care providers as sources of dissatisfaction for the

families of patients dying in hospital (Rogers, Karlsen, & Addington-Hall, 2000).

Poor communication and lack of consensus have also been identified as barriers to

end-of-life care in research undertaken in critical care settings (Espinosa et al., 2010;

Vanderspank-Wright et al., 2011).

26

Fragmented care delivery occurs as a result of the critical care team, the patient’s

primary medical team, nurses and allied health care providers independently

reviewing the patient and outlining a treatment plan. The consistency of information

provided to the family can also be threatened by the number of care providers

involved in the care of a critically ill patient (Nelson & Meier, 1999). Lack of

consensus around treatment goals and the provision of different information from

different care providers can create distress for the family (Nelson, 2006) and the

critical care nurse (Vanderspank-Wright et al., 2011). Previous research suggests that

the medical team withdraw their involvement with the patient following a decision to

withdraw life-sustaining treatment (Halcomb et al., 2004; Heland, 2006), indicating

that the provision of care to the patient at the end of life is delivered by critical care

nurses with limited input from the medical team.

In addition to the coordination of care by bedside critical care nurses and the health

care team, the inclusion of palliative care experts and consults within critical care

units can assist in facilitating the provision of optimal care for dying patients. In one

study that compared the care of a retrospective cohort of patients who died in an ICU

with a cohort of patients following the inclusion of palliative care consults and

palliative care strategies when treatment goals change to comfort only, the length of

time before the implementation of comfort care following identification of poor

prognosis, the length of hospital stay and the cost of care was reduced (Campbell &

Guzman, 2003). Although this study only considered patients with global cerebral

ischemia and multi-organ system failure, it provides some support for the active

implementation of an end-of-life care strategy to improve the quality of care for

patients at the end of life in critical care. Other papers have described the use of

training of intensive care nurses as end-of-life care opinion leaders to act as role

models and promote changes in end-of-life care attitudes (Treece et al., 2006) and the

inclusion of palliative care specialists on intensive care ward rounds to provide

consultation and initiate end-of-life care conversations (Billings et al., 2006). No

evaluation data of these interventions is provided in these papers (Billings et al.,

2006; Treece et al., 2006).

27

2.7 Summary

The domains of end-of-life care practice identified by Clarke et al. (2003) provides a

framework for identifying and describing the practices of critical care nurses in the

provision of care to patients at the end of life and their families. The key elements of

the PARIHS framework—context, evidence and facilitation—provide an overarching

framework to examine the factors that influence the provision of end-of-life care in

critical care settings. Identification of current practice and the factors influencing the

provision of end-of-life care provides a platform for practice development to improve

the care that patients and their families receive and support nurses in the provision of

this care. The current study aims to (1) identify the end-of-life care practices of

Australian critical care nurses, (2) identify the factors influencing the provision of

end-of-life care, and (3) identify the factors associated with specific end-of-life care

practices of Australian critical care nurses. The next chapter provides a description of

the methodology for the study.

28

3 METHODOLOGY

3.1 Introduction

A mixed methods approach was used for the current study. This chapter will define

mixed methods research and discuss key strengths and criticisms of the mixed

methods approach. A three-phase exploratory sequential mixed methods design will

be introduced as the design selected for the current study. Ethical considerations

pertaining to this study will then be discussed.

3.2 Mixed methods

The mixed methods approach has gained prominence as the third methodological

approach, alongside the traditional qualitative and quantitative research approaches

(Johnson & Onwuegbuzie, 2004; Johnson, Onwuegbuzie, & Turner, 2007).

Proponents of the mixed methods approach do not consider it a replacement, but

rather an approach that complements the traditional qualitative and quantitative

approaches (Johnson & Onwuegbuzie, 2004). All three research approaches are

considered valuable in answering different research questions in different

circumstances and contexts (Johnson et al., 2007).

Following analysis of 19 definitions provided by leading scholars in mixed methods,

Johnson et al. (2007) broadly defined mixed methods research as:

the type of research in which a researcher or team of researchers combined

elements of qualitative and quantitative research approaches (eg. Use of

qualitative and quantitative viewpoints, data collection, analysis, inference

techniques) for the broad purposes of breadth and depth of understanding and

corroboration. (p. 123)

As identified in Johnson et al.’s (2007) definition, a mixed methods study includes

both qualitative and quantitative research strands that are combined or mixed in some

29

way. The rationale for using a mixed methods approach is that the strengths of both

qualitative and quantitative research approaches complement each other and, when

combined in a single research study, a comprehensive investigation of an area of

interest can be undertaken (Creswell & Plano Clark, 2011). Specifically, in a mixed

methods approach, both exploratory and interpretative research can be undertaken in

a single study. This is advantageous in multidimensional complex areas of research

interest, such as those that arise in nursing practice (Tashakkori & Teddlie, 2010).

Traditionally, qualitative research approaches acknowledge that multiple realities

exist and explore phenomena of interest in depth using small non-probability

samples, limiting the ability to generalise findings. In contrast, quantitative research

approaches employ larger samples to test hypotheses and causal relationships,

allowing generalisability of the findings. The mixed methods approach values both

subjective and objective knowing. The transferability of knowledge generated is not

limited by the methods used, but by consideration and appropriate use of the findings

in relevant situations (Morgan, 2007). In addition, consideration of areas of

convergence and divergence between the qualitative and quantitative findings can

provide further insight into the area of research under investigation (Teddlie &

Tashakkori, 2010).

While the quantitative research approach dominated much of the 20th century,

during the so-called paradigm wars of the 1980s and 1990s, there was renewed

attention to the place and value of the qualitative research approach to social science.

Key proponents of the value of the qualitative research approach aligned their

arguments with the philosophy of knowledge to provide credence and strength (see

Morgan, 2007). Criticism of the mixed methods approach stems largely from the

view that the philosophical assumptions underpinning the qualitative and quantitative

research approaches are incompatible and, thus, it is inappropriate to mix qualitative

and quantitative research methods.

Morgan (2007) states that the incompatibility of methods thesis ‘fails at every level

except for debates about the nature of reality and truth’ (p. 64). The mixed methods

scholarly community has rejected this incompatibility of methods thesis, instead

arguing for methodological eclecticism (Teddlie & Tashakkori, 2010). Teddlie and

30

Tashakkori (2010) define methodological eclecticism as ‘selecting and

synergistically integrating the most appropriate techniques from a myriad of

qualitative, quantitative and mixed strategies to more thoroughly investigate a

phenomenon of interest’ (p. 8). The research purpose or research questions are of

central importance and the most appropriate methods are selected to achieve the

research purpose or answer the research questions (Johnson & Onwuegbuzie, 2004),

as opposed to a mono-methodological approach and adherence to methods

traditionally associated with these approaches.

The mixed methods research approach requires the researcher to have knowledge and

skills in both qualitative and quantitative research methods. Although this can

present a challenge, those that engage in this work can become ‘methodological

connoisseurs’ and have a broader range of knowledge and skills in research methods

that can then be utilised to select and use the most appropriate methods in future

research projects (Tashakkori & Teddlie, 2010, p. 275; Teddlie & Tashakkori, 2010).

Several authors have developed mixed methods research design typologies to assist

researchers to select and combine the most appropriate methods to address their

research purpose and questions (see Creswell & Plano Clark, 2011; Johnson &

Onwuegbuzie, 2004). Mixed methods research designs vary in their timing, weighing

and mixing of qualitative and quantitative approaches (Creswell & Plano Clark,

2011). Specifically, a three-phase exploratory sequential mixed methods design was

utilised in this study (see Figure 3–1). In exploratory sequential designs, timing is

sequential with one methodological strand undertaken, followed by another, and the

strands are unequally weighed with one approach receiving a greater emphasis

(Creswell & Plano Clark, 2011). In the current study, a qualitative approach was

used first, followed by a quantitative approach, with a larger focus placed on the

quantitative approach.

31

Figure 3-1: Three-phase exploratory sequential mixed methods design (adapted

from Creswell & Plano Clark, 2007)

Phase 1 Phase 2 Phase 3

The mixing of the qualitative and quantitative research strands, termed the point of

interface, in exploratory sequential mixed methods designs occurs when the two

approaches are connected with the analysis of the findings from one strand informing

the development of the next strand (Creswell & Plano Clark, 2011). The point of

interface occurs in phase two of the current study, when the findings of phase one

(the qualitative strand) inform the development of a survey instrument for use in

phase three, the quantitative strand.

As previously identified, limited research has been undertaken to identify and

describe the practices of critical care nurses providing end-of-life care to patients and

their families and, therefore, a qualitative approach is appropriate for exploring this

phenomenon. Phase one of this study involved individual interviews with a small

convenience sample of critical care nurses from a single critical care setting to

explore their end-of-life care experiences and practices. The point of interface

between the qualitative and quantitative research strands in this study occurred in

phase two of this study, when the findings of phase one, combined with a review of

the literature, informed the development of the content of the survey instrument, for

use in phase three of this study (the quantitative strand). In addition to developing the

survey instrument, phase two of this study included a review of the survey

instrument by an expert panel and a pilot test of the survey instrument with a

convenience sample of critical care nurses. The findings of phase two provided

preliminary evidence of the reliability and validity of the survey instrument. The

survey instrument was then used in the larger quantitative phase of this study—phase

32

three—to conduct a national survey of critical care nurses to answer the three

research questions posed in this study. Analysis of participant responses to the survey

in phase three identified the frequency of practices undertaken in the provision of

end-of-life care, critical care nurses’ level of agreement with factors influencing the

provision of end-of-life care and the factors associated with specific end-of-life care

practice areas.

3.3 Ethical considerations

This study used humans as participants. Care was required in the planning and

conduct of the study to maximise the benefits of the study and minimise any

potential risks (National Health and Medical Research Council [NHMRC], 2007b).

Ethical considerations specific to this study include consent, anonymity and

confidentiality, the pre-existing relationship between the pilot test participants and

the associate supervisor, security of data and dissemination of findings. Potential

participants in this study were informed about the requirements of participation,

expected benefits and risks. Participants in phase one had an opportunity to ask

questions prior to signing a consent form to indicate that they were informed about

the study and had voluntarily agreed to participate (NHMRC, 2007b). In phases two

and three, submission of the completed survey implied consent.

Participants were assured anonymity and confidentiality. Consistent with NHMRC

guidelines (2007b), identifying information within the data was deleted from the data

set. The survey maintained the participants’ anonymity by not requesting their name

or contact details. Some demographic information was collected and, although

unlikely, this could result in a response to a survey in which a participant may be

individually identified (NHMRC, 2007b). If a response could potentially identify a

participant, this information was withheld from publication of the results of this

study. Pilot test participants in phase two of this study were recruited from a

postgraduate nursing subject at an Australian university. Although the researcher

does not have an existing relationship with the participants, the associate supervisor

of this PhD study is the course convener for postgraduate nursing students enrolled in

33

this subject. Participant confidentiality was maintained by restricting access to data

sets that identify participants and their responses to the researcher.

Research data is stored in a locked filing cabinet and will be retained for a period of

five years following submission of the results for publication (NHMRC, 2007a).

Electronic files are stored on a password-protected computer. Stored data is

periodically checked to ensure integrity and access to data is limited to the researcher

and supervisory team. The findings of this study have and will be disseminated to the

profession through conference presentations and publications in peer-reviewed

nursing and critical care journals. In addition, some participants voluntarily provided

contact details to the researcher and requested copies of findings as they are

published. Personal emails will be sent to these participants as requested.

Ethical approval to undertake this study was obtained from the Queensland

University of Technology (QUT) Human Research Ethics Committee (see

Appendices 13.1 and 13.2). In addition, ethical approval from the ACT Health

Human Research Ethics Committee was obtained for phase one of this study (see

Appendix 13.3) and the ACCCN agreed to support phase three of this study. A copy

of an email from the ACCCN secretary indicating support for this study is attached

(see Appendix 13.4).

3.4 Summary

A mixed methods approach using a three-phase exploratory sequential design was

utilised in this study. This chapter has provided a broad overview of this

methodological approach and the ethical considerations relevant to the current study.

Subsequent chapters address the methods used in each phase of this mixed methods

study.

34

4 PHASE ONE METHODS

4.1 Introduction

The first phase of this mixed methods study employed a qualitative research

approach. The research question guiding this phase of the current study was: What is

the experience of end-of-life care as described by critical care nurses? In answering

this question, the experiences and practices of critical care nurses in the provision of

end-of-life care were identified. As noted in the previous chapter, the purpose of this

phase was to explore dimensions of critical care nurses’ end-of-life care experiences

and practices, to assist with development of content for a survey instrument to be

used in later stages of the study. This chapter will outline the methods used in phase

one of the current study including the study setting, participants, method of data

collection, procedure and data management and analysis.

4.2 The setting

A 500-bed tertiary level referral centre and teaching hospital, providing health care

services to a population of approximately 500,000 in one Australian city and

surrounding region, provided the setting for phase one of this study. This hospital has

a 22-bed general ICU, providing intensive care services to medical and surgical

patients (Drennan et al., 2010). The unit includes two cardiothoracic surgery beds

and provides intensive care services to paediatric patients. Patients requiring high

dependency services are cared for in the unit, while coronary care services are

provided in a separate unit of the hospital.

The Joint Faculty of Intensive Care Medicine (JFICM, 2003) provides definitions of

standards required by ICUs for accreditation. Three levels of intensive care services

are identified by these standards with level three facilities providing the highest level

of critical care ‘including complex multisystem life support for an indefinite period’

(JFICM, 2003, p. 1). Patient care in a level three ICU is managed by an intensive

care specialist and level three units contribute to education and research (JFICM,

35

2003). The ICU that provided the setting for phase one of the current study was

classified as a tertiary level three intensive care facility.

4.3 Participants

In January 2009, the ICU that provided the setting for phase one of this study

employed approximately 60 full-time equivalent (FTE) registered nurses (J. Rochow,

personal communication, November 20, 2009). The unit employs both generalist

registered nurses, including new graduate nurses, and registered nurses who have

completed postgraduate education in intensive care or critical care nursing. There

were no enrolled nurses employed in a clinical position within the unit. This phase of

the study aimed to include experiences of both junior and senior nurses. During

recruitment of potential participants, it was acknowledged that the researcher was

interested in ascertaining the experiences and practices of both junior and senior

nurses. A purposive sampling technique was used to select a small sample of critical

care nurses employed in the ICU in the study setting to participate in an individual

semi-structured interview to discuss their experiences of providing end-of-life care in

the ICU.

4.4 Method of data collection

A semi-structured interview format was chosen to provide junior and senior nursing

staff privacy to openly share their personal experiences of providing end-of-life care

to patients in the ICU. Semi-structured interviews allow interviewees the opportunity

to freely share their experiences, providing rich in-depth accounts (Fontana & Frey,

2005; Silverman, 2006). Based on the assumption that participants will provide an

unbiased and truthful account (Fontana & Frey, 2005) of their end-of-life care

experiences, semi-structured interviews allow participants to talk freely and deeply

about their personal experiences and practices. Death and dying is an emotive topic,

and this topic can be sensitive for participants sharing their experiences and

discussing their practices.

36

4.5 Procedure

Participants were recruited via a presentation at a nursing staff meeting. The

presentation provided information about the study and the requirements of

participation. At the end of the presentation, seven registered nurses placed their

name and contact details on an expression of interest form. Three of these registered

nurses participated in an individual interview. A bulletin board advertisement (see

Appendix 13.5) was posted in the staff tearoom, the staff communication book and

on the research noticeboard in the main corridor of the unit. After seeing these

advertisements, two registered nurses contacted the researcher to express interest in

participating in the research study and were subsequently interviewed. A study

information sheet was also distributed to potential participants. The information sheet

outlined the purpose of the study, the potential benefits of the study to the

contribution of nursing knowledge of end-of-life care in critical care units and the

requirements of participation (see Appendix 13.6).

Qualitative methods typically employ small samples of participants, for the purpose

of obtaining in-depth exploration of a phenomenon of interest. After five individual

interviews, a number of key themes had emerged related to end-of-life experiences

and practices of intensive care nurses with varying levels of experience and

qualifications. While no claim is made that data saturation occurred with this small

sample, the insight gained from the interviews was consistent with themes identified

from the literature review relating to these issues. Given this convergence, the data

generated from the five participants was considered sufficient for the purpose of

phase one of the current study.

Fontana and Frey (2005) summarise elements of interviewing that interviewers must

consider, including how to present themselves and establish rapport and trust with

the interviewees. In this study, each interview began with the researcher thanking the

participant for agreeing to be interviewed and identifying that the interview was

being undertaken as a component of a PhD research study. The study was briefly

described, including its aims and potential contribution to nursing practice and

patient care. The participant was then asked to briefly summarise his or her nursing

education and employment history. This provided the researcher with information

37

regarding the participant’s nursing background and acted as an icebreaker to

encourage the participant to feel comfortable and talk openly. The participant was

then asked to describe a recent or significant experience of caring for an adult patient

at the end of life following a decision to withdraw or withhold life-sustaining

treatment. The role of the researcher as the interviewer was to actively listen to the

participant’s narrative and clarify or explore participant responses as required. The

interviewer is considered an active participant in the interview, providing verbal cues

such as mm, oh, huh and non-verbal nodding, smiling and frowning to encourage the

participant to talk and expand on his or her experiences and stories (Silverman,

2006). When necessary, additional open-ended questions were asked of the

participant, such as: What do you think is important in caring for patients at the end

of life? What assists you to provide end-of-life care? What barriers or difficulties

have you experienced when providing end-of-life care?

Interviews were conducted in a room in the ICU that ensured privacy and was free

from interruptions. In one instance, an interview was conducted in the home of a

participant at the participant’s request. Approval from the director of nursing was

provided to allow participants to attend the interview during work time. Interviews

ranged in length from 13 to 72 minutes (mean 41 minutes). A digital voice recorder

was used to record each interview to enable a verbatim transcription of the discussion

to be produced and permit the researcher to return to the discussion at a later time.

4.6 Data management and analysis

Immediately following each interview, the digital voice recording was sent to a

professional transcription agency for conversion into a word document. The

transcription from each interview was read while concurrently listening to the digital

record of the interview to first verify the accuracy and reliability of the transcription

as a written record and, secondly, to permit the researcher to return to the discussion

and recapture the intent of the written text conveyed through the participant’s verbal

expression (Silverman, 2006). Emphasis or intonations as expressed by the

participant were added to the transcription if they were considered to provide

38

important contextual data to the written record. Each transcript was reread until the

researcher felt familiar with the text.

In their handbook of qualitative data analysis, Miles and Huberman (1994) identify

data analysis as consisting of three concurrent activities: data reduction, data display

and conclusion drawing/verification. In the current study, data reduction proceeded

using a manual inductive coding approach, as described by Miles and Huberman

(1994). Each transcript was reread and labels of one to two words were placed in the

margin of the transcript to identify ideas and/or concepts of interest. Margin labels

were reviewed and segments of text containing similar ideas and concepts were

copied and pasted together into a new Word document to create a data display (Miles

& Huberman, 1994). Each group was named using one to two words that capture the

essence of the content. These groups formed the preliminary categories.

The data display of content of each preliminary category was reviewed to ensure best

fit of content (Miles & Huberman, 1994). Groups containing similar content were

combined to create major categories and, when appropriate, subcategories were

identified within the major categories. The content of each category was then

reviewed again and key quotes that exemplify the category were identified.

Check-coding, as described by Miles and Huberman (1994), during the initial coding

of the transcripts was undertaken. Three of the interview transcripts were

independently read and coded by the researcher and supervisory team. Inter-coder

agreement was achieved between the researcher and study supervisors following

comparison, discussion and debate of the codes. During data analysis, the key

research questions and aims of phase one of the current study were reviewed to assist

in conclusion drawing. The researcher interpreted the findings by reflecting upon the

data and categories identified in relation to the study aims and research questions.

Claims about the data were made and supported by exemplars and quotations. In

addition, a data display consisting of a matrix comparing the key domains of end-of-

life care practice and the factors influencing the provision of end-of-life care

identified in the literature review with the findings from the phase one interviews

was created. The findings of phase one were presented at a nursing in-service in the

study setting to enable further verification of the study findings. Nurses present

39

during the in-service agreed that the findings generally reflected their own

experiences and practices of end-of-life care, providing further confirmation of the

categories developed during data analysis.

4.7 Summary

The purpose of phase one of this study was to obtain insight into the end-of-life care

experiences and practices of critical care nurses. Phase one utilised a qualitative

approach to interview a small convenience sample of critical care nurses from a

single ICU to achieve this purpose. The findings of the analysis of interview data are

presented in the next chapter.

40

5 PHASE ONE FINDINGS

5.1 Introduction

This chapter will briefly describe the characteristics of the registered nurses who

participated in phase one of the current study, completing an individual interview

with the researcher. The categories and subcategories identified from the interview

data will then be described. Excerpts from the interview transcripts are included as

exemplars to support the categories and subcategories identified.

5.2 Participant characteristics

Five female registered nurses participated in phase one of the current study. The

participants included both senior and junior members of nursing staff. Table 5–1

outlines the number of years that each participant had been a registered nurse and the

number of years that they had worked in an ICU. Three participants had completed

formal postgraduate qualifications (either graduate certificate or graduate diploma) in

intensive or critical care nursing.

Table 5-1: Phase one participant characteristics

Participant Years as a

nurse

Years as an ICU

nurse

Postgraduate

qualifications

1 5 3 No

2 25 15 Yes

3 6 3 Yes

4 29 22 Yes

5 3 2 No

Note. ICU = intensive care unit.

41

Participants in this phase of the current study were asked to discuss their experiences

of caring for a patient after a decision to withdraw or withhold treatment had been

made. It is acknowledged that practice is complex and multidimensional in nature;

however, due to constraints of the spoken and written word, this account of care of a

patient at the end of life, at times, appears linear. In addition, excerpts used to

provide evidence of a category or subcategory may also provide evidence of other

categories or subcategories. This is reflective of the interrelationships that exist

between the categories and subcategories developed from the interview data. Three

major categories were identified: Beliefs about end-of-life care, end-of-life care in

the intensive care context and facilitating end-of-life care (see Table 5–2).

Table 5-2: Categories and subcategories developed from phase one interview

data

Category Subcategory

Beliefs about end-of-life care Value of participating in end-of-life care

Complexity of end-of-life care

End-of-life care in the intensive care

context

Organisational support

Modifying the bedside environment

Facilitating end-of-life care Uncertainty and ambiguity

Emotional intensity

Providing comfort care

Humanising the patient

Facilitating the family’s experience

5.3 Category: Beliefs about end-of-life care

During the interviews, participants shared their personal beliefs about death and

dying and the provision of end-of-life care in the intensive care setting. In particular,

participants in phase one of this study expressed their beliefs about the value of

participating in end-of-life care and the complexity of end-of-life care. Variation in

42

critical care nurses’ personal beliefs about end-of-life care may influence the care

that is provided to patients and their families.

5.3.1 Value of participating in end-of-life care

The value of participating in end-of-life care was emphasised by participants.

Participants used terms such as privilege and honour to describe how they felt about

the opportunity to care for patients at the end of life and their families.

I think, depending on how you look at it, I guess it’s a little bit of a—it’s kind

of an honour to share that most intimate moment in someone’s life.

(Participant 5)

I’m very aware of the intimate position that we have in people’s lives in these

big, big moments. (Participant 2)

Some participants felt that the value of participating in end-of-life care was not

always recognised by their nursing colleagues. The intensive care setting is

dominated by a curative culture driven by a medical model of practice. The aim of

intensive care is to save lives and loss of life may be viewed as a failure. The

curative culture in addition to the nurse’s beliefs and comfort with spirituality and

death can pose as barriers to engaging in end-of-life care, influencing the care

provided to the patient and the patient’s family.

I think there’s a very strong part of the culture here in ICU that death is

failure. Nurses don’t seem to like looking after dying people, not only

perhaps because they don’t quite know how they feel about death themselves,

but I think also because it’s considered that you don’t need skills to look after

a dying person. I think that it’s quite the contrary. (Participant 2)

43

5.3.2 Complexity of end-of-life care

Participants in phase one of the current study believed that the provision of end-of-

life care is an advanced practice skill. However, participants did not elaborate in

detail about what was involved in the complexity of caring for a patient at the end of

life. Aspects of end-of-life care practice that were identified included interpersonal

skills, assessment skills and knowledge of symptom management.

I’ve seen it allocated to the junior nurses and it’s just, oh well, you just

extubate and then start the morphine or, and it is so much more than that, so

much more. (Participant 1)

Generally, participants in this study considered end-of-life care an advanced practice

skill and that the allocation of nurses inexperienced in the provision of end-of-life

care posed a barrier to the care provided. However, confidence and skill in the

provision of end-of-life care may be developed through experience in providing this

care.

I think that looking after a dying patient is an advanced practice skill. Even

just last week, we had some—quite a few deaths and some junior people who,

clearly, I could see, were struggling with—one, I mean, I guess with the care.

How to deal with the family and how to, I guess, even deal with it herself.

(Participant 4)

I think it does take time and skill and years to look after people who are dying

and to be comfortable. (Participant 4)

Some participants felt that the complexity of end-of-life care was not always

recognised by their nursing colleagues. Examples of inappropriate allocation of

junior nursing staff and higher nurse-to-patient ratios were cited as evidence of both

the lack of value and lack of understanding of the complexity involved in end-of-life

care.

44

I think what frequently happens is that the more junior nurses are allocated

the patients that we’re withdrawing on, which I think is completely, I guess

unfair to the nurse. But also, I guess it also impacts the family. (Participant 1)

In recent times I was given two patients who weren’t expected to survive.

They were both on ventilators . . . I had two students as well. I had two

patients on ventilators, which is normally unheard of in this unit. I also had a

student to mentor with each of these patients. I thought that was pretty overt

disregard for the complexity that is involved in end-of-life care . . . So that

was really evident to me that withdrawing and dying was seen as something

very simple. It’s not simple. It’s time intensive. There’s observation . . . I’m

collecting information all the time. I’m being vigilant all the time.

(Participant 2)

Variation in the personal beliefs held by individual nurses about the value of end-of-

life care and the complexity of this area of practice are factors that may influence

intensive care nurses’ approach to end-of-life care and the care provided to patients

and their families.

5.4 Category: End-of-life care in the intensive care context

The ICU is a unique practice setting and the factors influencing care provided in this

context may differ to those in other contexts such as an acute ward or oncology unit.

Participants reported a practice context of limited organisational supports relating to

end-of-life care issues. Interventions aimed at modifying the bedside environment

can assist nurses in creating a context within the intensive care setting that supports

the provision of care for patients at the end of life and their families.

45

5.4.1 Organisational support

Some participants in this study perceived a lack of guidance from their colleagues

and the organisation in regard to end-of-life care. Following the decision to withdraw

treatment, the medical team was perceived to withdraw from participating in the care

of the patient. Factors that may influence the disengagement of medical staff include

the curative culture, perception of death as failure and the personal comfort of

medical staff with death and dying.

Speaking to family in those circumstances and how to—basically, when a

family says we’re withdrawing the tubes out, the doctors don’t even go in

except to pronounce the death. There would be nothing wrong with them

popping in just to see how the family was going or—you know. I think to

show the ongoing care of doctors, as well as nurses. Yeah. I think that might

be fear as well, might be time constraints. Yeah. (Participant 4)

We don’t tend to have doctor input after the decision to withdraw care,

though, and maybe that’s something that we could look at as well, to do

better. Generally, after the family meeting, the decision to withdraw, the

family only deals with the nursing staff who, if they’re junior, can’t explain

what might be happening and why certain symptoms might be occurring or

don’t know that certain symptoms can be treated and your antimucolytics for

rattly chest and stuff is obviously something that I always advocate for.

(Participant 1)

The variation in perceptions of the value and complexity of end-of-life care was also

evident in the guidance provided by senior nursing staff. Some participants felt that

the guidance provided by nursing team leaders to junior nurses was superficial and

inadequate to address the complexity of care required by a patient at the end of life.

I’ve seen it allocated to the junior nurses and it’s just, oh well, you just

extubate and then start the morphine or, and it is so much more than that, so

much more. (Participant 1)

46

Nursing care should be individualised to the needs and personal preferences of the

patient, and care at the end of life is no different. Although some participants in this

study stated that care should be individualised, they also felt that policies and

guidelines about end-of-life care may assist staff in the provision of this care by

helping them to think about various aspects of care that may be required by the

patients and their families. Examples provided by participants included contacting

spiritual advisors and social workers and asking about the patient’s pets. No policies

or guidelines were available in the unit to guide nurses providing end-of-life care.

5.4.2 Modifying the bedside environment

As previously discussed, end-of-life care was highly valued by participants in this

study, and they wanted to create a positive experience for the patient and family. All

participants described interventions undertaken after a decision had been made to

withdraw or withhold treatment to modify the bedside environment and create a good

environment for the patient and the patient’s family to experience end-of-life care.

Participants attempted to create an intimate and peaceful setting that supported the

provision of end-of-life care by ‘removing the intensive out of intensive care’

(Participant 2) through interventions to provide privacy, remove clinical equipment

and create a homely environment.

In the transition to end-of-life care, participants indicated that patients were

transferred to a single room within the ICU if they were not already in one. Patient

care in the ICU that provided the study setting for phase one of this research study is

provided in open plan bays, with four beds in each bay, or in one of four single

rooms. Participants reported that admission to a single room provided space and

privacy for the patient and his or her family, allowed more visitors at the bedside and

permitted visitation of family outside regular visiting hours. The use of single and

family rooms was also perceived to be beneficial to other patients and families, to

lessen their exposure to the dying patient and the patient’s family.

47

Normally people go to a private room: one, to allow more visitors; secondly,

it’s just a quieter environment. Also, for other patients as well, to see people

upset and crying is very disturbing for the other relatives because they think

there but for the grace of god go I. They’re given a quiet room. (Participant

2)

One participant identified that single rooms are not always available and expressed

frustration at not being able to provide the family with adequate privacy.

I felt like I couldn't give them the privacy that they deserved because I was

standing behind those magical soundproof curtains. It’s not the same; it’s not

the same. (Participant 5)

Another factor that facilitated a positive experience of end-of-life care was the

availability of a private room for use by the family of the patient at the end of life.

Participants in this study identified that their ICU has a designated room, separate to

the intensive care visitor’s waiting room, which is used by families of dying patients.

This room allows family members to remain within the ICU, close to their relative,

and to congregate separately to the other visitors to intensive care.

I think the family room is fantastic. It’s a little bit of privacy that you’re not

in the visitors room crying or—you know. They’ve got a loo there, so you

don’t have to step out to come in to use the loo, that sort of thing. I think,

overall, they’re really good things . . . I think also our kitchen provides us

with sandwiches and drinks. So I think we can—I think we give the family a

good environment in which to experience death. (Participant 4)

Participants reported other interventions undertaken to modify the bedside

environment, such as shifting monitoring equipment out of view of the family and

removing clinical equipment that was not in use, including ventilators, infusion

pumps and intravenous lines. Participants felt that monitoring equipment was

48

distracting for the family as they spend their time watching the monitor rather than

looking at and talking to their relative.

Generally, I mean I tend to ask the family members what they want but we

can kind of dim the lights and have chairs all around and stuff but generally

I’ll turn the screen so that, if need be, I can see it from the other room. If

anyone in particular wants to be able to see it, then they can. I guess we do

minimise the equipment, the monitoring equipment, to sort of clear, clear out

the room as best we can to fit as many people in as we can and I guess make

it less clinical. (Participant 1)

I think that’s really important and getting rid of the monitoring is a big thing

for me because I don’t think that, I don’t want their last moments with their

loved one to be watching a screen like you would on an emergency show

where it goes blip, blip, beep. They sort of expect it and they’ll just watch the

monitor and I think that’s not fair. You shouldn’t be watching a monitor; you

should be watching your loved one and talking to your loved one and I think

we’re depriving them of that if we keep all this other stuff around.

(Participant 5)

Removing clinical equipment combined with modifications such as adjusting the

lighting, changing the bed linen, putting up photographs and playing music assisted

in creating an atmosphere that was less clinical and more homely and comfortable for

the patient and family. The staff on the unit had developed a resource kit containing

equipment such as linen, lamps and cushions for use when caring for a patient at the

end of life. The kit contained coloured linen and matching quilts that was used to

replace hospital linen. Participants stated that they turned off or dimmed the lights in

the room and used bedside lamps to create a peaceful atmosphere.

Get rid of the sterile environment, change that perception, it looks like

they’re relaxed, more like they would be at home. (Participant 5)

49

In that room we’ve got lamps to create an ambience of quietness and

intimacy, as much as you can in the clinical confines of an ICU unit. We use

linen. We try to make it as un-clinical as possible. (Participant 2)

Put up pictures, put up music, get their own blankets, get whatever, and make

it a personal experience, rather than the stark white and tippy-toe around.

(Participant 4)

End-of-life care was highly valued by participants in this study and they wanted to

create a positive experience for the patient and family. The practice context of

intensive care is a busy and noisy environment; however, an intimate, peaceful

setting in which to experience end-of-life care and death could be achieved by

transferring the patient to a private room and modifying the environment to create a

homely feel. The removal of monitoring and equipment encouraged the family to

direct their attention to the patient and change the focus of care to providing comfort

at the end of life.

5.5 Category: Facilitating end-of-life care

The category of facilitating end-of-life care incorporated five subcategories. The

subcategories uncertainty and ambiguity and emotional intensity captured nurses’

perceptions of the factors influencing the provision of end-of-life care and the

support available to nurses. The subcategories comfort care, humanising the patient

and facilitating the family’s experience captured the practices of critical care nurses

in the provision of care to patients at the end of life and their families.

5.5.1 Uncertainty and ambiguity

Some participants alluded to the uncertainty and ambiguity surrounding patient

prognosis and end-of-life decision making in this setting. This uncertainty and

50

ambiguity appeared to continue even after decisions had been made to withdraw

treatment. In a practical sense, participants reported that this would manifest in

waiting for family members to arrive at the patient’s bedside, resulting in a delay

between when a decision was made to withdraw treatment and when this actually

occurred.

We can have a dying patient here for two days because—well, first of all,

well the niece needs to come and she won’t be here for 12 hours . . . then the

second cousin, four times removed . . . I think a lot of times we give the

family a lot of—maybe too much time to—before withdrawal of treatment . .

. I think it’s disrespectful to the person who’s dying, to lie there, gasping

sometimes, because they’re waiting for someone to come. (Participant 4)

When a clinical decision has been made in relation to withdrawing and withholding

treatment, the patient’s family may not always be ready for this decision to occur.

The family may require time to accept the change in treatment plan to comfort only

and the pending loss of their loved one. Several factors may contribute to family

readiness, including the uncertainty and ambiguity surrounding the patient’s

prognosis and treatment in intensive care, the overwhelming experience of having a

family member in intensive care and the experience of grief and loss. Nurses’ role in

facilitating end-of-life care may include ascertaining family readiness and providing

adequate preparation and support for family members of patients at the end of life.

5.5.2 Emotional intensity

In the critical care setting, the difficulty in determining patient prognosis and the

availability of advanced treatment modalities that can maintain life can make

decisions regarding the continuation or termination of treatment difficult.

Participants were informed that the focus of this study was the experience of caring

for a patient at the end of life after a decision had been made to withdraw or withhold

life-sustaining treatment. Therefore, the data does not tend to capture the uncertainty

51

and ambiguity that surrounds this area of practice. However, participants alluded to

some of the challenges that they are faced with when they spoke of avoiding sending

mixed messages to the families of patients at the end of life and their participation in

family meetings to discuss patient prognosis and decisions to shift focus of treatment

to comfort only.

Intensive care is a high-intensity work environment. Caring for patients with

traumatic injuries and critical illness on a daily basis exposes nurses to grief and

human suffering not experienced in other workplaces. One participant highlighted

this situation, stating:

As a junior person, in particular, you don’t have anyone to talk to because

your nursing friends that you would normally confide in don’t understand.

Your family doesn’t understand, your partner doesn’t understand, they don’t

get it. You say, I’ve had a really horrible day. Oh, mine too. Column A

wouldn’t add up to column B and that made column C all out of whack. No,

someone’s died. No-one gets it except for the people that are here.

(Participant 5)

With everyday exposure to critically ill patients, nurses may be expected to be

resilient to the experience and emotional aspects of this work. One senior member of

nursing staff described how senior staff check in with each other after caring for a

patient at the end of life. This exemplar suggests that, although this support is

offered, it is expected that support would not be required.

Yep. But a lot of times it’s just, are you alright? They go, yep. Because a lot

of times we are, because we’ve done it so many times. Sometimes it isn’t.

(Participant 4)

In addition, the intensity of nursing work in intensive care may become second

nature to senior staff and they may not recall how overwhelming the experience may

be for junior staff. Some participants identified that as a nurse in intensive care, you

52

were expected to cope with death and dying. These participants felt that the

prevailing culture in intensive care made it difficult, particularly for junior staff, to

ask for help or seek emotional support during or after the provision of end-of-life

care to a patient and family.

It’s not okay here to say that you’re not quite coping. You are looked down

upon if you say I’m not really coping, I’m kind of struggling with this. I had

a really emotional day. This is hard. We don’t have that support. (Participant

5)

One participant also expressed concern that staff do not have a say in whom they are

allocated to care for and this may lead to situations that are emotionally confronting

for the nurse. This participant also discussed the lack of down time between the

provision of end-of-life care and the admission of another patient during the same

shift. These situations may occur because end-of-life care is not considered important

and/or the complexity of end-of-life care is not recognised by some. These situations

may not only influence the care provided, as identified by Participant 1, but may also

affect nurse coping and stress levels.

So if they’re not comfortable regarding a specific issue of dying or it may be

what the patient may be dying of, maybe we should reconsider allocation.

That’s a thing at the moment, we’re pre-allocated, we don’t have a say in who

we nurse and I think that may impact on care. (Participant 1)

I see a need for debrief for nurses, particularly after, and one of my examples

is coming on shift, having a really critically unwell patient and they die as

soon as you’ve sort of sorted out things with family, attended to paperwork

and stuff, then you can get the next post-op off the ranks sort of thing. I don’t

think we provide adequately for that changeover for nurses. No-one sort of

comes up to you and goes, gee, was there anything in that that struck a chord,

triggered something, you know, whether it was, if I had a previous

experience whether in nursing or in my personal life. You don’t really get

asked those questions of, how are you going after that this morning. I can

53

certainly see that there’s a need for debrief or even just a down time when a

person can change gears between patient A and patient B. (Participant 1)

The frequency of critical incidents in intensive care may mean that debriefing

sessions are overlooked as nurses continue to engage in the busyness of the work

environment. Participants in this study identified that more support for intensive care

nurses providing end-of-life care was needed. It was acknowledged that external

support services existed but these were not accessed. Staff identified that they often

rely on their collegial network for support and that this form of support may not be

available for junior staff who may not have established these networks.

Being supported, I guess. If you knew—even if you don’t have to take them

up on it—if you knew that there was support there, it would be better.

(Participant 5)

Just having support, junior nurses being supported in the role of doing end-

of-life care instead of expected just to be able to cope. (Participant 1)

I think us oldies just use each other, the way we’re used to, because we never

had anything like that. I guess the hospital does make things available

through whatever it is [external employee support programme]. But people

again are reluctant to use services like that. (Participant 4)

Intensive care is a unique practice context involving high-intensity and emotive

work. Uncertainty and ambiguity related to patient prognosis, the trajectory of

critical illness and treatment decisions surround practice in intensive care. Nurses in

phase one of this study described a culture of poor organisational and emotional

support. Lack of emotional support and inadequate guidance in end-of-life care

practice may influence intensive care nurses’ end-of-life care experiences, stress and

coping and may compromise the quality of care afforded to patients at the end of life

and their families.

54

5.5.3 Comfort care

As previously discussed, participants viewed end-of-life care as an advanced practice

skill and felt that the complexity of this area of practice was not always recognised

by their colleagues. End-of-life care in the ICU consisted of more than just the

removal of treatment.

It’s not just withdrawing; it’s actually providing comfort. (Participant 2)

Just because they’re dying, doesn’t—we might be withdrawing treatment, but

we don’t withdraw care. (Participant 4)

One participant defined comfort care provided at the end of life by offering the

following explanation:

We had moved her to comfort care. If comfort is the criteria, then you’re

looking at emotional, physical, psychological comfort, not only of her but

also of the extended family. The significant relationships are really important

in terms of end-of-life care and creating comfort so that the person dies with

as much ease and release as possible. (Participant 2)

A focus on caring for the patient’s family and facilitating the family’s experience of

end-of-life care dominated participants’ narratives of their end-of-life care

experiences in intensive care. The actual nursing interventions undertaken in the

provision of comfort care to the patient were not directly discussed by participants in

detail. Often, participants revealed tasks undertaken in the care of the patient in the

context of highlighting some other aspect of their end-of-life care practice or

experience. For example, the task of bathing a patient was mentioned by one

participant in the context of allowing family to be involved in the care of the patient.

The actual tasks undertaken in the provision of comfort care may be taken-for-

granted aspects of participants’ practice. Additionally, participants may have felt that

55

they did not need to describe these details of their practice to the researcher due to

the researcher’s experience and knowledge of intensive care work.

Tasks undertaken in the provision of comfort care that were mentioned in the

interviews but not described in detail included bathing, hair care, mouth care,

pressure area care, spiritual care, and the administration of analgesics, sedatives and

antimucolytics.

5.5.4 Humanising the patient

Participants in this study described interventions to humanise the patient and assist in

shifting the focus of care to comfort only. Interventions to humanise the patient

included removing equipment, personal cares and talking about the patient.

When caring for patients at the end of life, participants spoke about humanising the

patients to make them look more like themselves. Removing monitoring and

equipment or covering up equipment that could not be removed and providing

personal cares such as combing the patient’s hair assisted in de-medicalising the

patient and shifting the family’s focus to the patient as a person and comfort-only

care.

Try and make them comfortable. I don’t know. I think, in some ways, it can

be very intuitive. I just think what I would like or what humanness, human

beings would like, and try and do that. (Participant 4)

Make them look a bit more like themselves, get rid of as many lines as you

can. If stuff has to be in, it has to be in. Cover that arm that’s covered in stuff.

(Participant 5)

Get rid of all the extraneous monitoring equipment, to leave it at the bedside

in the main ward. So we go into the single room with absolutely minimal

stuff. We don’t need . . . Often we just have morphine and midaz running.

56

That’s all we need: just one pump, just one ventilator. Sometimes we just put

them on a bag to enter the single room so that they’re just breathing on room

air so that . . . There’s a medical reason for this in my practice. That is to

allow CO2 to build up so that they’re almost CO2 narcotised and sedated

almost naturally so that they don’t experience too much air hunger if they’re

not actually brain dead. So there are things that I do. When I get into the

room I change the settings on the ventilator to be as normal as possible. If

they’re on high amounts of PEEP or high amounts of pressure support, I sort

of crank it down incrementally so that by the time we actually get to

extubation . . . I don’t want the relatives to experience distress by seeing air

hunger. (Participant 2)

Patients are admitted to ICUs following trauma or an acute phase of illness, often

presenting with an altered level of consciousness. This poses a barrier to establishing

a relationship between the patient and the intensive care nurses. Nurses also shifted

their focus to the patient as a person by asking the family about the patient to enable

them to ‘get to know’ the patient. The process of the nurse ‘getting to know’ the

patient also provided an opportunity for the family to talk about their relative and

reflect on their lives together.

Yeah. I’ve had two where people just—I think they—not all, but a lot of

them do want to tell you what the person who’s lying in the bed. You know,

we had one lady who—I was just sitting in the empty room. She was sitting

with her dad. Often she physically came and grabbed me and pulled me

beside her dad’s bed. She said let me tell you about my dad. She said, oh, he

had a fantastic voice. I said, oh, he could sing. She just laughed and said, no,

he just had a nice voice. She just wanted to tell people about—and sometimes

I just ask too, depending on if you read the mood right. You know, what was

he like? Or blah-blah-blah. Something like that, just to try and make it that

they can speak about the person who’s dying or indeed has already died,

about what they were, about their life, that kind of stuff. (Participant 4)

57

I think you do find people wanting to share stories and that sort of stuff. So I

think allowing them to do that and to let us, the stranger, sort of know them.

Or I can just say I don’t know your dad, because he’s always just been

attached to tubes. So what was he like? There’s ways into those

conversations if it’s appropriate. (Participant 4)

Through interventions to humanise the patient, both families and nurses shift their

focus to the patient as a person and comfort-only care. These interventions create a

context conducive to the provision of end-of-life care in the intensive care setting,

and the value of participating in this moment in a person’s life for nurses is evident.

5.5.5 Facilitating the family’s experience

Care of the dying patient’s family was identified as a focus of care for participants in

phase one of this study, and a significant proportion of participant dialogue in each

interview was dedicated to this care. The findings of this study identified facilitating

the family’s experience during the provision of end-of-life care as an important

nursing role.

Participants in this study acknowledged the need to provide options and respect

choices of the dying patient’s family. Providing options and respecting choices were

primarily discussed in relation to the patient’s family. The limited or absent

discussion of the patient’s wishes by participants in the interviews may be in

response to the inability to interact or develop a relationship with patients in

intensive care due to their altered level of consciousness. In the absence of an

advance directive, the family may be able to indicate what they believe the patient

would want in the given circumstances. The intensive care nurse has the opportunity

to develop a relationship with the family and discuss possible options and choices.

Nurses are able to facilitate a family’s experience of end-of-life care by providing

options and respecting choices, enabling family involvement and participation.

58

Participants spoke of their ability to influence the family’s experience of end-of-life

care. Some participants stated that they asked the family ‘what do you want?’ to try

to ascertain the type of atmosphere the family wanted to create, the support required

from the nurse and the degree to which family wanted to be involved in providing

end-of-life care to their relative.

So I think it is personal, I think we should ask people what do you want?

This is your hell. You’re going through it. How do you want to go through it?

(Participant 4)

So I helped the family, before the withdrawal, helped them. One of them

wanted to bathe, help bathe, so we helped bathe and do all the general care of

their loved one. (Participant 3)

One participant described that a rapport with the family was needed to assist the

nurse to facilitate family experience.

It’s really hard looking after someone if you don’t have a rapport with the

family. I generally can create a rapport pretty well so I haven’t had too many

experiences where I haven’t but if the family’s really closed and you just feel

really awkward, even going in there to do anything at all. If the infusion

pump is beeping or something you feel really uncomfortable going in there

and I think that would then reflect to them. So I just find it such—it makes it

emotionally harder if you don’t have a rapport. In a way I know that sounds

strange but I sort of think, what are they getting from me, like what sort of a

vibe are they getting from me if I’m feeling uncomfortable in a situation.

Like that’s not fair on them either. (Participant 5)

All participants acknowledged their role in explaining and providing information to

the patients and their families. Participants in this study identified that they explained

what was happening to the patients and their families, answered questions posed by

family members and provided information.

59

You don’t have to really say much to them when they’re dying, when they’ve

got their loved one dying. Explaining when things happened. Just, I’d

switched the monitoring around so the family weren’t fixated on the

monitoring but I did say when their heart had stopped and things like that. I

explained, they said there was some beeping noises and I explained that was

normal and it could continue for a little while. So it was more just

reassurance and information that I gave the family at that time, without

information overloading them as well. (Participant 3)

It’s very satisfying to be able to give people information . . . Even though

they’re dying, they’re still not dead. (Participant 2)

The provision of end-of-life care can be an emotionally confronting area of practice.

It is a second-nature aspect of nursing practice to explain and provide information,

and this is no different when caring for a patient at the end of life. Although the

actual content of information and explanations provided could not be ascertained

from the interview data, explaining and providing information was considered a safe

area as the information needs and questions from family were similar. This was a

practical and comfortable area of practice for both senior and junior staff. In response

to her role in explaining and providing information about clinical equipment and

treatment interventions, one participant said:

I felt quite comfortable doing that, that was something I could do. Right,

okay, comfort zone again, I can explain this stuff, that’s cool. (Participant 5)

Facilitating end-of-life care included explaining and providing information to the

family. The findings suggest that explaining and providing information was a

comfortable area of practice for all participants in the often confronting and

emotional work of providing end-of-life care.

60

All participants described being there for the family through their presence with the

patient and family according to the family’s preference and the nurse’s comfort with

being there. The environment of the intensive care setting and the experience of

having a relative at the end of life may be overwhelming for family members.

Nursing presence or being there was integral to facilitating family experience as they

support family members to be with their relative at the end of life. This support

included a physical presence by the nurse at the patient’s bedside and encouraging

the family to sit, talk to and touch the patient.

Yes, the family knew what was happening anyway and when they came back

in, just stood with the wife for a while and, I don’t know, I think she wanted

me to be there for a little bit. Because we did have such a good rapport I

think she just sort of held on to my arm, you know, it was like, it’s okay, sit

down and encouraged her to talk to him and all that and of course she asked,

can he hear me and you have to say, well, I don’t know but personally I

believe, I’d like to think so. So she sat down and held his hand and had a chat

and then I left, I stepped out and all the family sort of had their little time in

there and basically they were all in there, yes, until he passed away.

(Participant 5)

I think that’s another big thing, is permitting people to touch and talk and cry.

I think those are the three biggest things. I think with all the lines and tubes,

patients’ families generally hold back, concerned that they might knock

something out or do the wrong thing and a cranky nurse is going to jump all

over them. But I think if we explain, well, you can touch them here or touch

them there if there is risk of something becoming dislodged. (Participant 1)

Being there was not only a physical presence, but also involved emotional

investment of the nurse.

I mean we still cry. I mean there was a team leader crying with those deaths

last week. But that’s alright. I don’t think crying is bad. I think crying is

good. Yeah. So I think if something’s sad, there’s nothing wrong to cry, even

61

with the family seeing her . . . I must admit, I haven’t done that for a while.

But you know, even—you know, in this unit, there’s one family where I

think I took the tissue box out of the mother’s hand because I was the one

bawling. They just laughed. But they enjoyed that I still—that you still could

feel, that you could still care and that sort of stuff. We had a—I had a real

special bond with family. So it’s alright to cry. I don’t apologise for it. I’m

not embarrassed by it either. In front of my colleagues, I’m not either. If it’s

sad and I’ve got tears in my eyes, well, that’s how it is. Someone comes and

gives me a hug and off we go. Yeah. (Participant 4)

Without this emotional investment, the nurse may risk focusing on the process and

tasks of end-of-life care at the expense of caring for the family. Inexperience and

discomfort with providing end-of-life care are factors that contribute to nurses

avoiding being physically and emotionally present for the patient and family.

Other just little subtle things. She has a body bag and the thing’s in the room,

just as the family was leaving. So yeah. Maybe that was just a little bit

premature . . . I don’t think the family knew what it was. But yeah, but just—

she was just trying to think of being efficient. She was thinking of doing her

job, rather than, I think, thinking with her heart a little bit. But I think she

was too afraid to think with her heart. Not that—I don’t think that she

intentionally would have done anything like that. But I think it was just her

inexperience. (Participant 4)

I guess she was afraid—it seemed like she was afraid to speak to the family

or to be in the room. Or she’d go in the room and be tiptoe in, tiptoe out, sort

of thing, instead of—I don’t know. She obviously didn’t have much death

experience. (Participant 4)

I think nurses as well can be very task orientated and sort of can hide behind

being busy doing tasks. (Participant 1)

62

Physical and emotional presence by the intensive care nurse is important in

facilitating end-of-life care by supporting family members. For a nurse to be there at

the end of life for patients and their families, they need to feel comfortable with

being there themselves.

As previously outlined, participants in this study valued participating in end-of-life

care and identified their role in influencing the end-of-life experience for patients and

their families. Creating memories and positive experiences were practices undertaken

by intensive care nurses in facilitating end-of-life care. All nurses can contribute to

the creation of memories and positive experiences that families may carry with them

through their lives.

This is the most horrific, horrible thing that they’re ever going to face, this

family. They’re losing someone that’s most important to them in their life but

it doesn’t have to be so bad. I think there’s little things you can do that

change someone’s perception of the whole thing even though it sucks.

(Participant 5)

I think often families—I think we have a lot of—not the word controller part,

but we have influence, I think, on how people die and the memories that

those families take away through the death process. (Participant 4)

All participants in this study described actions that they undertook to assist in

creating memories for the patients’ families. Participants described examples of

creating memories such as taking the patients’ hand prints and collecting a lock of

the patients’ hair and their identity bands for the families to keep. These activities

can be offered to all families, regardless of the nurse’s level of skill and expertise in

end-of-life care.

Some participants in this study described experiences in which the role of the nurse

was pivotal in supporting families to have meaningful experiences with their relative

at the end of life. These stories indicated an advanced level of practice whereby the

nurse involved could intuitively grasp the family situation, identify a meaningful

63

moment and respond appropriately to assist in creating a positive experience. The

importance of these experiences for the family was highlighted when the family

acknowledged the nurse for their work.

I wrote a book for this little girl, for both of them. After hearing the

discussion . . . I drew upon what I had heard the father talking to the little girl

about . . . I didn’t get the husband’s permission and I didn’t tell anybody that

I was doing this. I don’t know where I stood ethically. I just was really driven

by concern for this little four year old, and also just the lack of vocabulary

that this man had for explaining death. The book was created . . . Then I went

to the funeral and he actually read that book out in lieu of a eulogy. So he

was still so without vocabulary or anything that this was just something. I

was struck by the significance of what the book became. Then he actually

wrote a letter to the CEO about the care that his wife had got. I got a phone

call from the CEO. That really blew me away. (Participant 2)

Most recently there was a 27 year old that was dying of leukaemia. She was

very strong willed. She knew what she wanted and what she didn’t want. By

hook or by crook, she wanted a cigarette before she died because that’s what

she and her dad used to do; they used to smoke together, even with the

leukaemia. I thought it was really important to give dad and the daughter the

opportunity to hang out on the balcony. It’s very against the rules here to

have cigarettes on the balcony. We wheeled her out there. We sat her up. She

could hardly sit up. But the significance of lighting . . . We weren’t

obstructed. She was no longer . . . As far as I was concerned; we had moved

to comfort care, which is a euphemism for palliation. We had moved her to

comfort care. If comfort is the criteria, then you’re looking at emotional,

physical, psychological comfort, not only of her but also of the extended

family. The significant relationships are really important in terms of end-of-

life care and creating comfort so that the person dies with as much ease and

release as possible. So we got her out to the balcony. It was just through

wilfulness that she managed to sit on the edge of that bed with her dad. She

lit up. I don’t think she inhaled, but she lit up. That was a significant thing.

64

They were there for the duration. We came back in and she died within 12

hours of that. (Participant 2)

I think it’s a very personal—I think it’s an intensely personal time. I think

people should have a choice in how they do that . . . We had one lady whose

birthday the next week and they’d bought presents. They asked will we bring

them in? So they brought in the presents and they played music. They had

like a picnic. It sounds maybe disrespectful to other people, but for them it

was the right way to do it. I guess personal life is—they don’t—people don’t

know that they can personalise the death. Put up pictures, put up music, get

their own blankets, get whatever, and make it a personal experience, rather

than the stark white and tippy-toe around. So there was laughter. There was

crying. I think they made a real—and the mother wrote me afterwards and

said thank you, because it was something that she could then take with her,

rather than the tiptoeing around and the whispers in the corridor and the—

you know, sort of stuff. So anyway. So I guess that’s how I think of lots of—

I guess it’s my own experience, seeing somebody else’s. Then just this

woman saying that she appreciated being able to be involved in the care.

(Participant 4)

The findings of this phase of the study suggest that in the provision of end-of-life

care, all nurses can facilitate the creation of memories for families. At an advanced

level of practice, personalised meaningful experiences could be facilitated for

families of patients at the end of life. A relationship with the family and an intuitive

grasp of the situation is necessary for this to occur. Creating memories and positive

experiences is a satisfying aspect of end-of-life care practice for intensive care

nurses.

65

5.6 Discussion

Findings from phase one of this study identified categories relating to the value

nurses placed on the opportunity to participate in end-of-life care and their beliefs

that end-of-life care is an advanced practice skill. Participants felt that the value and

complexity of end-of-life care was not always recognised by their nursing colleagues

and other health care professionals. Variation in personal beliefs about end-of-life

care can thus influence how one approaches end-of-life care and the care provided to

patients and their families.

The ICU has some unique features and the provision of end-of-life care in this

context has some differences from end-of-life care in other settings. Specifically,

organisational and emotional support, the intensive care bedside environment and the

intensive care nurses relationship with the patient and the patient’s family are factors

that emerged from interviews with critical care nurses that may influence critical care

nurses’ end-of-life care experiences and practices. Nurses in this phase of the study

described interventions to modify the bedside environment and humanise the patient

to create a context that supported the provision of end-of-life care.

Nurses in this phase of the study described practices undertaken in the facilitation of

end-of-life care that included providing options and respecting choices, explaining

and providing information, providing comfort care, being there, and creating

memories and positive experiences. The practices that nurses described varied

according to their own personal beliefs, their comfort with death and dying and their

expertise in providing end-of-life care to patients and their families.

Importantly, from the findings of phase one, aspects of critical care nurses’ end-of-

life care practices similar to those presented in Clarke et al.’s (2003) domains of end-

of-life care were identified. In addition, other aspects of end-of-life care, such as the

value of participating in end-of-life care and the emotional support obtained through

collegial support networks, were consistent with evidence in the literature review

(Calvin et al., 2007; Fridh et al., 2009; Halcomb et al., 2004; Jones & FitzGerald,

1998; Kirchhoff et al., 2000; Shorter & Stayt, 2010).

66

5.7 Summary

The findings of phase one of this study provide insight into the end-of-life care

experiences and practices of critical care nurses. The content of the survey

instrument developed in phase two of this study was informed by the literature

review and the findings of phase one presented in this chapter. The next chapter

documents the development of the survey instrument and the methods used to obtain

expert review and pilot test the survey instrument.

67

6 PHASE TWO METHODS

6.1 Introduction

The purpose of phase two of this study was to develop a survey instrument informed

by an extensive review of the literature related to end-of-life care in critical care

settings and the findings of phase one of this study. Phase two included a review of

the survey instrument by a panel of experts and a pilot test of the survey instrument

to determine the reliability of the instrument and feasibility of methods planned for

use in phase three of this study. This chapter will first describe the development of

the survey instrument. The methods used in obtaining expert review of the survey

instrument including selection of panel members, review procedures and data

management and analysis will then be described. The process of pilot testing the

survey instrument will then be discussed, including the setting and sample,

procedures and data management and analysis.

6.2 Development of the survey instrument

A suitable survey instrument was not available to identify the end-of-life care

practices and the factors influencing these practices as reported by Australian critical

care nurses. Therefore, a survey instrument needed to be developed for use in this

study. Lynn (1986) describes a three-step process for the development of a survey

instrument. The three steps are identification of domains and subdomains of the

content of interest, item generation for each domain and subdomain, and formatting

of the instrument into a usable form. To answer the research questions in phase three

of this study, three main sections were required in the survey instrument: critical care

nurses’ end-of-life care practices, factors influencing the provision of end-of-life care

and demographic information. The first two steps in developing the survey

instrument—identifying the content of interest and item generation for each of the

sections of the survey instrument—is outlined in further detail below. The third step

in instrument development, as suggested by Lynn (1986), involves developing the

instrument into a usable format. Key Survey provided the platform for the online

survey (WorldAPP Key Survey, 2009, 2011). The existing template was used for the

68

survey with some minor adjustments made to the shading of alternate items to

improve readability. The guidelines for web questionnaires outlined by Dillman,

Smyth, and Christian (2009) were reviewed, and appropriate formatting changes

were made to enhance the readability and usability of the survey for participants.

6.2.1 Critical care nurses’ end-of-life care practices

The Critical care nurses’ end-of-life care practices section of the survey instrument

aimed to address the research question: What are the most frequently reported end-

of-life care practices of Australian critical care nurses? Following a review of the

literature, the domains of end-of-life care identified by Clarke et al. (2003) were

chosen as a framework for the Critical care nurses’ end-of-life care practices section

of the survey instrument. These domains provided comprehensive coverage of all

aspects of end-of-life care in the critical care setting identified in an extensive review

of the literature and the findings of interviews with critical care nurses in phase one

of this study (see Table 6–1). The six domains that incorporated clinicians’ practices

directly related to the care of the patients and their families were included in this

section of the survey instrument. The seventh domain—emotional and organisational

support for clinicians—was not included in this section of the survey instrument as it

does not relate directly to care of the patient; however, this domain was included in

the Factors influencing the provision of end-of-life care section of the survey

instrument.

The items in the Critical care nurses’ end-of-life care practices section of the survey

instrument were empirically derived from an extensive review of the literature in

relation to end-of-life care in critical care settings and the findings of phase one of

this study. The Critical care nurses’ end-of-life care practices section of the

instrument requires respondents to rate how often they undertake select practices

when caring for a patient at the end of life. These items were measured using a five-

point rating scale in which 1 = never, 2 = rarely, 3 = sometimes, 4 = often and 5 =

always. This rating scale was selected to provide evidence of the frequency with

69

which specified practices are undertaken in the provision of end-of-life care by

respondents to the survey.

Table 6-1: Comparison of Clarke et al.’s (2003) domains of end-of-life care,

subcategories developed from phase one interview data and initial a priori

scales in the Critical care nurses’ end-of-life care practices section of the survey

instrument

Domain (Clarke et al.

2003)

Phase one findings A priori scale in

instrument

Patient and family

centered decision making

Facilitating the family’s

experience: Providing

options and respecting

choices

Patient and family centred

decision making

Communication within the

team and with patients and

families

Facilitating the family’s

experience: Explaining

and providing information

Explaining and providing

information

Continuity of care — Continuity of care

Emotional and practical

support for patients and

families

Humanising the patient

Facilitating the family’s

experience

Emotional and practical

support for patients and

families

Symptom management

and comfort care

Modifying the bedside

environment

Comfort care—

environmental factors

Comfort care Symptom management

and comfort care

Spiritual support for

patients and families — Spiritual support

Note. The domains continuity of care and spiritual support were not identified in the

categories and subcategories of interview data in phase one of this study.

70

6.2.2 Factors influencing the provision of end-of-life care

The PARIHS framework was originally developed to describe factors influencing the

uptake of evidence in clinical practice (Kitson et al., 1998). As noted in the literature

review chapter, this framework provides a suitable framework for this study to

identify factors that may influence nursing practice in relation to end-of-life care.

Specifically, the elements of context, evidence and facilitation were identified as the

key domains to be explored with items in the survey instrument that identify the

respondents’ attitudes and beliefs, and perceptions of their workplace in relation to

factors that may influence the provision of care to patients at the end of life and their

families. The a priori scales within each domain in this section of the survey

instrument, presented in Table 6–2, were empirically derived from an extensive

review of the literature and the findings from phase one of this study. This section of

the survey instrument aimed to address the research question: What are the select

context, evidence and facilitation factors influencing end-of-life care as perceived by

Australian critical care nurses?

Table 6-2: Domains and associated a priori scales in the Factors influencing the

provision of end-of-life care section of the survey instrument

Domain A priori scales

Context Organisational culture

Resources

Evidence

Patient and family preferences

Nursing education and experience

Nurse practice knowledge

Facilitation

Emotional support

Nurses’ end-of-life care values and beliefs

Coordination of care delivery

The Factors influencing the provision of end-of-life care section of the survey

instrument included items from the Neonatal Palliative Care Attitude Scale

71

(NiPCAS; Kain, Gardner, & Yates, 2009). The 26 items of the NiPCAS (Kain et al.,

2009) were reviewed and considered to address some of the contextual elements

pertinent to end-of-life care in an adult critical care unit. Four items from the

NiPCAS were not included in the survey instrument for the current study as they

were deemed not relevant to the purpose of the current study or context of critical

care in the adult patient population. These items were about society’s attitude to

palliative care, pain management and the perception that caring for dying patients is

traumatic. The remaining items from the NiPCAS were adapted to reflect the adult

patient population and the term end-of-life care was substituted for palliative care.

Following review by the expert panel and pilot testing of the survey instrument, only

14 items from the NiPCAS were retained in the final version of the survey

instrument used in phase three of this study.

The Factors influencing the provision of end-of-life care section of the survey

instrument also includes the Death Anxiety Scale (DAS; Templer, 1970). The DAS

consists of 15 items and each is measured dichotomously (true/false) and then a score

out of 15 is calculated. Nine of the items are keyed true and six of the items are

keyed false. The DAS has also been converted to be measured on a seven-point

Likert scale (McMordie, 1979). To maintain consistency with the rest of the survey

instrument, in the current study, a five-point Likert scale was used. The DAS has

previously been tested for reliability and validity (Templer, 1970; McMordie, 1979)

and was included in the current study to provide evidence of critical care nurses’

comfort with death and dying.

In addition to the items from the NiPCAS and DAS, the Factors influencing the

provision of end-of-life care section of the survey instrument included items

generated from a review of the literature and the findings of phase one of this study.

The items in the Factors influencing the provision of end-of-life care section on the

survey instrument are measured using a five-point Likert scale in which 1 = strongly

disagree, 2 = disagree, 3 = neutral, 4 = agree and 5 = strongly agree. Although Kain

et al. (2009) located the neutral response to the far right of the scale in the NiPCAS

instrument to minimise response bias with participants choosing this option, in the

current study, the neutral response was located in the centre of the scale to maintain

consistency with the other sections of the instrument, avoiding respondent confusion.

72

6.2.3 Demographic information

Items in the demographic information section were developed to provide consistency

with the Australian Institute of Health and Welfare (AIHW, 2011) Nursing and

Midwifery Labour Force 2009 survey and the Intensive Care Resources and Activity:

Australia and New Zealand 2007/2008 report (Drennan et al., 2010) to permit

comparison of the sample to the broader nursing and critical care nursing

populations. The demographic information collected included age, gender, location,

years of experience as a registered nurse, years of experience as a critical care nurse,

qualifications, current workplace, type of work, length of time since the participant

had provided direct patient care and the length of time since the participant had

provided end-of-life care.

6.3 Expert panel

In addition to the steps outlined by Lynn (1986), DeVellis’s (2003) guidelines for

developing a survey instrument include additional steps related to review of the

survey instrument by a panel of experts and pilot testing. This section outlines the

methods used in obtaining expert review of the survey instrument, while the methods

used in pilot testing are discussed in Section 6.4.

6.3.1 Participants

A purposive sampling technique was used to invite six registered nurses identified as

having expertise in critical care, palliative care and/or instrument development to

form a panel to review the survey instrument. The sample included an international

expert in end-of-life care in critical care settings. The purposive sampling technique

is an appropriate sampling strategy for the purpose of selecting participants who can

provide feedback on the content of the survey instrument. A sample size of six

experts is recommended when ascertaining the validity of an instrument by

calculating the content validity index (CVI). A sample of six enables at least one

73

person to disagree with an item, while still maintaining the content validity of the

item within an acceptable range (Lynn, 1986; Polit & Beck, 2006).

6.3.2 Procedure

The panel were contacted via email (see Appendix 13.7) and invited to participate in

the expert review of the survey instrument. To enable the panel of experts to assess

the instrument, they were provided with an information sheet (Appendix 13.8)

detailing the purpose and scope of the survey, research questions, definitions of key

terms (end-of-life care, practices, context, evidence and facilitation) and the

measurement scale on which items in the survey are measured (Davis, 1992; Grant &

Davis, 1997; Lynn, 1986). Each panel member was also provided with a uniform

resource locator (URL) link to the online survey, where they could complete their

review. Key Survey software version 7.2 provided the platform for the online survey

during phase two of the current study.

The validity of an instrument is the degree to which it measures what it was

developed to measure (Polit & Beck, 2008). The panel of experts assessed the

content validity of the online survey instrument. The panel were asked to rate each

item for the relevance to end-of-life care in the critical care setting using a four-point

ordinal rating scale in which 1 = not relevant, 2 = somewhat relevant, 3 = quite

relevant and 4 = highly relevant (Davis, 1992; Grant & Davis, 1997; Lynn, 1986;

Polit & Beck, 2006). These relevance ratings were used to calculate a CVI to indicate

the perceived relevance of individual items (I-CVI) and the overall content validity

of the instrument, or scale content validity index (S-CVI), as assessed by the experts

(Lynn, 1986; Polit & Beck, 2006).

Face validity of the instrument was also assessed by asking the panel of experts to

provide comment on the content in regard to coverage of the domains of interest and

to suggest the addition, deletion and revision of individual items (Grant & Davis,

1997; Lynn, 1986; Polit, Beck, & Owen, 2007). The experts were also asked to

comment on the readability of the survey instrument.

74

6.3.3 Data management and analysis

The I-CVI was calculated by hand from the number of experts that rated the item as

3 or 4 divided by the total number of experts. These individual item scores were then

combined and divided by the total number of items to produce an average CVI, that

reflects the overall content validity of the instrument (S-CVI) (Polit & Beck, 2006).

An I-CVI score of 0.80 or greater and an S-CVI score of 0.90 were considered

satisfactory (Grant & Davis, 1997; Lynn, 1986; Polit & Beck, 2006; Polit et al.,

2007). Individual items with an I-CVI less than 0.80 were deleted or modified.

Comments by the expert panel guided the development of additional new items (Polit

et al., 2007).

The findings of the analysis and the expert panel’s comments about the survey’s

content and readability were reviewed and discussed with the supervisory team.

Following analysis of the expert panel’s responses, the survey instrument was

revised and resent to the panel for further comment.

6.4 Pilot test

The final step in the development of the survey instrument in this study involved a

pilot test, to obtain preliminary evidence of the reliability and validity of the newly

developed survey instrument. Pilot testing also assessed the feasibility of methods

planned for use in the collection and analysis of quantitative data in phase three of

this study. An initial pilot test was undertaken in Semester 1, 2010. Some of the

scales of the instrument were not found to be robust measures of the constructs of

interest in this study. As a result, the instrument was revised and a repeat pilot test-

retest was undertaken in Semester 1, 2011, to reassess the validity and reliability of

the instrument. The setting and sample, procedure, and data management and

analysis used in pilot testing the survey instrument are described below.

75

6.4.1 The setting and sample

The pilot test was conducted on a sample of registered nurses with experience in

critical care. A convenience sampling technique was used to invite registered nurses

currently employed in a critical care setting and enrolled in a postgraduate critical

care nursing subject at an Australian university to participate in the pilot test of this

study.

6.4.2 Procedure

Potential participants were contacted via an email distributed through the Key Survey

version 7.2 software (WorldAPP Key Survey, 2009) to individual students’

university email addresses. A copy of the email invitation, reminder email and study

information sheet sent to pilot test participants is provided in Appendices 13.9 and

13.11. The ‘capture respondent email address’ feature was used to enable the

researcher to view the list of email addresses of respondents who took the survey

(completes and incompletes) to enable follow-up reminder emails to be sent to

selected participants. This feature also captures individual email addresses on survey

responses, enabling the researcher to link data for the purposes of test-retest

reliability. In addition to completing the survey, the participants were provided with

space to comment on the readability and content of the survey instrument.

Reminder emails were sent to non-responders two and four weeks after initial

distribution. The survey was open for six weeks. Two weeks later, all students who

had participated by completing the survey were invited to complete the survey again

to permit test-retest reliability. This time frame is considered a reasonable period of

time to limit the possibility of participant responses changing as a result of

experience or changes in practices (Marx, Menezes, Horovitz, Jones, & Warren,

2003).

76

6.4.3 Data management and analysis

Survey responses were exported from Key Survey into PASW version 18 for data

analysis. Cronbach’s alpha was calculated to identify the internal consistency of

items in each scale identified a priori. The corrected item-scale correlation and the

Cronbach’s alpha if an item was deleted were reviewed for each item and items were

considered for deletion if they were found to influence the reliability of the scale

(DeVellis, 2003).

Test-retest reliability was conducted by calculating the Pearson product-moment

correlation coefficient (r) to determine the reliability of the survey instrument in

obtaining consistent results on repeated measures (Polit & Beck, 2008). Cohen’s

(1988) criteria for interpreting the size of correlation coefficients with ±.1 a small

effect, ±.3 a medium effect and ±.5 a large effect, were used in the reporting of

results. Reliability coefficients of greater than .70 were considered satisfactory (Hair,

Black, Babin, Anderson, & Tatham, 2006), although lower reliability coefficients

may be considered satisfactory when measuring affective domains or in exploratory

research (Field, 2009; Hair et al., 2006).

6.5 Summary

In summary, the development of a survey instrument was necessary to achieve the

purpose of this study and answer the research questions. The scales in the survey

instrument and the items generated to capture the content of the scales were

empirically derived from the findings of phase one of the current study and an

extensive review of the literature. A selection of items from the NiPCAS and the

DAS were also included in the survey instrument. Review of the survey instrument

by an expert panel provided evidence of the content validity. Pilot testing was also

undertaken to obtain preliminary evidence of the reliability of the survey instrument

and feasibility of methods planned for use in phase three of this study. The results

obtained from expert review and pilot testing of the survey instrument are provided

in the next chapter.

77

7 PHASE TWO RESULTS

7.1 Introduction

Review by an expert panel and pilot testing of the survey instrument provided

preliminary evidence of the validity and reliability of the survey instrument

developed for this study. The results of phase two of this study also informed the

modifications undertaken to the survey instrument and minor changes to the methods

planned for use in phase three. This chapter documents the findings from the expert

review and pilot test of the survey instrument. Specifically, this chapter will address

the composition of the expert panel, the content validity of the survey instrument

based on analysis of the ratings provided by the expert panel, the characteristics of

participants involved in pilot testing the instrument and the results of internal

consistency and test-retest reliability of the scales of the survey instrument proposed

a priori.

7.2 Expert panel

Initially, seven registered nurses were identified as potential participants and invited

to join the current study’s expert panel and participate in the review of the survey

instrument. Four of these registered nurses agreed to participate, completing a review

of the survey instrument online. Two registered nurses declined to participate, citing

workload issues and perceived lack of expertise in the topic under investigation.

Another registered nurse declined to participate and instead referred the researcher to

a nursing colleague who was more closely aligned with the topic under investigation

and, upon invitation, this person agreed to participate. A further two potential

participants were then approached to participate, and one agreed to participate and

the other did not respond to the invitation. The final panel, therefore, comprised six

participants. These participants were employed in academic or research positions and

their primary areas of expertise are outlined in Table 7-–1. Clinician views on the

survey instrument were obtained during the pilot testing phase of instrument

development in this study.

78

Table 7-1: Expert panel members

Reviewer Area of expertise

1 Neonatal intensive care and instrument development

2 Critical care nursing

3 Palliative care

4 International expert in end-of-life care in intensive care settings

5 Palliative care

6 Critical care, end-of-life care and instrument development

The CVI for the survey instrument overall (S-CVI) and for each item (I-CVI) was

calculated from the relevance ratings provided by the expert panel. The initial S-CVI

for the survey instrument was .88. The initial S-CVI for the Critical care nurses’

end-of-life care practices section of the survey instrument was .90 and the CVI for

each item (I-CVI) in this section of the survey instrument is provided in Table 7–2.

Table 7-2: Content validity index for each item (I-CVI) in the Critical care

nurses’ end-of-life care practices section of the survey instrument

Item I-CVI

Patient and family centred decision making

Provide the family with options about the care of the patient 1.00

Ask the family if they would like to be involved in the care of the patient 1.00

Ask the family their preferences for the patient’s care 1.00

Attend family meetings with the medical team 1.00

Reassure the family that they are not responsible for any EOL decisions made .66

Ask the family about their knowledge of the patient’s wishes for EOLC .83

Explaining and providing information

Answer the family’s questions about the patient’s condition 1.00

Fully inform the family about all aspects of the plan of care 1.00

Explain all interventions undertaken in the care of the patient 1.00

Provide information to the family about the dying process 1.00

79

Item I-CVI

Explain to the family what is happening to the patient 1.00

Encourage the family to reminisce about the patient 1.00

Listen to the family reminisce about the patient .83

Continuity of Care

Arrange for transfer of the patient out of the ICU after a decision to withdraw

treatment .50

Document the family’s preference for the care of their relative .83

Introduce the nursing staff on the oncoming shift to the patient and family .83

Handover the family’s preference for the patient’s EOLC to the oncoming

nurse 1.00

Emotional and practical support for patients and families

Permit the family to visit outside of regular visiting hours 1.00

Adhere to unit policy on the number of visitors at the patient’s bedside .33

Encourage the family to talk to the patient 1.00

Encourage the family to touch the patient 1.00

Provide the family with access to a private room to congregate 1.00

Arrange for a counsellor or social worker to meet with the family .83

Stay in the room with the patient and family .66

Ask the family about their feelings .83

Ask the family how they are coping 1.00

Provide opportunities for the family to create positive memories .83

Comfort care—environmental factors

Move the patient to a more private area of the unit 1.00

Create a calm, peaceful environment 1.00

Turn off monitoring equipment 1.00

Silence alarms on all equipment 1.00

Symptom management and comfort care

Administer fluids to the patient to maintain hydration .66

Keep the patient sedated .66

Provide adequate pain management 1.00

Provide care to maintain patient hygiene 1.00

Provide pressure area care 1.00

80

Item I-CVI

Remove the endotracheal tube if the patient is breathing spontaneously 1.00

Maintain the patient’s airway using oro/endotracheal suction .66

Spiritual support

Ask the family about spiritual needs 1.00

Offer to arrange for a spiritual advisor or pastoral care representative to visit

the patient and family 1.00

Support the family to participate in cultural practices related to dying 1.00

Ask the family about cultural needs 1.00

Note. EOL = end of life; EOLC = end-of-life care; ICU = intensive care unit.

The initial S-CVI for the Factors influencing the provision of end-of-life care section

of the survey instrument was .84 and the I-CVI in this section of the survey

instrument is provided in Table 7–3.

Table 7-3: Content validity index for each item (I-CVI) in the Factors

influencing the provision of end-of-life care section of the survey instrument

Item I-CVI

Organisational culture

The medical staff support EOLC for patients in my unit 1.00

In my Unit, the team expresses its opinions, values and beliefs about

providing care to dying patients .83

All members of the health care team in my Unit agree with and support EOLC

when it is implemented for a patient 1.00

After a decision to withdraw life-sustaining treatment is made, nurses are left

to manage the care of the patient 1.00

After a decision is made to withdraw treatment, the intensive care medical

team in my unit reduce their contact with the patient .66

In my Unit, the staff go beyond what they feel comfortable with in using

technological life support .83

In my Unit, staff are asked by families to continue extending care beyond .50

81

Item I-CVI

what they feel is right

EOLC should be allocated to experienced nurses .80

Resources

The physical environment of my Unit is ideal for providing EOLC 1.00

My Unit is adequately staffed for providing the needs of dying patients and

their families 1.00

When a patient dies in my Unit, I have sufficient time to spend with the

family 1.00

There are policies/guidelines to assist in the delivery of EOLC in my Unit .83

Lack of staff results in no ‘down time’ between the death of a patient and the

admission of the next patient .83

Patient and family preferences

In my Unit, families are involved in decisions about their dying relative 1.00

In my Unit, withdrawal of treatment is delayed whilst waiting for relatives to

arrive 1.00

Nursing education and experience

I have received in-service education that assists me to support and

communicate with families of dying patients .83

I have had experience of providing EOLC to patients and their families 1.00

I am often exposed to death in the critical care environment .83

EOLC is necessary in critical care nursing education .83

Emotional support

When a patient dies in my Unit, counselling is available if I need it 1.00

In my Unit, nurses are expected to cope with death and dying 1.00

I feel supported when caring for a patient at the EOL 1.00

After caring for a patient at the EOL, my colleagues will ask me if I am OK 1.00

Formal debriefing occurs after providing EOLC 1.00

Staffing is arranged to allow nurses to have a rest break after the death of a

patient .66

Nurses’ end-of-life care values and beliefs

My previous experiences of providing EOLC have been rewarding 1.00

I feel a sense of personal failure when a patient dies .50

82

Item I-CVI

It is a privilege to care for a patient at the EOL and their family 1.00

Nurses can influence the patient and family’s experience of EOLC 1.00

EOLC is against the values of critical care nursing .16

EOLC is as important as curative care in the critical care environment .83

Curative care is more important than EOLC in the critical care environment .33

My personal attitudes about death affects my willingness to deliver EOLC .83

Coordination of care delivery

Patients should be transferred out of intensive care after a decision is made to

withdraw life-sustaining treatment .66

Nurses that know the patient and family should be allocated to care for the

patient to maintain continuity of care .83

Note. EOL = end of life; EOLC = end-of-life care.

Items with a CVI below .80 were reviewed. Seven items were deleted as they were

not considered relevant to end-of-life care in the adult critical care context, were

similar to other items rated as relevant by the expert panel and/or achieved a very

low CVI (see Table 7–4). Following deletion of these seven items, the overall

content validity index of the survey instrument (S-CVI) was recalculated at .91 and

identified as at a satisfactory level (Grant & Davis, 1997; Lynn, 1986; Polit & Beck,

2006; Polit et al., 2007). The revised S-CVI for the Critical care nurses’ end-of-life

care practices section of the survey instrument was .93 and the revised S-CVI for the

Factors influencing the provision of end-of-life care section of the survey instrument

was .89.

83

Table 7-4: Items deleted from the survey instrument following review by the

expert panel

Item

Reassure the family that they are not responsible for any EOL decisions made

Arrange for transfer of the patient out of the ICU after a decision to withdraw

treatment

Adhere to unit policy on the number of visitors at the patient’s bedside

In my Unit, staff are asked by families to continue extending care beyond what they

feel is right

EOLC is against the values of critical care nursing

Staffing is arranged to allow nurses to have a rest break after the death of a patient

Patients should be transferred out of intensive care after a decision is made to

withdraw life-sustaining treatment

Note. EOL = end of life; EOLC = end-of-life care; ICU = intensive care unit.

Although the CVI of the items I feel a sense of personal failure when a patient dies

(I-CVI = .50) and After a decision is made to withdraw treatment, the intensive care

medical team in my unit reduce their contact with the patient (I-CVI = .66) were low,

these items were retained in the survey. Death as failure and the withdrawal of the

medical team following withdrawal of life-sustaining treatment were identified as

important items that may influence the provision of end-of-life care by participants

currently working as clinicians in an ICU in phase one of this study. In addition, the

perception of death as failure and the withdrawal of the medical team following

withdrawal of life-sustaining treatment have been identified in other previous

qualitative research undertaken in Australia (Halcomb et al., 2004; Heland, 2006).

The CVI of items pertaining to symptom management and the physical care of the

patient, Keep the patient sedated (I-CVI = .66), Maintain the patient’s airway using

oro/endotracheal suction (I-CVI = .66) and Administer fluids to the patient to

maintain hydration (I-CVI = .66) were below the level deemed satisfactory (I-CVI =

.80). However, these items were retained in the survey as they had been identified in

84

the previous literature and/or the phase one findings of this study as practices

undertaken by critical care nurses in the provision of end-of-life care.

The item Stay in the room with the patient and family (I-CVI = .66) was modified to

include if the family wants, acknowledging that practices may be adapted to meet the

needs of different families. The item Curative care is more important than end-of-life

care in the critical care environment (I-CVI = .33) was revised to read End-of-life

care is as important as curative care in the critical care environment. Following

review by the expert panel, two additional items were added to the survey

instrument: Ask family members if there is anyone that they would like to be there

and In my unit, when a diagnosis with a likely poor outcome is made, families are

informed of their end-of-life care options.

After modifications were made to the survey instrument, it was recirculated to the

expert panel and additional feedback from four of the experts was provided by return

email. This feedback consisted of suggestions for minor changes such as deleting

‘Australian’ from the survey title, standardising the use of either ‘unit’ or ‘Unit’ and

the formatting and location of response options within the survey instrument.

7.3 Pilot test

During Semester 1, 2010, an initial pilot test of the survey instrument was

undertaken. In total, 17 out of 51 critical care nurses enrolled in a postgraduate

nursing subject at an Australian university participated in the initial pilot test of the

instrument, a response rate of 33%. Retest of the survey was completed by eight

participants. A further three accessed the survey and provided demographic

information but did not complete the rating scales or subsequent questions.

The findings of this initial pilot test revealed that some of the a priori scales of the

survey instrument were not robust measures of the constructs of interest in this study.

As a result, the instrument was significantly revised. The item I have had experience

providing end-of-life care to dying patients and their families was deleted as this

85

information was obtained from participant response to the item in the demographic

section: Please specify the last time that you provided end-of-life care to a patient in

a critical care unit. The items My previous experiences of providing end-of-life care

have been rewarding and Formal debriefing occurs after providing end-of-life care

were also deleted. These items were considered ambiguous and their intent was

captured by other items in the survey instrument. Two items were also modified to

read Keep the patient sedated, if appropriate and Administer analgesia via

intravenous infusion.

Additional items were added to the survey instrument following reflection on the

findings of the initial pilot test, the findings of phase one of this research study and

further review of the previous literature of end-of-life care in critical care settings.

Review of the PARIHS framework guiding the development of the Factors

influencing the provision of end-of-life care section and consideration of the items

within each scale proposed a priori assisted in identifying additional items relevant to

the constructs of interest in this research study. Although these additional items,

presented in Table 7–5 and Table 7–6, are listed under the scale headings proposed a

priori, only section headings (Demographics, Practices and Factors) were included in

the final survey instrument circulated to participants in phases two and three of this

study and the a priori scale headings were not included.

Table 7-5: Additional items included in the Critical care nurses’ end-of-life care

practices section of the survey instrument

A priori scale Additional items

Continuity of care Ask the critical care team to document the plan of care /

treatment orders

Comfort care—

Environmental

factors

Adjust (dim or turn off) the lighting around the patient’s bed

space

Play music

Put up photographs of the patient, family members and/or pets

Cluster/coordinate care to minimise interruptions and provide

privacy for the patient and family

86

Draw curtains to provide privacy for the patient and family

Place chairs around the bed for the family

Symptom

management and

comfort care

Remove the endotracheal tube if the patient is breathing

spontaneously

Initiate (or implement) weaning from mechanical ventilation

once the decision to withdraw mechanical ventilation has been

made (ie. reduce FiO2, rate and/or tidal volume)

Administer medications to prevent/minimise dyspnoea and

respiratory secretions

Regularly obtain and document patient vital signs during the

provision of end-of-life care

Table 7-6: Additional items included in the Factors influencing the provision of

end-of-life care section of the survey instrument

A priori scale Additional items

Organisational

culture

In my unit, EOLC is mostly allocated to junior nurses

In my unit, nurses view the provision of EOLC as an important

component of critical care nursing

Resources Private rooms are available to care for the patient at the EOL

My unit is adequately equipped (chairs/recliners/waiting room

etc.) to support the comfort needs of the family during EOLC

When a patient dies in my Unit, nurses have sufficient time to

spend with the family

Patient and family

preferences

In my unit, the patient’s preferences (or their likely preference

according to family) for EOLC are considered

In my unit, families are given adequate time to consider

decisions to withhold/withdraw treatment for the patient (their

relative)

In my unit, family meetings with the health care team are held

to resolve differing views about a patient’s treatment

Nursing education

and experience

I have received in-service education that assists me to provide

care and manage patient symptoms at the EOL

87

My undergraduate nursing education included EOLC content

My postgraduate nursing education included EOLC content

relevant to the critical care context

I draw on evidence from the literature to provide EOLC to

patients and their families

I feel adequately prepared to care for patients at the EOL

I don’t know what to say to the family of a patient at the EOL

I have limited knowledge of symptom management for

patient’s at the EOL

I have learnt from other nurses how to provide EOLC

Nurse practice

knowledge

Analgesia should be titrated to keep the patient comfortable

even if this hastens death

Pressure area care should continue to be provided for a patient

at the EOL

EOLC should include care of the patient’s family

Abrupt withdrawal of mechanical ventilation may contribute to

dyspnoea

When withdrawing mechanical ventilation at the EOL,

ventilator weaning (such as reductions in FiO2 and PEEP)

should occur prior to extubating the patient or placing the

patient on a T-piece

Intravenous fluids should be provided to maintain hydration for

the patient at the EOL

The provision of EOLC requires emotional detachment

It is OK to cry when a patient dies

Emotional support My supervisors/managers ensure staff caring for patients at the

EOL are supported

My colleagues appreciate the stressors associated with caring

for patients at the EOL

My supervisors/managers provide guidance that assists me to

provide EOLC

Nurses’ end-of-life

care values and

My own personal experiences of death have influenced the care

I provide to patients at the EOL

88

beliefs

Coordination of

care delivery

In my unit, a plan of care for the patient is clearly documented

by the medical team after a decision to withhold or withdraw

treatment is made

In my unit, the intensive care team and the patient’s primary

medical or surgical team agree on a plan of care

In my unit, patients remain in ICU for EOLC following a

decision to withhold or withdraw treatment

Note. EOL = end of life; EOLC = end-of-life care; ICU = intensive care unit.

Due to the significant adjustments made to the survey instrument, it was determined

that a repeat pilot test-retest was required to reassess the validity and reliability of the

instrument prior to circulation of the instrument to a national sample in phase three

of this research study. The findings presented in the remainder of this chapter refer to

the results of the repeat pilot test and retest of the survey instrument.

7.3.1 Demographic information

During Semester 1, 2011, 62 potential participants were invited to participate in the

repeat pilot test of the survey instrument. In total, 31 critical care nurses accessed and

commenced the survey, a response rate of 50%. Five participants only partially

completed the survey. Of these, responses from the two participants who completed

both the demographic information section and the Factors influencing the provision

of end-of-life care section were included in the data analysis. Responses from the

three participants who only completed the demographic information section and

partially completed the Factors influencing the provision of end-of-life care section

were deleted and not included in the data analysis. The final response rate was 45%

(n = 28). Test-retest reliability of the survey instrument was obtained by inviting

participants to complete the survey instrument again, two weeks later. This retest of

the survey was completed by 11 participants. A further two accessed the survey and

89

provided demographic information but did not complete the rating scales or

subsequent questions.

The mean age of registered nurses completing the Semester 1, 2011, pilot test of the

survey instrument was 32.2 years (SD = 7.9) and 17.8% of participants were male (n

= 5). Years of experience in nursing ranged from one to 39 years with the majority of

participants (57.1%) having five years or less experience in nursing. Years of

experience working as a nurse in critical care ranged from one to 12 years with a

mean of 4.7 years (SD = 3.3). The participants for the pilot test were recruited from a

postgraduate critical care nursing subject, and their experience in nursing and critical

care most likely reflects registered nurses pursuing further studies in their chosen

specialty area of nursing practice. Table 7–7 outlines the characteristics of

participants in the Semester 1, 2011, pilot test. The majority of participants reported

that their highest qualification in nursing was an undergraduate degree or hospital

certificate (89.2%). Only 10.7% of participants had a postgraduate degree; however,

all participants were currently enrolled in a postgraduate critical care nursing course.

Table 7-7: Demographic characteristics of pilot test participants (n = 28)

Characteristic %

Age

<25 28.6

25–29 10.7

30–34 25

35–39 17.9

40–44 14.3

45+ 3.6

Gender

Female 82.1

Male 17.9

Nursing experience

0–5 57.1

6–10 17.9

90

11–15 14.3

16–20 0

20+ 10.7

Critical care experience

0–5 75

6–10 14.3

11–15 10.7

The majority of participants were currently employed in a critical care unit located in

a major city (78.6%). The remainder of participants reported employment in a critical

care unit in a regional area (21.4%). Most of the participants (92.9%) were employed

in a general ICU, while the remainder were employed in a combined ICU, CCU and

HDU (high dependency unit). Direct patient care was identified as the primary role

of the majority of participants in their current employment (92.9%). All participants

in the Semester 1, 2011, pilot test had provided direct patient care within the last

month and end-of-life care within the last year (see Table 7–8).

Table 7-8: Pilot test participants’ current workplace and participation in

nursing work (n = 28)

Characteristic %

Current workplace

General ICU (integrated medical/surgical including

ICU managed HDU and specialty ICUs)

92.9

Integrated ICU/CCU/HDU 7.1

Type of work

Clinical (direct patient care) 96.4

Nursing management (administration) 3.6

Time since provision of direct patient care

Within the last week 92.9

Within the last month 7.1

Time since provision of end-of-life care

91

Within the last week 3.6

Within the last month 32.1

Within the last 12 months 64.3

Note. ICU = intensive care unit; CCU = coronary care unit; HDU = high dependency

unit.

7.3.2 Critical care nurses’ end-of-life care practices

In the pilot test, the Critical care nurses’ end-of-life care practices section of the

survey instrument consisted of 50 items across seven a priori scales conceptually

derived from evidence in the literature and supported by the findings of phase one of

this study. In the following section, each of these scales will be described and the

results of internal consistency and test-retest reliability will be reported.

Scale: Patient and family centred decision making

The items in the patient and family centred decision making scale pertain to the

practices undertaken by the individual nurse (survey respondent) to facilitate

involvement of the family in care planning and decision making for the patient (their

relative) at the end of life. Family was selected as the focus due to the inability or

limited ability of the patient in critical care to participate. In the survey instrument,

participants were asked to rate the frequency with which they engage in the practice

identified in each item. The items contributing to the patient and family centred

decision making scale are presented in Table 7–9. Internal consistency of this scale

was satisfactory with a Cronbach’s alpha of .81; however, only a moderate

correlation coefficient (r = .35) that was not statistically significant was obtained on

test-retest reliability of this scale.

92

Table 7-9: Items in the patient and family centred decision making scale

Item

Provide the family with options about the care of the patient

Ask the family if they would like to be involved in the care of the patient

Ask the family their preferences for the patient’s care

Attend family meetings with the intensive care team

Ask the family about their knowledge of the patient’s wishes for EOLC

Note. EOLC = end-of-life care.

Scale: Explaining and providing information

The items in this scale include practices undertaken by the individual nurse (survey

respondent) to explain and provide information to the family of the patient at the end

of life (see Table 7–10). Participants rated the frequency with which they undertake

each item in the provision of end-of-life care. This scale achieved a Cronbach’s alpha

of .84 indicating the scale is internally consistent; however, once again the test-retest

reliability of this scale obtained a moderate correlation coefficent (r = .53) that was

not statistically significant (see Table 7–16).

Table 7-10: Items in the explaining and providing information scale

Item

Answer the family’s questions about the patient’s condition

Fully inform the family about all aspects of the plan of care

Explain all interventions undertaken in the care of the patient

Provide information to the family about the dying process

Explain to the family what is happening to the patient

93

Scale: Continuity of care

The continuity of care scale was designed to measure the practices undertaken by the

individual nurse (survey respondent) to smooth transitions between individual care

providers (both nurses and other health care providers) for the family and coordinate

care for the patient at the end of life. This scale was originally included in the

instrument as it is one of the seven domains of quality end-of-life care identified by

Clarke et al. (2003). The indicators and associated behaviours in this domain relate

mainly to the transfer or non-transfer of patients from critical care (Clarke et al.,

2003). In phase one of this study, participants reported that it was usual practice that,

after withdrawal of treatment, patients remained in their ICU for end-of-life care and

this view was supported by the expert panel.

The items in the continuity of care scale, listed in Table 7–11, were conceptually

derived and developed from the literature on maintaining continuity of care and

coordination of care as well as the findings from phase one of this study. The items

in this section are designed to measure practice, rather than barriers to or facilitators

of continuity of care, which is addressed in the Factors influencing the provision of

end-of-life care section of the survey instrument. This scale obtained a Cronbach’s

alpha of .61 and a moderate correlation coefficient (r = .46) on test-retest reliability

that was not statistically significant and, therefore, does not appear to provide a

robust measure of the construct continuity of care. Although the internal consistency

for this scale was weak, it was considered a theoretically important construct, and

further testing of this construct with a larger sample in phase three of this study was

needed. Additionally, in response to these results an item: Liaise with other members

of the healthcare team to meet identified needs of the patient and family, was added

to the survey instrument for further testing in phase three of this study.

94

Table 7-11: Items in the continuity of care scale

Item

Document the family’s preference for the care of the patient

Introduce the nursing staff on the oncoming shift to the patient and family

Hand over the family’s preferences for the patient’s EOLC to the oncoming nurse

Ask the intensive care team to document the plan of care/treatment orders

Note. EOLC = end-of-life care.

Scale: Emotional and practical support for patients and families

The items in this scale include practices undertaken by the individual nurse (survey

respondent) to provide emotional and practical support for patients and families. The

12 items in this scale, listed in Table 7–12, obtained a Cronbach’s alpha of .69 and a

moderate correlation coefficient (r = .45) that was not statistically significant on test-

retest reliability (see Table 7–16). All items were retained in the survey instrument

for circulation to potential participants in phase three of this study.

Table 7-12: Items in the emotional and practical support for patients and

families scale

Item

Encourage the family to reminisce about the patient

Listen to the family reminisce about the patient

Permit family to visit outside of regular visiting hours

Encourage the family to talk to the patient

Encourage the family to touch the patient

Provide the family with access to a private room to congregate

Ask family members if there is anyone they would like to be there

Arrange for a counsellor or social worker to meet with the family

Stay in the room with the patient and family, if the patient and family wants

Ask the family about their feelings

95

Ask the family how they are coping

Provide opportunities for the family to create positive memories

Scale: Comfort care—environmental factors

The comfort care—environmental factors scale consisted of items related to the

practices undertaken by the individual nurse (survey respondent) to create a

comfortable physical environment for patients at the end of life and their families.

The items included in this scale are outlined in Table 7–13. A Cronbach’s alpha of

.66 was obtained on pilot testing of this scale. A moderate correlation coefficient (r =

.48) that was not statistically significant was obtained on test-retest reliability of this

scale. Despite these results, the environment is considered important to the provision

of a positive experience of end-of-life care for patients and their families and,

therefore, the items in this scale were retained in the final version of the survey

instrument.

Table 7-13: Items in the comfort care—environmental factors scale

Item

Move the patient to a more private area of the unit

Create a calm, peaceful environment

Turn off monitoring equipment

Silence alarms on all equipment

Adjust (dim or turn off) the lighting around the patient’s bed space

Play music

Put up photographs of the patient, family members and/or pets

Cluster/coordinate care to minimise interruptions and provide privacy for the patient

and family

Draw curtains to provide privacy for the patient and family

Place chairs around the bed for the family

96

Scale: Symptom management and comfort care

The items in the symptom management and comfort care scale incorporated the

practices undertaken by the individual nurse (survey respondent) to relieve

symptoms and provide physical care to the patient at the end of life (see Table 7–14).

The internal consistency of this scale was considered satisfactory with a Cronbach’s

alpha of .73. Although a large correlation coefficient was obtained on test-retest

reliability (r = .58), it was not statistically significant. The practices undertaken in the

provision of physical care of the patient including symptom management at the end

of life are important, and these items were retained in the final version of the survey

instrument.

Table 7-14: Items in the symptom management and comfort care scale

Item

Administer fluids to the patient to maintain hydration

Keep the patient sedated, if appropriate

Administer analgesia via intravenous infusion

Provide care to maintain patient hygiene

Provide pressure area care

Maintain the patient’s airway using oro/endotracheal suction

Remove the endotracheal tube if the patient is breathing spontaneously

Initiate (or implement) weaning from mechanical ventilation once the decision to

withdraw mechanical ventilation has been made (i.e., reduce FiO2, rate and/or tidal

volume)

Administer medications to prevent/minimise dyspnoea and respiratory secretions

Regularly obtain and document patient vital signs during the provision of EOLC

Note. EOLC = end-of-life care.

97

Scale: Spiritual support

The items in the spiritual support scale, presented in Table 7–15, include practices

undertaken by the individual nurse (survey respondent) to identify and support the

spiritual and cultural needs of the family. The items in this scale were found to be

internally consistent (α = .90), although the test-retest reliability of this scale was low

(r = .33).

Table 7-15: Items in the spiritual support scale

Item

Ask the family about spiritual needs

Offer to arrange for a spiritual advisor or pastoral care representative to visit the

patient and family

Ask the family about cultural needs

Support the family to participate in cultural practices related to dying

Although the majority of a priori scales in the Critical care nurses’ end-of-life care

practices section of the survey instrument were found to have internal consistency,

only low to moderate correlation coefficients were obtained on test-retest reliability

on all scales in the practices section (see Table 7–16). A limitation of test-retest as a

measure of reliability is that participants’ responses can change over time due to

changes in knowledge, experience and attitude, for example (Polit & Beck, 2010). It

is possible that participants’ awareness of end-of-life care practices was heightened

following initial completion of the survey and their engagement in selected practices

changed between completing the surveys. These changes in practice may only be

short term in response to raised awareness. In addition, only a very small number of

participants completed the survey twice. This small sample size may contribute to

variability in responses, as slight changes in participant responses to individual items

can affect the overall reliability of the scales. Despite the low to moderate correlation

coefficients obtained on test-retest reliability in this section of the survey instrument,

the internal consistency within the scales was generally satisfactory. Therefore, this

98

section of the survey instrument was considered adequate to achieve the purpose of

this study, identifying the frequency of engagement in selected end-of-life care

practices as reported by critical care nurses.

Table 7-16: Internal consistency and test-retest reliability of the a priori scales

in the Critical care nurses’ end-of-life care practices section of the survey

instrument

Scale α r

Patient and family centred decision making .81 .35

Explaining and providing information .84 .53

Continuity of care .61 .46

Emotional and practical support for patients and families .69 .45

Comfort care—environmental factors .66 .48

Symptom management and comfort care .73 .58

Spiritual support .90 .33

*p < .05, two-tailed.

7.3.3 Factors influencing the provision of end-of-life care

As previously discussed, the three key elements of context, evidence and facilitation

proposed in the PARIHS framework provided the framework guiding the design of

the Factors influencing the provision of end-of-life care section of the survey

instrument. This section comprised of 58 individual items across eight scales

proposed a priori within the three key elements. Each scale will be described and the

results of internal consistency and test-retest reliability will be outlined below.

Context

The key element of context was represented in the survey instrument in this study by

the scales organisational culture and resources.

99

Scale: Organisational culture

In this study, organisational culture is defined as the values and beliefs in relation to

end-of-life care generally shared within the critical care unit as perceived by the

individual nurse (survey respondent). The items in this scale are listed in Table 7–17.

The first three items in this scale are originally from Kain et al.’s (2009) organisation

subscale. Kain et al.’s organisation subscale consisted of five items and achieved a

Cronbach’s alpha of .73. Three of Kain et al.’s items were retained in the

organisational culture scale in the current study; however, two of the items did not fit

within the construct context, as defined in the framework guiding this study (Kitson

et al., 1998) and were instead included in the scale of ‘Patient and family

preferences’ within the key element of ‘Evidence’.

Four additional items were developed and included, based on the findings from phase

one of this study and the literature review. The internal consistency of this scale was

.58. Although deletion of the item In my Unit, end-of-life care is mostly allocated to

junior nurses from this scale increased the alpha to .66, this item was retained in the

survey instrument for phase three of this study to identify the relevance of this item

in the national critical care population.

The correlation coefficient obtained from test-retest reliability was slightly lower

than the generally accepted satisfactory level of .70; however, it was statistically

significant (r = .61, p < .05).

Table 7-17: Items in the organisational culture scale

Item

The medical staff support EOLC for dying patient in my unit

In my unit, the team expresses its opinions, values and beliefs about providing care

to dying patients

All members of the healthcare team in my Unit agree with and support EOLC when

100

it is implemented

After a decision is made to withdraw life-sustaining treatment, less time is spent with

the patient by the critical care medical team

After a decision is made to withdraw life-sustaining treatment, nurses are left to

manage the care of the patient

In my unit, EOLC is mostly allocated to junior nurses

In my unit, nurses view the provision of EOLC as an important component of critical

care nursing

Note. EOLC = end-of-life care.

Scale: Resources

Resources refer to the availability of factors, both physical and human, within the

critical care unit to support the provision of end-of-life care as perceived by the

individual nurse (survey respondent). This scale proposed a priori comprised of eight

items (see Table 7–18). The first, fourth and fifth items in this scale were from Kain

et al.’s (2009) instrument. The remaining items were developed from the findings of

phase one of this study and the literature review. The item Lack of staff results in no

‘down time’ between the death of a patient and the admission of the next patient was

reverse coded prior to data analysis.

The Cronbach’s alpha of this scale in pilot testing was .60 and achieved a large

correlation coefficient on test-retest reliability that was statistically significant (r =

.80, p < .01). Based on these results, all items were retained in the survey instrument

for use in phase three of this study.

Table 7-18: Items in the resources scale

Item

The physical environment of my unit is ideal for providing EOLC

Private rooms are available to care for the patient at the EOL

My unit is adequately equipped (chairs/recliners/waiting room etc.) to support the

101

comfort needs of the family during EOLC

My unit is adequately staffed for providing the needs of dying patients and their

families

When a patient dies in my Unit, nurses have sufficient time to spend with the family

There are policies/guidelines to assist in the delivery of EOLC in my Unit

Lack of staff results in no ‘down time’ between the death of a patient and the

admission of the next patient

When a patient dies in my Unit, families have sufficient time to spend with the

patient

Note. EOL = end of life; EOLC = end-of-life care.

Evidence

The key element of evidence was represented by three scales in the survey

instrument developed for the current study. The scales patient and family

preferences, nursing education and experience and nurse practice knowledge are

described in the following section and the results of internal consistency and test-

retest reliability for each scale are presented.

Scale: Patient and family preferences

Patient and family preferences is defined as the involvement of families in care

planning and decision making in relation to end-of-life care of patients (their

relatives) in the critical care unit as perceived by the individual nurse (survey

respondent). The first two items in this scale were drawn from Kain et al.’s (2009)

instrument, while the remaining items, presented in Table 7–19, were empirically

derived from the findings of phase one of this study and the review of the literature.

The Cronbach’s alpha of this scale in pilot testing was .79, indicating satisfactory

internal consistency. Test-retest reliability of this scale was also satisfactory with a

large correlation coefficient that was statistically significant obtained in pilot testing

(r = .68, p < .05). No changes to this scale were considered necessary prior to its

inclusion in the final survey instrument for use in phase three of this study.

102

Table 7-19: Items in the patient and family preferences scale

Item

In my unit, families are involved in decisions about the dying patient

In my unit, when a diagnosis with a likely poor outcome is made, families are

informed of EOLC options

In my unit, withdrawal of treatment is delayed whilst waiting for relatives of the

patient to arrive

In my unit, the patient’s preferences (or their likely preference according to family)

for EOLC are considered

In my unit, families are given adequate time to consider decisions to

withhold/withdraw treatment for the patient (their relative)

In my unit, family meetings with the health care team are held to resolve differing

views about a patient’s treatment

Note. EOLC = end-of-life care.

Scale: Nursing education and experience

The items in this scale refer to the individual nurses’ participation in educational

activities incorporating end-of-life care content and their knowledge of end-of-life

care that may be gained from either educational opportunities or personal

experiences of death and dying (see Table 7–20). In this scale, the items I don’t know

what to say to the family of a patient at the end of life and I have limited knowledge

of symptom management for patient’s at the end-of-life were reverse coded.

Following analysis of the pilot test participants’ responses, the internal consistency

of this scale was deemed satisfactory with a Cronbach’s alpha of .75 and a large

correlation coefficient that was statistically significant was obtained through test-

retest reliability (r = .77, p < .01). Once again, all items were retained, unchanged for

inclusion in the final version of the survey instrument.

103

Table 7-20: Items in the nursing education and experience scale

Item

I have received in-service education that assists me to support and communicate with

families of dying patients

I have received in-service education that assists me to provide care and manage

patient symptoms at the end-of-life

My undergraduate nursing education included EOLC content

My postgraduate nursing education included EOLC content relevant to the critical

care context

I draw on evidence from the literature to provide EOLC to patients and their families

I feel adequately prepared to care for patients at the EOL

I don’t know what to say to the family of a patient at the EOL

I have limited knowledge of symptom management for patient’s at the EOL

I am often exposed to death in the critical care environment

I have learnt from other nurses how to provide EOLC

Note. EOL = end of life; EOLC = end-of-life care.

Scale: Nurse practice knowledge

The items in this scale were designed to measure aspects of nurse knowledge in

relation to end-of-life care (see Table 7–21). The item The provision of end-of-life

care requires emotional detachment was taken from the Palliative Care Quiz for

Nursing (PCQN; Ross, McDonald, & McGuinness, 1996) and required reverse

coding prior to testing the internal consistency and reliability of this scale. The

remaining items were empirically derived from the literature and the findings of

phase one of this study.

The internal consistency of this scale was found to be low, with a Cronbach’s alpha

of .41 obtained. This finding indicates that the items in this scale may be measuring

discrete concepts rather than a single construct. A large correlation coefficient that

was statistically significant (r = .86, p < .01) was achieved on test-retest reliability,

indicating, as expected, that participant responses to these items remain stable across

104

time. Although the internal consistency of the scale was not satisfactory, each of the

items was considered to measure an aspect of end-of-life care of interest to this

study. Therefore, all items were retained in the final version of the survey instrument

distributed in phase three of this study.

Table 7-21: Items in the nurse practice knowledge scale

Item

Analgesia should be titrated to keep the patient comfortable even if this hastens death

Pressure area care should continue to be provided for a patient at the EOL

EOLC should include care of the patient’s family

Abrupt withdrawal of mechanical ventilation may contribute to dyspnoea

When withdrawing mechanical ventilation at the EOL, ventilator weaning (such as

reductions in FiO2 and PEEP) should occur prior to extubating the patient or placing

the patient on a T-piece

Intravenous fluids should be provided to maintain hydration for the patient at the

EOL

The provision of EOLC requires emotional detachment

It is OK to cry when a patient dies

Note. EOL = end of life; EOLC = end-of-life care.

Facilitation

The key element of facilitation was represented by three scales within the survey

instrument: the emotional support for nurses, nurses’ end-of-life care values and

beliefs and coordination of care delivery scales. The items representing these scales

and the results of internal consistency and test-retest reliability for each of these

scales are presented below.

105

Scale: Emotional support for nurses

The items in the emotional support for nurses scale (see Table 7–22) aim to measure

the individual nurses’ (survey respondents) perceptions of the support and assistance

they receive from their colleagues and their workplace when providing end-of-life

care. The first item in this scale was taken from Kain et al.’s (2009) instrument. The

remaining items were derived from the findings of phase one of this study and the

literature review. The second item, In my unit, nurses are expected to cope with

death and dying, was reverse coded prior to undertaking statistical analysis. Analysis

of the pilot test participant responses identified that this scale had satisfactory

internal consistency (α = .86) and achieved a large correlation coefficient that was

statistically significant on test-retest reliability (r = .92, p < .01). Therefore, the items

in this scale were included in the final version of the survey instrument.

Table 7-22: Items in the emotional support for nurses scale

Item

When a patient dies in my Unit, counselling is available if I need it

In my unit, nurses are expected to cope with death and dying

I feel supported when caring for a patient at the EOL

After caring for a patient at the end-of-life, my colleagues will ask me if I am OK

My supervisors/managers ensure staff caring for patients at the EOL are supported

My colleagues appreciate the stressors associated with caring for patients at the EOL

My supervisors/managers provide guidance that assists me to provide EOLC

Note. EOL = end of life; EOLC = end-of-life care.

Scale: Nurses’ end-of-life care values and beliefs

This scale contained items designed to measure the personal views of individual

nurses (survey respondents) in relation to end-of-life care. The items comprising this

scale are presented in Table 7–23. The fifth item in this scale was drawn from Kain

et al.’s (2009) instrument. The remaining items were generated from the literature

106

review and the findings of phase one of this study. The item I feel a sense of personal

failure when a patient dies was reverse coded.

The internal consistency of this scale was low (α = .39). Although deleting selected

items from this scale moderately improved the internal consistency, all items were

retained for further testing in phase three of this study, as it was considered that,

individually, the items may provide important descriptive information even though

they did not hold together as a scale. In an attempt to improve the internal

consistency of this scale in phase three, three additional items were added to the

survey: After a decision is made to withdraw treatment, I spend less time with the

patient, Patients at the end-of-life require little nursing care and End-of-life care

requires advanced nursing skills. In the pilot testing of this scale, a large correlation

coefficient that was statistically significant (r = .85, p < .01) was obtained on test-

retest reliability.

Table 7-23: Items in the nurses’ end-of-life care values and beliefs scale

Item

I feel a sense of personal failure when a patient dies

It is a privilege to care for a patient at the EOL and their family

I can influence the patient and family’s experience of EOLC

My own personal experiences of death have influenced the care I provide to patients

at the EOL

EOLC is as important as curative care in the critical care environment

After a decision is made to withdraw treatment, I spend less time with the patient

Patients at the EOL require little nursing care

EOLC requires advanced nursing skills

Note. EOL = end of life; EOLC = end-of-life care.

107

Scale: Coordination of care delivery

In this study, coordination of care delivery is defined as the strategies undertaken

within the unit to smooth transitions between individual care providers (both nurses

and other health care providers) for the family and coordinate care for the patient at

the end of life as perceived by the individual nurse (survey respondent). All four

items in this scale, presented in Table 7–24, were included in the final version of the

survey instrument. The Cronbach’s alpha of this scale was .67 and a large correlation

coefficient that was statistically significant (r = .76, p < .01) was achieved between

the test and retest scores of this scale.

Table 7-24: Items in the coordination of care delivery scale

Item

In my unit, nurses that know the patient and family are allocated to care for the

patient to maintain continuity of care

In my unit, a plan of care for the patient is clearly documented by the medical team

after a decision to withhold or withdraw treatment is made

In my unit, the intensive care team and the patient’s primary medical or surgical team

agree on a plan of care

In my unit, patients remain in ICU for EOLC following a decision to withhold or

withdraw treatment

Note. EOLC = end-of-life care; ICU = intensive care unit.

Overall, each of the a priori scales in the Factors influencing the provision of end-of-

life care section of the survey instrument achieved a satisfactory level of internal

consistency and test-retest reliability. A summary of the results of internal

consistency and test-retest reliability of the a priori scales in this section of the

survey instrument is provided in Table 7–25.

108

Table 7-25: Internal consistency and test-retest reliability of the a priori scales

in the Factors influencing the provision of end-of-life care section of the survey

instrument

Scale α r

Organisational culture .58 .61*

Resources .60 .80**

Patient and family preferences .79 .68*

Nursing education and experience .75 .77**

Nurse practice knowledge .41 .86**

Emotional support for nurses .86 .92**

Nurses’ end-of-life care values and beliefs .39 .85**

Coordination of care delivery .67 .76**

*p < .05, two-tailed. **p < .01, two-tailed.

Participants in the pilot test indicated that the readability of the survey instrument

was satisfactory, although it was acknowledged that the survey was quite long. All

items in the survey instrument that referred to intensive care were modified to read

critical rather than intensive, reflecting the broader practice settings in which critical

care nurses may work. Pilot test participants commented positively about the need

for research into end-of-life care in critical care settings and expressed an interest in

the findings of the research.

7.4 Summary

In summary, the purpose of phase two of this study was to develop and test a survey

instrument. A panel of six experts reviewed the content of the survey instrument and

provided relevance ratings permitting the calculation of both individual item and

survey CVIs. An initial pilot test indicated that several of the scales proposed a priori

were not robust measures of the constructs of interest. Significant modification to the

instrument was undertaken, and it was determined that a second pilot test was

necessary to obtain preliminary evidence of the reliability and validity of the survey

109

instrument. Pilot testing of the survey instrument revealed satisfactory internal

consistency between the items in the scales proposed; however, in many cases,

correlation coefficients from test-retest reliability were low. Overall, the survey

instrument was considered to provide a satisfactory instrument to achieve its purpose

and only minor modifications were undertaken prior to its use in phase three of this

study.

110

8 PHASE THREE METHODS

8.1 Introduction

Phase three of this mixed methods study employed a quantitative approach to

identify the end-of-life care practices of critical care nurses and the factors

influencing these practices. This phase utilised the survey instrument developed in

phase two, based on the findings of phase one of this study and a review of the

literature, to address the key research questions of this study, which are:

1. What are the most frequently reported end-of-life care practices of Australian

critical care nurses?

2. What are the factors influencing the provision of end-of-life care as perceived

by Australian critical care nurses?

3. What factors are associated with critical care nurses’ end-of-life care

practices?

This chapter will describe the setting and sample, procedure, and data management

and analysis used in phase three of this study.

8.2 The setting and sample

The target population in phase three of this study is Australian registered nurses with

experience in a critical care setting. In the Intensive Care Resources and Activity:

Australia and New Zealand 2007/2008 report, the term critical care unit encompasses

general and specialty ICU, combined ICU/CCU, paediatric ICU, HDU, and a

combination of any of the above (Drennan et al., 2010). This report identifies that

there were 8,975 registered nurses working in a critical care unit in Australia. Of

these, 4,605 (56.1%) held a critical care nursing qualification (Drennan et al., 2010).

It should be noted that these figures were obtained from staffing figures supplied by

critical care units that were eligible and responded to the survey (n = 153; response

rate 95%). The survey did not include data from smaller units that did not meet the

inclusion criteria of invasive ventilation of 20 or more patients per annum over three

consecutive survey periods (Drennan et al., 2010). These figures also did not include

111

registered nurses who may be working in critical care units as a result of employment

by nursing agencies or casual pools, nor does it include nurses employed in other

areas that may be considered critical care, including emergency departments and

retrieval units.

The accessible population is a subset of the target population that is available to

invite to participate in the study (Polit & Beck, 2010). The accessible population in

phase three of this study was the membership of the ACCCN. Membership of

ACCCN is open to anyone with an interest in critical care nursing. The ACCCN is a

professional organisation that represents critical care nurses on a national level and

provides its members with education and information pertinent to critical care.

Approximately 2,300 registered nurses are members of ACCCN. Members of

ACCCN are employed in clinical, research and/or academic positions.

A non-probability sampling technique was chosen to recruit the total population of

ACCCN members who had agreed to be contacted for research purposes to

participate in phase three of this study. It is recognised that this sampling strategy is

subject to a potential selection bias. The sample may not accurately represent the

total population of critical care nurses in Australia, as it will not include critical care

nurses who are not members of ACCCN, and this may potentially limit the

generalisability of findings. Members of ACCCN are a unique group; for example,

membership suggests they are more likely to be interested in professional issues and

education than non-members. The sample will represent approximately 25% of the

target population, based on the figures provided in the Intensive Care Resources and

Activity: Australia and New Zealand 2007/2008 report (Drennan et al., 2010).

8.3 Procedure

During May 2011, a convenience sampling technique was utilised to invite all

members of the ACCCN who had agreed to be contacted for research purposes to

participate in phase three of this study. An email inviting members of ACCCN to

participate in this study was distributed by an ACCCN administrative assistant

112

through their member database (see Appendix 13.10). An information sheet outlining

the requirements of participation, expected benefits and risks was attached to the

email (Appendix 13.11). Participants completed and submitted the survey online.

The survey takes approximately 20 minutes to complete and a copy of the survey

content is provided in Appendix 13.12. A reminder email was circulated to potential

participants three weeks following intial contact. The survey was available online for

six weeks.

The survey software, Key Survey version 7.3 (WorldAPP Key Survey, 2011)

provided the online platform for the survey. To permit the survey to be distributed by

ACCCN, a link to the online survey in the form of a master URL was used. Use of a

master URL required an ‘allow multiple responses’ rule to be set to permit the survey

to be taken multiple times on the same computer. The rule allowed participants who

are at work in a clinical setting to use the same computer terminal to complete the

survey.

8.4 Data management and analysis

As data were collected, they were periodically checked to identify any response

errors. The data were exported from Key Survey into PASW software version 18 for

statistical analysis. A log of decisions regarding data cleaning was maintained. The

data set was reviewed to identify any extreme values or outliers, and missing values.

Where statistical tests do not permit missing data, missing values were managed by

listwise deletion (Hair et al., 2006). Items worded in the opposite direction (see Table

8–1) were reverse scored prior to undertaking factor analysis.

113

Table 8-1: Items reverse scored prior to factor analysis

Item

Lack of staff results in no ‘down time’ between the death of a patient and the

admission of the next patient

I don’t know what to say to the family of a patient at the EOL

I have limited knowledge of symptom management for patient’s at the EOL

The provision of EOLC requires emotional detachment

In my Unit, nurses are expected to cope with death and dying

I feel a sense of personal failure when a patient dies

Patients at the EOL require little nursing care

The thought of death seldom enters my mind

It doesn’t make me nervous when people talk about death

I am not at all afraid to die

I am not particularly afraid of getting cancer

The thought of death never bothers me

I feel that the future holds nothing for me to fear

Note. EOL = end of life; EOLC = end-of-life care.

The analytical approach used in phase three of this study included the use of:

(1) descriptive statistics to summarise the data,

(2) factor analysis for data reduction and the identification of the underlying

constructs of the survey instrument,

(3) bivariate statistical tests to identify relationships between the independent and

dependent variables, and

(4) modelling of the data using a general linear model to explain the relationship

between each outcome variable and multiple explanatory variables.

8.4.1 Descriptive statistics

Descriptive statistics were used to summarise the data obtained from the survey

instrument. The continuous variables in the data set were the summated scales

114

(generated from the findings of factor analysis) for each of the outcome and explanatory

variables and participant age. Data from observations of continuous variables were

summarised by calculating the mean, median, range and standard deviation. These

statistics provided a summary of the measure of central tendency and spread of the data

for each variable and assisted in determining normality when an assumption of normal

distribution needed to be met. A histogram of the values of the variable was provided

when a graphical representation of the data was needed, and this graph was also used to

assist in determining normality of the distribution.

Data from categorical variables were summarised by calculating the frequency and

percentage for each category and were presented in a frequency table or bar chart.

Categorical variables in the data set include: gender, location, highest qualification in

nursing, formal education in critical care, years of experience as a nurse, years of

experience in critical care, current workplace, location of workplace, type of work, time

since the provision of direct patient care and time since the provision of end-of-life care.

In addition to the descriptive statistics of the summated scales, the frequency and

percentage of participant responses for each item in the Critical care nurses’ end-of-life

care practices and the Factors influencing the provision of end-of-life care sections of

the survey instrument were calculated and presented in a frequency table. The results of

these analyses are used to answer the research question: What are the most frequently

reported end-of-life care practices of Australian critical care nurses? In addition, the level

of agreement for items in the Factors influencing the provision of end-of-life care section

provides some evidence relevant to the research question: What are the factors

influencing the provision of end-of-life care as perceived by Australian critical care

nurses?

8.4.2 Exploratory factor analysis

EFA was undertaken on the data obtained from the Critical care nurses’ end-of-life care

practices and the Factors influencing the provision of end-of-life care sections of the

survey instrument for the dual purposes of data reduction and identifying the latent

115

constructs within each section. Although principal components analysis is widely used in

the literature and is the default EFA extraction method in many statistical software

packages, it is a data reduction technique, which only identifies the components of the

data and does not differentiate between common, unique and error variance (Bandalos &

Finney, 2010; Costello & Osborne, 2005). Common factor analysis methods such as

principal axis factoring (PAF) only analyse common (shared) variance between the

variables (Costello & Osborne, 2005; Hair et al., 2006), enabling latent constructs in the

data to be identified. The generation of summated scales from these latent constructs

creates a more parsimonious set of variables for use in bivariate analysis and multivariate

modelling of the data (Hair et al., 2006).

EFA as opposed to confirmatory factor analysis (CFA) was selected as the most

appropriate factor analytical technique, as the items in the survey instrument were

specifically developed for this study and limited previous research of the end-of-life care

practices of critical care nurses and the factors influencing the provision of end-of-life

care has been undertaken. Although a priori scales were specified during development of

the survey instrument, Bandalos and Finney (2010) note that:

A common situation in which one must choose between EFA and CFA is the

investigation of items that have been written to measure a construct hypothesized

to have several dimensions. Researchers often claim a priori knowledge of the

underlying structure based on the fact that the items have been written to

measure specific aspects of the construct. However, in our experience, items are

rarely aware of the scale for which they have been written and often fail to

behave as they should. (p. 96)

The results of EFA are dependent on the items included in the analysis. Inadequate

inclusion of items may result in a latent construct not being identified (Bandalos &

Finney, 2010), while the inclusion of items not conceptually relevant may result in the

identification of latent constructs not representative of the domain of interest (Fabrigar,

Wegener, MacCallum, & Strahan, 1999). A further limitation of EFA is the

subjectiveness involved in arriving at a factor solution (Bandalos & Finney, 2010; Hair

et al., 2006; Kim, 2008). Although some guidelines are available, the researcher

116

ultimately selects the methods used in factor analysis and interprets the theoretical and

conceptual meaningfulness of the solution (Bandalos & Finney, 2010; Hair et al., 2006;

Kim, 2008). The factorability of the data set, method of extraction, type of rotation and

number of factors to retain are key decisions to be made when undertaking EFA. The

remainder of this section will report the EFA methods used in this study.

Prior to commencing EFA, the factorability of the data set and the adequacy of the

sample size were ascertained. Correlation matrices of data from the Critical care nurses’

end-of-life care practices and the Factors influencing the provision of end-of-life care

sections of the survey instrument were reviewed to ensure correlations greater than .30

between variables, indicating factorability of the data (Hair et al., 2006). It is noted that

an inadequate sample size can result in a factor solution that fits the data but is not

generalisable to the wider population (Hair et al., 2006). Generally, a sample of five to

10 participants per item is considered adequate (Hair et al., 2006). Statistically, the

adequacy of the data set for factor analysis was also determined by reviewing the

measure of sampling adequacy (MSA) for each item and ensuring that it is greater than

.50 and obtaining a significant result on Bartlett’s test, indicating correlation among the

variables (Hair et al., 2006). Missing values in the data set were managed by listwise

deletion, whereby the data for the entire respondent were deleted and not included in the

factor analysis (Hair et al., 2006).

As previously identified, PAF is a common factor analysis method that identifies the

latent constructs in the data through analysis of shared variance. Therefore, PAF was

selected as the extraction method used in this research study. Oblique rotation methods

permit correlation of the factors by enabling the axis to fit the data and not restricting

them to be at 90 degrees (Hair et al., 2006). Oblique (direct oblimin) rotation was used,

as correlation of the factors in the data set was anticipated. Oblique rotation produces an

orthogonal result when no correlation exists (Bandalos & Finney, 2010; Costello &

Osborne, 2005).

Use of multiple methods to determine the number of factors to retain has been

recommended (Bandalos & Finney, 2010; Fabrigar et al., 1999; Hair et al., 2006). In this

study, the number of factors to be retained was determined using Kaiser’s criterion, a

scree plot and the number of factors proposed a priori (Costello & Osborne, 2005;

Fabrigar et al., 1999). Kaiser’s criterion retains factors with an eigenvalue greater than

117

1.0 and has been criticised for overextraction of factors (Fabrigar et al., 1999). The scree

plot is a graph of factors against their corresponding eigenvalues. The number of factors

retained according to the scree plot is determined by the point at which the curve begins

to plateau (Costello & Osborne, 2005; Hair et al., 2006). Interpretation of the scree plot

has been criticised for its subjectiveness (Fabrigar et al., 1999). Kaiser’s criterion, the

scree plot and the a priori factor structure, as represented by the scales identified during

the development of the survey instrument, provided a basis for selecting the number of

factors to retain. Comparison of multiple factor solutions for the best fit of the data in

regard to number of items loading on each factor, minimal cross loadings and theoretical

or conceptual meaningfulness was ultimately used to select the final factor solution

(Bandalos & Finney, 2010; Costello & Osborne, 2005; Fabrigar et al., 1999; Floyd &

Widaman, 1995).

When interpreting the factor solution, an item with a pattern coefficient greater than .30

was considered to contribute to the factor. A pattern coefficient of .30 equates to

approximately 10% shared variance between the variables (items) in the factor

(Bandalos & Finney, 2010; Costello & Osborne, 2005; Skerman, Yates, & Battistutta,

2009). The factor correlation matrix was reviewed to identify correlation between factors

(Costello & Osborne, 2005).

The internal consistency of scales (factors) identified by factor analysis was further

assessed by calculating Cronbach’s alpha coefficient for the survey items in each scale.

A reliability coefficient greater than .70 is considered satisfactory (Hair et al., 2006).

Summated scales were calculated by adding together the scores for items on a scale.

Descriptive statistics, including the mean and range, for each summated scale were

calculated. The summated scales developed from the Factors influencing the provision

of end-of-life care section of the survey instrument, together with the demographic

variables, formed the explanatory variables in subsequent bivariate and multivariable

modelling of the data. The summated scales developed from items in the Critical care

nurses’ end-of-life care practices section of the survey instrument were used as the

outcome variables. The findings from the EFA are used to answer the research question:

What are the factors influencing the provision of end-of-life care as perceived by

Australian critical care nurses?

118

8.4.3 Bivariate statistical analysis

Bivariate analyses were undertaken to identify possible relationships between each

outcome variable (critical care nurses’ end-of-life care practice areas derived from the

summated scales identified during factor analysis) and each explanatory variable

(demographic variables and summated scales representing select factors influencing the

provision of end-of-life care). The findings of the bivariate analyses informed the

selection of variables for inclusion in subsequent multivariable modelling.

The normality of each outcome variable was ascertained by reviewing the descriptive

statistics, including the mean, median, range and skewness and kurtosis coefficients,

combined with a visual inspection of the histogram for each outcome variable (Hair et

al., 2006; Tabachnick & Fidell, 2007). The normality assumption was also tested for

each outcome variable by calculating the Shapiro-Wilk statistic. A significant result

on this test indicates that the data are not normally distributed; however, this test may

produce significant results with small departures in normality in large samples (Hair

et al., 2006).

As a result of violation of the normality assumption by each of the outcome variables,

non-parametric tests were undertaken in all bivariate analyses. When the outcome and

explanatory variables were measured as continuous variables (for example, age), a

scatter plot of the relationship between the variables was graphed and a Spearman’s rank

correlation coefficient calculated. When testing the relationship between an outcome

variable and categorical variable with two groups (gender, location, postgraduate

qualification and formal education in critical care variables), a Mann-Whitney test was

undertaken. When testing the relationship between an outcome variable and categorical

variable with three or more groups (years of nursing experience, years of critical care

experience, workplace, primary nursing role, time since direct patient care and time since

end-of-life care variables), a Kruskal-Wallis ANOVA was used. Multiple pairwise

comparisons were undertaken when a significant difference was identified to determine

which of the groups were different.

Side-by-side box plots were used to provide a graphical summary of the relationship

between the continuous outcome variable and a categorical explanatory variable. In each

119

of the bivariate statistical tests undertaken, a p-value of less than .05 was selected as the

criterion for reporting statistical significance. An adjusted level of statistical significance

was calculated and reported for tests involving multiple pairwise comparisons to

minimise Type I error (Tabachnick & Fidell, 2007).

8.4.4 Multivariable statistical analysis

Multivariable regression modelling is undertaken for the purpose of prediction and/or

explanation of the relationships between each outcome (dependent) and the explanatory

(independent) variables (Hair et al., 2006; Tabachnick & Fidell, 2007). Prediction is

achieved by identifying the explanatory variables that contribute to changes in the

outcome variable. Explanation occurs when the most important predictors (explanatory

variables) are identified, based on the strength, direction of relationship (positive or

negative) and statistical significance (Hair et al., 2006). In this phase of the study, the

primary purpose of undertaking multiple regression modelling was prediction to answer

the research question: What factors are associated with critical care nurses’ end-of-life

care practices?

This section will describe the model-building approach used in this study, including

estimation of the model, model fit and the assumptions tested in evaluating the model.

Using the univariate general linear model procedure in PASW software version 18, a

model was built for each of the six outcome variables. As previously identified, the

outcome variables were the six summated scales of critical care nurses’ end-of-life care

practices. The potential explanatory variables of these practices were the summated

scales of factors influencing the provision of end-of-life care and the demographic

variables.

Any explanatory (independent) variable that had a significance level of p < .2 in

bivariate analysis with the outcome (dependent) variable was considered a variable of

importance and included in the initial model (Tabachnick & Fidell, 2007). Using a

backward elimination approach, non-significant explanatory variables were removed

from the model one at a time until all explanatory variables remaining in the model were

120

significant (Hair et al., 2006; Tabachnick & Fidell, 2007). A p-value of less than .05 was

selected as the criterion for reporting statistical significance. This p-value was used to

report overall model significance and the significance of explanatory variables. Post hoc

pairwise comparison with Bonferroni correction was undertaken when a significant

difference was identified for a categorical explanatory variable to determine which of the

groups were different.

The explanatory power of the general linear model for the data was reported using the

adjusted R-squared. The R-squared value increases as the number of variables in the

model increases, while the adjusted R-squared adjusts for the number of variables in the

model, providing a more accurate measure of explanatory power (Hair et al., 2006). The

explanatory power for individual variables was reported using partial eta squared values.

Partial eta squared explains the variance unique to each explanatory variable in the

model (Hair et al., 2006).

The assumptions to be met in the general linear model are that of independence of

observations, normality, linearity and equality of variance. The assumption of

independence of observations is met in this cross-sectional study design where each

participant is observed on only one occasion. However, the possibility of participants

being alike as an observational unit, for example being employed in the same workplace

and having access to the same resources, should be considered during interpretation of

the findings of this study.

Graphical representations of the residuals for each model were reviewed to check the

assumptions of normality, linearity and equality of variance. Identification of normal

distribution of the residuals, around zero with a constant variance, indicated that these

assumptions were met (Hair et al., 2006; Tabachnick & Fidell, 2007). In addition, the

assumption of equality of variance was met when a non-significant result on Levene’s

test of equality of variance was obtained (Hair et al., 2006).

Finally, multicollinearity occurs when explanatory variables are highly correlated (Hair

et al., 2006). Correlation matrices of explanatory variables were reviewed to identify if

any of the explanatory variables were too highly correlated (r > .9) (Tabachnick &

Fidell, 2007).

121

8.5 Summary

This chapter has described the methods used to conduct a national online survey of a

convenience sample of critical care nurses. An analytical approach involving

descriptive statistics, EFA and multivariable analysis was described and

interpretation of the results from these analyses will be used to answer the three key

research questions proposed in the current study. The results of phase three of the

study are presented in the next chapter.

122

9 PHASE THREE RESULTS

9.1 Introduction

The survey instrument was circulated to and available for completion by potential

participants in phase three of this study during May and June 2011. This chapter will

first outline participant recruitment and response to the survey instrument. The

demographic characteristics of the sample and the descriptive statistics of individual

items will then be presented. The results of EFA on the Critical care nurses’ end-of-life

care practices and the Factors influencing the provision of end-of-life care sections of

the survey instrument will be outlined. Bivariate and multivariable analyses will then be

presented, with particular attention to the six practice models and their explanatory

variables identified using the univariate general linear model procedure.

9.2 Phase three participants

Approximately 2,300 critical care nurses are members of ACCCN. When the survey was

initially distributed in May 2011, 1,553 of these members had agreed to be contacted for

research purposes and were invited to participate. Due to an increase in membership, the

invitation to participate reminder notice was sent to 1,580 members (L. Herson, personal

communication, June 21, 2011).

In total, 404 members of ACCCN responded to the survey invitation (response rate

25.5%). Eight persons responded via return email declining participation, stating that

they were not currently working in clinical practice and could not complete the survey.

As the survey was distributed through the ACCCN membership database, the researcher

was unable to ascertain if any emails were undelivered.

The survey instrument was completed by 396 critical care nurses. The survey software

did not capture those persons who partially completed the survey, so the response rate

only reflects those who completed the survey. Responses from three participants were

deleted as they identified that they had never provided end-of-life care. A duplicate

response was noted in one instance, and the duplicate response was deleted. Therefore,

123

the final response rate was 25% (392/1568). A flow chart representing participant

recruitment and response is provided in Figure 9–1.

Figure 9-1: Flow chart of participants in phase three

Note. EOL = end of life.

9.3 Demographic characteristics

The mean age of critical care nurse participants in phase three of this study was 43.1

years (SD = 9.2) and 13.5% of participants were male. Years of experience in nursing

ranged from two to 43 years (M = 20.9; SD = 10.1), with 77.1% of participants

124

having more than 10 years of experience in nursing. Years of experience working as

a nurse in a critical care unit ranged from 0 to 39 years (M = 14.4; SD = 8.2) and

64% reported more than 10 years of experience in the critical care setting. In addition

to extensive years of experience, the majority of participants in phase three of this

study had postgraduate qualifications in nursing (82.7%) and had completed or were

currently undertaking formal education in critical care (90.3%). Table 9–1 outlines

the age, gender, nursing experience, qualifications and education of phase three

participants.

Table 9-1: Demographic characteristics of phase three participants (n = 392)

Characteristic %

Age

<25 2.0

25–29 8.7

30–34 10.5

35–39 12.8

40–44 16.1

45–49 23.2

50–54 15.6

55–59 6.9

60+ 2.0

Gender

Female 86.5

Male 13.5

Nursing Experience

0–5 7.7

6–10 15.2

11–15 8.2

16–20 15.5

20+ 53.4

Critical Care Experience

0–5 18.1

125

6–10 17.9

11–15 21.8

16–20 18.1

20+ 24.1

Highest Qualification in Nursing

Postgraduate 82.7

Hospital certificate/

Undergraduate degree

10.7

Formal Education in Critical Care

Yes 90.3

No 9.7

The majority of participants in phase three of this study were employed in a major

city (71.4%). The remainder reported employment in a regional area (27.0%) or

remote area (1.5%). All jurisdictions were represented by participants in phase three

of this study (see Table 9–2). The current workplace of participants reflected the

variety of settings in which critical care nurses are employed. The majority of

participants were currently employed in a general ICU (49.2%) or an integrated

ICU/CCU/HDU unit (23.5%). Overall, 90% of phase three participants had provided

direct patient care within the last 12 months and, for 86% of participants, this care

included the provision of end-of-life care (see Table 9–2).

Table 9-2: Phase three participants’ location, workplace and participation in

work (n = 392)

Characteristic %

Location

Queensland 18.9

New South Wales 22.2

Australian Capital Territory 2.8

Victoria 34.7

Tasmania 1.5

126

South Australia 8.9

Western Australia 7.4

Northern Territory 1.8

Overseas 1.8

Current workplace

General ICU (integrated medical/surgical including ICU managed HDU

and specialty ICUs)

51.3

Integrated ICU/CCU/HDU 23.5

Paediatric ICU 7.4

Emergency Department 4.6

Other clinical 3.8

Other non-clinical 9.4

Type of work

Clinical (direct patient care) 55.4

Nursing management (team leader/supervision of clinicians) 16.1

Nursing management (administration) 7.1

Education 14.8

Research / Quality / Project work 6.6

Time since provision of direct patient care

Within the last week 74.5

Within the last month 8.9

Within the last 12 months 6.6

In the last 1–5 years 6.6

More than 5 years 3.3

Time since provision of end-of-life care

Within the last week 19.1

Within the last month 30.4

Within the last 12 months 36.5

In the last 1–5 years 10.2

More than 5 years 3.8

Note. ICU = intensive care unit; CCU = coronary care unit; HDU = high dependency

unit.

127

9.4 Descriptive statistics

Descriptive statistics were calculated for each of the items in the Critical care

nurses’ end-of-life care practices section and the Factors influencing the provision of

end-of-life care section of the survey instrument. Table 9–3 presents the mean and

standard deviation for items in the Critical care nurses’ end-of-life care practices

section that achieved the highest ranking based on their mean score, indicating the 10

end-of-life care practices most frequently undertaken as identified by critical care

nurse participants in this study. Participant responses as a percentage of the sample,

means and standard deviations for all items in the Critical care nurses’ end-of-life

care practices section of the survey instrument are provided in Appendix 13.13.

Table 9-3: The mean and standard deviation for the 10 items in the Critical care

nurses’ end-of-life care practices section most frequently undertaken based on

their mean score (n = 392)

Item M (SD)

Permit the family to visit outside of regular visiting hours 4.7 (0.5)

Place chairs around the bed for the family 4.7 (0.5)

Provide care to maintain patient hygiene 4.7 (0.5)

Introduce the nursing staff on the oncoming shift to the patient and family 4.7 (0.5)

Encourage the family to touch the patient 4.7 (0.6)

Handover the family’s preference for the patient’s end-of-life care to the

oncoming nurse 4.7 (0.6)

Draw curtains to provide privacy for the patient and family 4.6 (0.6)

Encourage the family to talk to the patient 4.6 (0.6)

Answer the family’s questions about the patient’s condition 4.6 (0.6)

Silence alarms on all equipment 4.6 (0.6)

Note. M = mean; SD = standard deviation

Table 9–4 presents the mean and standard deviation for the items in the Factors

influencing the provision of end-of-life care section of the survey instrument that

128

achieved the highest ranking based on their mean score, indicating the items that

achieved the highest level of agreement as factors influencing the provision of end-

of-life care in critical care settings. Participant responses as a percentage of the

sample, means and standard deviations for all items in the Factors influencing the

provision of end-of-life care section of the survey instrument are provided in

Appendix 13.14.

Table 9-4: The mean and standard deviation for the 10 most strongly agreed

items in the Factors influencing the provision of end-of-life care section of the

survey instrument based on their mean score (n = 392)

Item M (SD)

EOLC should include care of the patient’s family (n = 391) 4.7 (0.5)

Analgesia should be titrated to keep the patient comfortable even if this

hastens death (n = 391)

4.6 (0.6)

EOLC is as important as curative care in the critical care environment (n

= 389)

4.5 (0.7)

It is a privilege to care for a patient at the EOL and their family 4.5 (0.7)

I can influence the patient and family’s experience of EOLC (n = 391) 4.5 (0.7)

Pressure area care should continue to be provided for a patient at the EOL

(n = 391)

4.4 (0.7)

It is OK to cry when a patient dies (n = 390) 4.2 (0.8)

In my Unit, families are involved in decisions about the dying patient 4.2 (0.7)

In my Unit, withdrawal of treatment is delayed whilst waiting for

relatives of the patient to arrive

4. 2 (0.7)

In my Unit, family meetings with the health care team are held to resolve

differing views about a patient’s treatment

4.1 (0.8)

Note. M = mean; SD = standard deviation; EOL = end of life; EOLC = End-of-life

care

129

9.5 Exploratory factor analysis

Factorability of the data was verified by reviewing the correlation matrix of the data

set to ensure the data contained correlations greater than .30 (Hair et al., 2006).

Although a number of items did not correlate greater than .30, the items included in

the survey instrument were deemed conceptually important and/or had been

validated by the expert panel as relevant to the domains of interest in this study.

Therefore, all items were included in factor analysis.

The factorability of the data set was further verified by obtaining a significant result

for Bartlett’s test of sphericity for the Critical care nurses’ end-of-life care practices

section [χ2 (1275) = 10413.76, p < .001] and the Factors influencing the provision of

end-of-life care section [χ2 (1653) = 7368.51, p < .001] of the survey instrument,

indicating adequate correlations between the items.

The Critical care nurses’ end-of-life care practices section of the survey instrument

consisted of 51 items and there were no missing values in this section. The Factors

influencing the provision of end-of-life care section of the survey instrument

consisted of 58 items and 9.4% of participants had a missing value for at least one

item. Despite the missing values in this section of the survey instrument, the ratio of

between five and 10 participants per item recommended in the literature was

achieved (Hair et al., 2006). The adequacy of the sample size for factor analysis was

also assessed by calculating the Kaiser-Meyer-Oklin (KMO) MSA. For the Critical

care nurses’ end-of-life care practices section, the KMO MSA was .91. For the

Factors influencing the provision of end-of-life care section of the survey instrument,

the KMO MSA was .81. The KMO statistic for all individual items was above .50.

Thus, the sample size was considered adequate for factor analysis.

9.5.1 Exploratory factor analysis: Critical care nurses’ end-of-life care practices

Factor analysis using PAF and oblique rotation (direct oblimin) was undertaken on

the Critical care nurses’ end-of-life care practices section of the survey instrument

130

using PASW software version 18. Kaiser’s criterion suggested a solution of 12

factors. The a priori factor structure suggested seven factors, while interpretation of

the scree plot suggested a solution of five or seven factors (see Figure 9–2).

Therefore, comparison of factor solutions using five, six, seven, eight and 12 factors

was undertaken.

Figure 9-2: Scree plot of eigen values for factors in the Critical care nurses’ end-

of-life care practices section of the survey instrument

The seven factor solution was deemed to have the best fit of the data in terms of

theoretical and conceptual meaningfulness. However, factor seven was not accepted

as a scale as it only contained two items. The two items that load together on factor

seven seem conceptually similar and perhaps measure a latent construct that is not

adequately covered by the items in the current survey instrument.

The pattern coefficient (factor loading) and communality (h2) for each item in the

Critical care nurses’ end-of-life care practices section of the survey instrument is

provided in Table 9–5. In this solution, 47 out of the 51 items had a pattern

coefficient greater than .30, the level required for an item to be considered to

131

contribute to a factor. The structure matrix for the Critical care nurses’ end-of-life

care practices section of the survey instrument is provided in Appendix 13.15. In the

seven factor solution, there were 162 (12.0%) non-redundant residuals with absolute

values greater than 0.05. Review of the factor correlation matrix (see Table 9–6)

revealed moderate correlations between factor one and a number of the other factors.

Moderate correlations between factor six and a number of the other factors were also

identified.

132

Table 9-5: Factor loadings for exploratory factor analysis with oblique rotation of Critical care nurses’ end-of-life care practices section of

the survey

Items

Factor loadings

h2 1 2 3 4 5 6 7

Explain all interventions undertaken in the care of the patient .72 –.03 –.01 .04 .05 .01 –.12 .60

Explain to the family what is happening to the patient .70 –.07 .13 .09 .06 –.07 –.01 .57

Fully inform the family about all aspects of the plan of care .64 –.06 –.02 .12 –.01 .17 –.06 .60

Answer the family’s questions about the patient’s condition .58 .11 –.09 .04 .08 .10 –.08 .51

Provide information to the family about the dying process .51 –.05 .14 .22 –.03 –.00 .12 .41

Introduce the nursing staff on the oncoming shift to the patient and family .42 .06 .12 .01 –.03 .25 –.28 .59

Liaise with other members of the health care team to meet identified needs of the patient and family .41 .03 .08 .16 .12 .17 .06 .46

Handover the family’s preferences for the patient’s EOLC to the oncoming nurse .33 .14 .04 .24 –.09 .22 –.23 .54

Ask the critical care team to document the plan of care/treatment orders .30 .02 –.08 .19 .04 .23 –.04 .32

Attend family meetings with the critical care team .30 .22 –.08 .07 –.07 .14 .06 .23

Adjust (dim or turn off) the lighting around the patient’s bed space –.02 .64 .10 .03 .03 .07 .10 .49

Silence alarms on all equipment .06 .63 –.09 .02 –.10 .04 –.13 .45

Create a calm, peaceful environment .05 .56 .14 .07 .06 .05 .25 .48

Turn off monitoring equipment –.07 .54 .06 .07 –.14 –.03 –.14 .32

Place chairs around the bed for the family .22 .43 .07 –.13 .05 .23 –.21 .58

Cluster/coordinate care to minimise interruptions and provide privacy for the patient and family .02 .42 .28 –.10 .03 .25 –.00 .49

Remove the endotracheal tube if the patient is breathing spontaneously .11 .41 –.03 .08 .02 –.11 –.06 .21

133

Items

Factor loadings

h2 1 2 3 4 5 6 7

Draw curtains to provide privacy for the patient and family .15 .41 .07 –.09 .08 .17 –.21 .46

Move the patient to a more private area of the unit –.08 .36 .17 .06 .10 .03 .17 .23

Permit family to visit outside of regular visiting hours .16 .26 –.00 .02 .03 .18 –.19 .29

Administer medications to minimise dyspnoea and respiratory sections .00 .25 –.03 .20 .18 .00 –.12 .21

Provide the family with access to a private room to congregate .15 .24 .11 .01 .07 .06 .20 .19

Encourage the family to reminisce about the patient .10 –.06 .69 .01 –.03 –.02 –.14 .51

Ask the family about their feelings –.00 –.03 .55 .26 .00 .12 –.14 .53

Provide opportunities for the family to create positive memories .11 –.04 .53 .16 .13 .18 .03 .62

Ask the family how they are coping –.02 .05 .47 .19 .08 .13 –.14 .47

Put up photographs of the patient, family members and/or pets –.07 .22 .47 –.08 .09 .11 .09 .38

Listen to the family reminisce about the patient .08 .01 .45 .06 .04 .17 –.13 .40

Play music –.02 .17 .44 .02 .02 .12 .24 .38

Ask the family their preferences for the patient’s care .07 .06 .06 .72 .01 .03 –.04 .63

Ask the family if they would like to be involved in the care of the patient .15 –.01 .12 .66 .07 .04 .03 .62

Provide the family with options about the care of the patient .17 –.01 .09 .58 .07 .05 .11 .52

Ask the family about their knowledge of the patient's wishes for EOLC .24 .16 .00 .44 –.10 .14 –.00 .47

Document the family’s preference for the care of the patient .29 –.08 .16 .38 –.03 .16 –.01 .47

Initiate (or implement) weaning from mechanical ventilation… –.10 .17 –.05 .30 .18 .05 –.01 .18

Administer fluids to the patient to maintain hydration –.06 –.17 .02 .11 .58 –.03 –.08 .35

Maintain the patient’s airway using oro/endotracheal suction .13 .03 .04 .07 .57 –.09 .07 .39

134

Items

Factor loadings

h2 1 2 3 4 5 6 7

Keep the patient sedated, if appropriate .05 .20 –.13 .05 .46 .17 –.18 .45

Regularly obtain and document patient vital signs during the provision of EOLC –.01 –.17 .13 –.01 .45 .10 .07 .26

Administer analgesia via intravenous infusion –.02 .28 –.18 .01 .41 .20 –.20 .43

Provide pressure area care .37 .11 .13 –.20 .39 –.08 –.01 .40

Provide care to maintain patient hygiene .31 .19 –.01 –.25 .34 .16 –.11 .47

Offer to arrange for a spiritual advisor or pastoral care representative to visit… –.04 –.05 –.03 .01 –.07 .90 –.03 .71

Ask the family about spiritual needs –.04 –.08 .06 .03 –.00 .86 –.01 .74

Ask the family about cultural needs .01 –.08 .04 .06 .06 .81 .07 .71

Support the family to participate in cultural practices related to dying .05 –.03 .06 .05 .05 .77 .06 .70

Ask family members if there is anyone they would like to be there .20 .10 .11 .03 –.02 .44 –.04 .47

Stay in the room with the patient and family, if the patient and family wants .12 .05 .22 .03 –.01 .41 –.08 .43

Arrange for a counsellor or social worker to meet with the family .20 .11 .16 –.04 .04 .25 .11 .28

Encourage the family to touch the patient .12 .11 .29 –.01 .10 .04 –.63 .68

Encourage the family to talk to the patient .06 .08 .33 –.03 .16 –.00 –.56 .56

α .87 .80 .83 .84 .71 .90 .87

Note. Factor loadings > .30 are in boldface; EOLC = end-of-life care.

135

Table 9-6: Factor correlation matrix of Critical care nurses’ end-of-life care

practices section of the survey instrument

Factor 1 2 3 4 5 6 7

1 1.00 .30 .30 .32 .25 .48 –.20

2 .30 1.00 .22 .21 .21 .35 –.16

3 .30 .22 1.00 .22 .27 .45 .01

4 .32 .21 .22 1.00 .16 .33 –.04

5 .25 .21 .27 .16 1.00 .27 –.12

6 .48 .35 .45 .33 .27 1.00 –.14

7 –.20 –.16 .01 –.04 –.12 –.14 1.00

The reliability of the scales identified through EFA was assessed by calculating

Cronbach’s alpha to determine the internal consistency of the scales. The Cronbach’s

alpha for each factor is provided in the final row of Table 9–5. All scales derived

from EFA in the Critical care nurses’ end-of-life care practices section of the survey

instrument achieved a satisfactory reliability coefficient greater than .70. Therefore,

factors one through to six were accepted as scales and a summary of these factors,

the label of the scale, number of items and Cronbach’s alpha are provided in Table

9–7.

Factor one represented a combination of the scales explaining and providing

information and continuity of care proposed a priori. It was broader than the

explaining and providing information scale in that it refers to information sharing

between nurses and patients and families, and also between health care professionals,

which was originally captured in the continuity of care scale.

The items representing the latent construct, environmental modification, are similar

to those proposed a priori in the comfort care—environmental factors scale. Some of

the items from the a priori scale, emotional and practical support, were represented in

the latent construct emotional support. Items in the patient and family centred

decision making, symptom management and spiritual support latent constructs were

similar to those in the scales capturing these areas of practice proposed a priori.

136

Table 9-7: Final factors, labels and Cronbach’s alphas for Critical care nurses’

end-of-life care practices section of the survey instrument

Factor Label Items α

1 Information sharing 10 .87

2 Environmental modification 9 .80

3 Emotional support 7 .83

4 Patient and family centred decision making 5 .84

5 Symptom management 7 .71

6 Spiritual support 6 .90

9.5.2 Exploratory factor analysis: Factors influencing the provision of end-of-

life care

Factor analysis using PAF and oblique rotation (direct oblimin) was undertaken on

the Factors influencing the provision of end-of-life care section of the survey

instrument in PASW version 18. For this section of the survey instrument, Kaiser’s

criterion suggested a solution of 16 factors. The a priori factor structure suggested

eight factors, while interpretation of the scree plot suggested a solution of six or eight

factors (see Figure 9–3). Therefore, comparison of factor solutions using six, seven,

eight, nine and 16 factors was undertaken.

137

Figure 9-3: Scree plot of eigenvalues for factors in the Factors influencing the

provision of end-of-life care section of the survey instrument

Solutions of nine and 16 factors both contained a number of factors with only two or

three items loading on the factor. The eight factor solution most closely resembled

the a priori scales of the survey instrument and provided the best fit of the data in

terms of conceptual and theoretical meaningfulness of the factor structure. The

pattern coefficient (factor loading) and communality for each item in the Factors

influencing the provision of end-of-life care section of the survey instrument is

provided in Table 9–8. In this solution, 42 out of the 58 items had a pattern

coefficient greater than .30, the level required for an item to be considered to

contribute to a factor. The structure matrix for the Factors influencing the provision

of end-of-life care section of the survey instrument is provided in Appendix 13.16. In

the eight factor solution, there were 259 (15.0%) non-redundant residuals with

absolute values > 0.05.

The reliability of the scales identified through EFA was assessed by calculating

Cronbach’s alpha to determine the internal consistency of the scales. The Cronbach’s

alpha for each factor is provided in the final row of Table 9–8.

138

The factor correlation matrix (see Table 9–9) was reviewed to identify correlation

between factors. Moderate correlation (> .30) (Cohen, 1988; Field, 2009) was noted

between factors one and four and factors two and eight.

139

Table 9-8: Factor loadings for exploratory factor analysis with oblique rotation of Factors influencing the provision of end-of-life care

section of the survey instrument

Items

Factor loadings

h2 1 2 3 4 5 6 7 8

My supervisors/managers ensure staff caring for patients at the EOL are supported .82 –.05 –.01 –.06 –.04 –.03 .09 –.04 .70

My supervisors/managers provide guidance that assists me to provide EOLC .74 –.10 –.07 .01 –.03 –.00 .13 .06 .62

After caring for a patient at the EOL, my colleagues will ask me if I am OK .70 .21 –.01 –.05 .09 .06 –.03 –.02 .53

I feel supported when caring for a patient at the EOL .67 .04 –.01 .06 –.01 .09 .04 –.10 .59

My colleagues appreciate the stressors associated with caring for patients at the EOL .59 .17 –.04 –.06 –.02 .17 –.09 –.13 .51

When a patient dies in my Unit, counselling is available if I need it .53 –.14 .04 .01 .01 .04 .24 .05 .39

In my Unit, nurses view the provision of EOLC as an important component of critical care nursing .28 .22 –.02 .12 .00 .17 –.00 –.04 .25

Lack of staff results in no ‘down time’ between the death of a patient and the admission of the next patient .28 –.17 –.15 .13 –.16 –.17 –.04 –.04 .22

In my Unit, nurses are expected to cope with death and dying .24 –.21 .04 .06 –.22 –.16 –.03 .01 .19

EOLC is as important as curative care in the critical care environment –.02 .64 –.03 –.02 .01 –.04 .07 –.05 .46

I can influence the patient and family’s experience of EOLC .04 .58 .07 .04 –.06 –.04 –.05 –.15 .41

EOL care should include care of the patient’s family .00 .54 –.11 –.13 .05 .10 –.05 –.04 .37

It is a privilege to care for a patient at the EOL and their family .14 .53 –.00 .02 –.01 –.14 .02 –.12 .39

My own personal experiences of death have influenced the care I provide to patients at the EOL and their family –.08 .38 .04 .08 .00 –.02 .07 –.02 .16

Analgesia should be titrated to keep the patient comfortable even if this hastens death .05 .35 –.07 –.03 .08 –.01 –.01 –.11 .19

Pressure area care should continue to be provided for a patient at the EOL –.04 .31 .01 –.02 –.01 .09 .02 –.03 .11

It is OK to cry when a patient dies .04 .29 –.04 –.00 –.13 –.07 .01 –.08 .14

140

Items

Factor loadings

h2 1 2 3 4 5 6 7 8

EOLC requires advanced nursing skills .08 .28 .01 –.00 –.00 –.00 .27 –.00 .18

I have learnt from other nurses how to provide EOLC .15 .20 –.11 .15 .15 .15 –.00 .14 .18

In my Unit, families are given adequate time to consider decisions to withhold/withdraw treatment for the patient –.02 –.12 –.76 .02 –.04 .03 .07 –.01 .57

In my Unit, family meetings with the health care team are held to resolve differing views about a patient’s treatment .08 –.15 –.71 .04 –.02 .09 .05 –.04 .58

In my Unit, the patient’s preferences (or their likely preference according to family) for EOLC are considered .12 –.03 –.57 –.07 –.01 .27 .09 .10 .49

In my Unit, families are involved in decisions about the dying patient .05 .07 –.53 .11 –.07 .08 –.03 .10 .39

In my Unit, withdrawal of treatment is delayed whilst waiting for relatives of the patient to arrive –.08 .12 –.50 .05 .10 .02 –.04 –.01 .29

When a patient dies in my Unit, families have sufficient time to spend with the patient .19 –.06 –.33 .27 –.15 .04 –.21 –.10 .42

I am often exposed to death in the critical care environment .04 .17 –.21 –.02 .20 .02 –.05 –.04 .13

Abrupt withdrawal of mechanical ventilation may contribute to dyspnoea –.03 .19 –.19 .04 .14 –.07 .03 –.12 .14

The physical environment of my Unit is ideal for providing EOLC –.09 –.04 .06 .77 –.05 –.07 .02 –.02 .54

My Unit is adequately equipped to support the comfort needs of the family during EOLC –.14 .01 –.07 .73 –.02 .01 .11 –.00 .53

Private rooms are available to care for the patient at the EOL –.01 .13 .04 .71 .10 –.01 .02 .05 .48

My Unit is adequately staffed for providing the needs of dying patients and their families .13 –.18 –.21 .59 –.03 –.04 –.03 –.13 .56

When a patient dies in my Unit, nurses have sufficient time to spend with the family .23 –.22 –.27 .37 –.10 –.01 –.21 –.37 .62

In my Unit, nurses that know the patient and family are allocated to care for the patient to maintain continuity… .19 .05 –.03 .28 –.18 .12 .14 .12 .29

Following a decision to withhold or withdraw treatment, patients remain in my Unit for EOLC .07 .22 –.06 .27 –.08 .09 –.02 –.01 .18

After a decision is made to withdraw treatment, less time is spent with the patient by the critical care medical team .00 .01 –.01 –.06 .75 –.13 –.10 –.08 .61

After a decision is made to withdraw treatment, nurses are left to manage the care of the patient –.03 –.02 .05 –.01 .72 –.15 –.03 –.09 .57

In my Unit, EOLC is mostly allocated to junior nurses –.09 –.12 –.02 –.04 .32 –.15 .10 .14 .21

141

Items

Factor loadings

h2 1 2 3 4 5 6 7 8

After a decision is made to withdraw treatment, I spend less time with the patient .13 –.21 .15 .02 .32 .16 –.28 .12 .35

The medical staff support EOLC for dying patients in my Unit .08 .01 –.17 –.04 –.117 .55 –.07 –.09 .45

In my Unit, when a diagnosis with a likely poor outcome is made, families are informed of EOLC options –.00 –.04 –.30 –.02 –.17 .53 .07 –.02 .48

In my Unit, the team expresses its opinions, values and beliefs about providing care to dying patients .20 .04 .01 .01 –.04 .51 .01 –.03 .38

All members of the healthcare team in my Unit agree with and support EOLC when it is implemented .08 .09 –.07 .19 –.00 .47 –.18 –.04 .38

In my Unit, a plan of care for the patient is clearly documented by the medical team .01 –.09 –.17 .11 –.24 .43 –.01 –.10 .40

In my Unit, the critical care team and the patient’s primary medical or surgical team agree on a plan of care .06 –.03 –.21 .06 –.14 .42 .00 –.04 .34

The provision of EOLC requires emotional detachment .04 .22 .06 –.02 –.21 –.24 –.12 –.09 .17

I have received in-service education that assists me to support and communicate with families of dying patients .15 –.20 .13 .05 .06 .26 .57 –.28 .60

I have received in-service education that assists me to provide care and manage patient symptoms at the EOL .11 –.23 .16 .08 .04 .22 .54 –.33 .57

I draw on evidence from the literature to provide EOLC to patients and their families .02 .08 –.25 –.02 –.07 –.10 .43 –.13 .33

My postgraduate nursing education included EOLC content relevant to the critical care context .16 .04 .05 .05 –.04 .04 .43 –.01 .26

Patients at the EOL require little nursing care –.07 .21 –.17 –.20 –.21 –.16 .34 .01 .31

My undergraduate nursing education included EOLC content .12 .01 .12 .13 .04 .01 .28 .02 .14

There are policies/guidelines to assist in the delivery of EOLC in my Unit .16 –.01 –.00 .11 –.03 .26 .28 .04 .26

When withdrawing mechanical ventilation at the EOL, ventilator weaning should occur… –.01 .03 –.05 .00 .04 –.12 .27 –.01 .09

Intravenous fluids should be provided to maintain hydration for the patient at the EOL –.02 .02 –.03 .05 –.10 .06 .23 –.03 .08

I feel adequately prepared to care for patients at the EOL .02 .06 –.02 .03 .07 .08 .08 –.67 .52

I don’t know what to say to the family of a patient at the EOL .00 .06 .04 .05 –.02 –.02 .05 –.64 .45

I have limited knowledge of symptom management for patient’s at the EOL –.06 .07 .07 –.02 .10 .05 .07 –.52 .29

142

Items

Factor loadings

h2 1 2 3 4 5 6 7 8

I feel a sense of personal failure when a patient dies .15 .14 –.09 –.13 –.11 –.06 –.08 –.29 .22

α .85 .70 .79 .81 .64 .80 .66 .69

Note. Factor loadings > .30 are in boldface; EOL = end of life; EOLC = end-of-life care.

143

Table 9-9: Factor correlation matrix for the Factors influencing the provision of

end-of-life care section of the survey instrument

Factor 1 2 3 4 5 6 7 8

1 1.00 .01 –.20 .32 –.23 .29 .13 –.20

2 .01 1.00 –.22 –.05 .04 –.02 .08 –.33

3 –.20 –.22 1.00 –.19 .14 –.21 .00 .17

4 .32 –.05 –.19 1.00 –.13 .24 .10 –.07

5 –.23 .04 .14 –.13 1.00 –.02 –.07 .14

6 .29 –.02 –.21 .24 –.02 1.00 .05 –.02

7 .13 .08 .00 .10 –.07 .05 1.00 –.13

8 –.20 –.33 .17 –.07 .14 –.02 –.13 1.00

Most of the scales achieved a satisfactory Cronbach’s alpha of greater than .70.

Factor one, consisting of six items, represented the latent construct labelled

emotional support for nurses.

Deletion of two items from factor two that do not conceptually fit within this scale

(Pressure area care should continue to be provided for a patient at the end of life

and My own personal experiences of death have influenced the care I provide to

patients at the end of life and their family) increases the alpha of this factor to .73.

The five remaining items in factor two represented the latent construct labelled

palliative values.

Factor three consisted of six items and was labelled patient and family preferences.

The five items in factor four represented the latent construct labelled resources.

Initially, four items loaded on factor five. The item After a decision is made to

withdraw treatment, I spend less time with the patient refers to the self, as opposed to

the other items on factor five, which refer to the larger organisational culture.

Deletion of this item increases the alpha of this scale to .69. The three items in factor

five reflected the latent construct labelled organisational culture.

144

Factor six comprised six items reflecting support for end-of-life care and organising

a plan of care. These items represented the latent construct labelled care planning.

Factor seven was labelled knowledge, to represent the content of the four items that

loaded on this factor. The item Patients at the end-of-life require little nursing care

was conceptually a poor fit with the other items that loaded on factor seven. Deletion

of this item from the scale resulted in an increase of the alpha to .69.

Lastly, the label preparedness was chosen for factor eight to represent the latent

construct measured by the three items in this factor.

Table 9-10: Final factors, labels and Cronbach’s alpha for the Factors

influencing the provision of end-of-life care section of the survey instrument

Factor Label Items α

1 Emotional support for nurses 6 .85

2 Palliative values 5 .73

3 Patient and family preferences 6 .79

4 Resources 5 .81

5 Organisational culture 3 .69

6 Care planning 6 .80

7 Knowledge 4 .69

8 Preparedness 3 .69

Death Anxiety Scale

The internal consistency of the DAS in the current study is the same as that reported

by Templer (1970), who developed the true-false version of DAS (Cronbach’s alpha

.76). McMordie (1979) revised and tested the DAS using a seven-point Likert scale

response format and achieved a slightly higher Cronbach’s alpha of .84. A five-point

Likert scale was utilised in the current study to maintain consistency in response

format across all items in the survey instrument.

145

9.6 Bivariate analysis

The psychometric properties of each of the summated scales generated from EFA are

provided in Tables 9–11 and 9–12. The summated scales from the Critical care

nurses’ end-of-life care practices section of the survey instrument formed the

outcome (dependent) variables in all bivariate and multivariable analysis. The

descriptive statistics and histogram for each of these scales were reviewed to

determine if they met the assumption of normal distribution. An example of the

histogram for the information sharing summated scale is provided in Figure 9–4.

Table 9-11: Psychometric properties of the summated scales from the Critical

care nurses’ end-of-life care practices section of the survey instrument (n = 392)

Scale M SD Median

Range

Skew Potential Actual

Information sharing 44.7 4.7 46 10–50 20–50 –1.09

Environmental modification 38.7 4.0 39 9–45 20–45 –.81

Emotional support 26.7 4.3 27 7–35 11–35 –.43

Patient and family centred

decision making 20.4 3.5 21 5–25 9–25 –.64

Symptom management 29.4 3.4 29 7–35 14–35 –.45

Spiritual support 26.4 3.9 28 6–30 11–30 –1.22

Note. M = Mean; SD = Standard deviation.

146

Table 9-12: Psychometric properties of the summated scales from the Factors

influencing the provision of end-of-life care section of the survey instrument

Scale n M SD Median

Range

Skew Potential Actual

Emotional support for

nurses 383 21.7 4.2 23 6–30 6–30 –.66

Palliative values 386 22.7 2.2 23 5–25 13–25 –.97

Patient and family

preferences 392 24.6 3.1 24 6–30 7–30 –.98

Resources 392 15.2 4.2 16 5–25 5–25 –.15

Organisational culture 392 8.5 2.6 9 3–15 3–15 –.25

Care planning 392 22.7 3.9 23 6–30 9–30 –.63

Knowledge 388 12.1 3.2 12 4–20 4–20 .10

Preparedness 387 11.8 2.3 12 3–15 3–15 –.54

Death Anxiety Scale 392 41.7 6.9 42 15–75 15–61 –.25

Note. M = Mean; SD = Standard deviation.

Figure 9-4: Histogram of information sharing summated scale

147

The normality assumption was also tested using the Shapiro-Wilk test. A significant

result was obtained on the Shapiro-Wilk test for each summated scale, indicating that

the data were significantly different from a normal distribution. However, it is noted

that a significant result on this test can occur in large samples when the distribution is

only slightly different from normal (Hair et al., 2006).

Based on the descriptive statistics, histograms and Shapiro Wilk tests, deviations

from normality were noted for each of the summated scales from the Critical care

nurses’ end-of-life care practices section of the survey instrument. However, it has

been acknowledged that bivariate and multivariate statistical methods are robust against

violations of this assumption (Hair et al., 2006; Tabachnick & Fidell, 2007).

9.6.1 Bivariate analysis of critical care nurses’ end-of-life care practices

summated scales and demographic variables

Information sharing

Statistically significant differences between male and female participants (U =

10837.5, z = 2.43, p < .05; Male Mdn = 43, Female Mdn = 46) and between

participants with postgraduate qualifications (PG) and those without postgraduate

qualifications (non-PG) in nursing (U = 13161.5, z = 2.54, p < .05; PG Mdn = 46,

non-PG Mdn = 44) on the information sharing summated scale were found.

A statistically significant difference across the categories of the demographic

variable years of experience in nursing and the outcome variable of information

sharing [H(4) = 10.88, p < .05] was also identified. Pairwise multiple comparisons

revealed that participants in the 0–5 years category were significantly different from

those in the 11–15 years and 16–20 years of experience categories. However, when

adjustment for multiple comparison was made, a statistically significant difference

was only found between participants with 0–5 years and 16–20 years of experience

(T = –70.82, z = –2.84, p < .05). A boxplot of years of experience in nursing and

information sharing practices is provided in Figure 9–5.

148

Although not statistically significant, the demographic variables of age, location,

years of critical care experience, and time since provision of end-of-life care

achieved a significance level of p < .2 and were identified as potential variables of

importance to information sharing practices.

Figure 9-5: Boxplot of years of experience in nursing and information sharing

practices

Environmental modification

Statistically significant differences between male and female participants (U =

10581.5, z = 2.09, p < .05; Male Mdn = 37, Female Mdn = 39), between participants

working in a major city (Mdn = 39) and those in a regional/remote area (Mdn = 40)

(U = 17935.0, z = 2.23, p < .05) and between participants with postgraduate

qualifications (PG) and those without postgraduate qualifications (non-PG) in

nursing (U = 13256.5, z = 2.65, p < .01; PG Mdn = 39, non-PG Mdn = 38) on the

environmental modification summated scale were found.

149

Although not statistically significant, the demographic variables of age, formal

education in critical care and time since direct patient care achieved a significance

level of p < .2 and were identified as potential variables of importance to

environmental modification practices.

Emotional support

A small positive relationship was found between age and emotional support (rs = .13,

p < .05, n = 391). In addition, a statistically significant difference between male and

female participants was found on the emotional support (U = 11958.0, z = 3.89, p <

.001; Male Mdn = 24, Female Mdn = 27) summated scale.

A statistically significant difference across the categories of the demographic

variable of years of experience in nursing and the outcome variable of emotional

support [H(4) = 11.29, p < .05] was identified. However, no adjusted statistically

significant differences were noted following multiple pairwise comparisons. A box

plot of the categories of years of experience in nursing and the outcome variable of

emotional support is provided in Figure 9–6.

Figure 9-6: Boxplot of years of experience in nursing and emotional support

150

A statistically significant difference across workplaces on the emotional support

[H(5) = 24.01, p < .001] summated scale was identified. Multiple pairwise

comparisons found an adjusted statistically significant difference between the

emergency department workplace and the workplaces general/specialty ICU (T = –

82.79, z = –2.98, p < .05), ICU/CCU/HDU (T = –90.43, z = –3.12, p < .05),

paediatric ICU (T = –152.27, z = –4.49, p < .001) and other clinical (T = 136.87, z =

3.47, p < .01). An adjusted statistical significance was also identified between

general/specialty ICU and paediatric ICU (T = 69.48, z = 3.10, p < .05).

Although not statistically significant, the demographic variable of years of

experience in critical care achieved a significance level of p < .2 and was identified

as a potential variable of importance to emotional support practices.

Patient and family centred decision making

A statistically significant difference between participants with postgraduate

qualifications (PG) and those without postgraduate qualifications (non-PG) in

nursing on the patient and family centred decision making (U = 13793.5, z = 3.29, p

< .01; PG Mdn = 21, non-PG Mdn = 19) summated scale was identified.

Bivariate analysis also revealed a statistically significant difference across

workplaces on the patient and family centred decision making [H(5) = 13.41, p <

.05] summated scale. Multiple pairwise comparisons found an adjusted statistically

significant difference between paediatric ICU and general/specialty ICU for patient

and family centred decision making (T = 71.40, z = 3.19, p < .05).

Although not statistically significant, the demographic variables of gender, location

and time since provision of end-of-life care achieved a significance level of p < .2

and were identified as potential variables of importance to patient and family centred

decision making practices.

151

Symptom management

A small positive relationship was found between age and symptom management (rs =

.17, p < .01, n = 391). A statistically significant difference across the categories of

the variable years of experience in nursing and the outcome variable of symptom

management [H(4) = 16.04, p < .01] was also identified. However, no adjusted

statistically significant difference was found following multiple pairwise

comparisons. A box plot of the relationship between years of experience in nursing

and symptom management is provided in Figure 9–7.

Figure 9-7: Boxplot of years of experience in nursing and symptom management

A statistically significant difference was found between the categories of years of

experience in critical care and the outcome variable of symptom management H(4) =

10.97, p < .05. Pairwise comparison of the categories identified those with 0–5 and

6–10 years of experience were significantly different from both the 16–20 years and

more than 20 years of experience in critical care categories. However, these

significant results were not retained when adjusting for multiple comparisons.

152

A statistically significant difference across workplaces was found on the symptom

management [H(5) = 20.32, p < .01] summated scale. Follow-up analysis using

multiple pairwise comparisons revealed an adjusted statistically significant

difference between the emergency department and paediatric ICU (T = –108.06, z =

–3.19, p < .05) and between the emergency department and workplace category other

clinical (T = 131.50, z = 3.34, p < .05) for the symptom management summated

scale.

Although not statistically significant, the demographic variable of time since

provision of end-of-life care achieved a significance level of p < .2 and was

identified as a potential variable of importance to symptom management practices.

Spiritual support

Bivariate analysis identified statistically significant differences between male and

female participants (U = 11292.5, z = 3.06, p < .01; Male Mdn = 26, Female Mdn =

28) and participants with postgraduate qualifications (PG) and those without

postgraduate qualifications (non-PG) in nursing (U = 13172.5, z = 2.58, p < .05; PG

Mdn = 28, non-PG Mdn = 26) on the spiritual support summated scale.

A statistically significant difference across the categories of the variable years of

experience in nursing and the outcome variable of spiritual support [H(4) = 13.08, p

< .05] was found. Multiple pairwise comparisons revealed an adjusted statistically

significant difference between 0–5 years and 11–15 years of experience in nursing (T

= –85.52, z = –3.05, p < .05) and 6–10 years and 11–15 years of experience in

nursing (T = –69.85, z = –2.88, p < .05) for spiritual support.

Although not statistically significant, the demographic variables of workplace and

time since provision of end-of-life care achieved a significance level of p < .2 and

were identified as potential variables of importance to spiritual support practices.

153

No statistically significant or potentially important results were found between the

demographic variable of primary nursing role and any of the outcome (dependent)

variables.

9.6.2 Bivariate analysis of critical care nurses’ end-of-life care practices

summated scales and factors influencing the provision of end-of-life care

summated scales

Spearman’s correlation coefficient was calculated between the summated scales of

each outcome variable and the explanatory variables. A number of statistically

significant relationships were identified (see Table 9–13). All variables correlated at

the minimum level of p < .2 indicating potential variables of importance with the

exception of the relationships between environment modification and knowledge,

symptom management and emotional support for nurses, and symptom management

and resources.

154

Table 9-13: Spearman’s rho correlation coefficients between outcome variables and explanatory factors

Information

sharing

Environmental

modification

Emotional

support

Patient and family centred

decision making

Symptom

management

Spiritual

support

Emotional support for nurses

n = 383 .20** .12* .20** .15** .05 .13*

Palliative values

n = 386 .38** .38** .32** .33** .25** .37**

Patient and family preferences

n = 392 .31** .24** .28** .31** .16** .31**

Resources

n = 392 .09 .20** .20** .12* .06 .12*

Organisational culture

n = 392 –.06 .08 –.13* –.08 -.09 -.10

Care planning

n = 392 .32** .10* .19** .20** .11* .20**

Knowledge

n = 388 .18** –.01 .24** .29** .10* .11*

Preparedness

n = 387 .35** .18** .28** .35** .15** .31**

155

Death Anxiety Scale

n = 392 –.15** –.09 –.14** –.17** –.08 -.17**

*p < .05, two-tailed. **p < .01, two-tailed.

156

9.7 Multivariable analysis

A univariate general linear model was developed using a backward elimination

approach in PASW software version 18 for each of the six outcome variables to

identify the factors associated with critical care nurses’ end-of-life care practices. A

description of each model and model parameters is outlined below. An overview of

the results of assumption testing is then provided.

9.7.1 Information sharing practices model

A statistically significant model of critical care nurses’ information sharing practices

was developed F(3, 377) = 40.53, p < .001. This model explained (adjusted R2)

23.8% of the variation in information sharing practices. The model consisted of three

variables that partially explain information sharing: palliative values F(1, 377) =

41.21, p < .001, partial η2 = .10, care planning F(1, 377) = 27.74, p < .001, partial η2

= .07 and preparedness F(1, 377) = 17.32, p < .001, partial η2 = .04. The parameter

estimates for each explanatory variable included in the final model are provided in

Table 9–14. A positive relationship existed between each of the explanatory

variables and the outcome variable. For example, as palliative values increase by one

unit, information sharing practices increase by .64 when all other variables are held

constant.

Table 9-14: Parameter estimates for information sharing model

Parameter B SE B t 95% CI Partial η2

Palliative values 0.64** 0.10 6.42 [0.45, 0.84] .099

Care Planning 0.29** 0.06 5.27 [0.18, 0.40] .069

Preparedness 0.41** 0.10 4.16 [0.22, 0.61] .044

**p < .001

157

9.7.2 Environmental modification practices model

The environmental modification practices model was significant F(5, 380) = 19.55, p

< .001 and explained (adjusted R2) 19.4% of the variation in environmental

modification practices. The overall model consisted of five variables that partially

explain environmental modification practices: Palliative values F(1, 380) = 43.00, p

< .001, partial η2 = .10, patient and family preferences F(1, 380) = 6.91, p < .01,

partial η2 = .02, resources F(1, 380) = 20.47, p < .001, partial η2 = .05,

organisational culture F(1, 380) = 8.45, p < .01, partial η2 = .02 and work location

F(1, 380) = 5.49, p < .05, partial η2 =.01. The parameter estimates for the

environmental modification model are provided in Table 9–15.

Table 9-15: Parameter estimates for environmental modification model

Parameter B SE B t 95% CI Partial η2

Work location—Major city –0.98* 0.42 –2.34 [–1.80, –0.16] .014

Work location—

Regional/remote 0a . . . .

Palliative values 0.57*** 0.09 6.56 [0.40, 0.74] .102

Patient and family preferences 0.18** 0.07 2.63 [0.04, 0.31] .018

Resources 0.22*** 0.05 4.53 [0.12, 0.32] .051

Organisational culture 0.22** 0.08 2.91 [0.07, 0.37] .022

a. This parameter is set to zero because it is redundant.

*p < .05 **p < .01 ***p < .001

The model identified a statistically significant difference in environmental

modification practices between regional areas and major cities. Major cities had a

difference of –.98 in environmental modification practices compared with regional

areas. A positive relationship existed between the parameters palliative values,

patient and family preferences, resources and organisational culture and the outcome

variable of environmental modification.

158

9.7.3 Emotional support practices model

A model of critical care nurses’ emotional support practices was developed. This

model was significant F(10, 366) = 12.10, p < .001 and explained (adjusted R2)

22.8% of the variation in emotional support practices. The model consisted of six

variables that partially explain emotional support practices: Palliative values F(1,

366) = 6.92, p < .01, partial η2 = .02, patient and family preferences F(1, 366) =

10.18, p < .01, partial η2 = .03, knowledge F(1, 366) = 8.75, p < .01, partial η2 = .02,

preparedness F(1, 366) = 12.03, p < .01, partial η2 = .03, gender F(1, 366) = 13.12, p

< .001, partial η2 = .04 and workplace F(5, 366) = 4.62, p < .001, partial η2 = .06.

Post-hoc pairwise comparison identified a significant mean difference in emotional

support practices between participants working in an emergency department and all

other clinical areas (p < .05). Emotional support practices were lower in emergency

departments than in other clinical areas. No other comparisons between workplaces

were found to be significant.

Table 9–16 provides the parameter estimates for the emotional support practices

model. Male gender was associated with a statistically significant decrease in

emotional support practices by 2.12, compared with females. Each of the explanatory

variables, palliative values, patient and family preferences, knowledge and

preparedness, had a positive relationship with emotional support practices.

Table 9-16: Parameter estimates for emotional support model

Parameter B SE B t 95% CI Partial η2

Male –2.12*** 0.59 –3.62 [–3.27, –0.97] .035

Female 0a . . . .

Other non-clinical –0.22 0.71 –0.31 [–1.62, 1.17] .000

Other clinical 2.43* 1.04 2.34 [0.39, 4.47] .015

Emergency department –3.57*** 0.99 –3.60 [–5.51, –1.62] .034

Paediatric ICU 0.99 0.80 1.24 [–0.59, 2.58] .004

ICU/CCU/HDU 0.29 0.49 0.59 [–0.68, 1.25] .001

General/Specialty ICU 0a . . . .

159

Palliative values 0.25** 0.10 2.63 [0.06, 0.44] .019

Patient and family preferences 0.22** 0.07 3.19 [0.08, 0.35] .027

Knowledge 0.19** 0.07 2.96 [0.07, 0.32] .023

Preparedness 0.33** 0.10 3.47 [0.14, 0.52] .032

Note. ICU = intensive care unit; CCU = coronary care unit; HDU = high dependency

unit.

a. This parameter is set to zero because it is redundant

*p < .05 **p < .01 ***p < .001

9.7.4 Patient and family centred decision making practices model

A univariate general linear model was developed to identify the factors associated

with critical care nurses’ patient and family centred decision making practices. The

final model was significant F(8, 362) = 17.61, p < .001 and explained (adjusted R2)

26.4% of the variation in patient and family centred decision making practices. The

overall model consisted of eight variables that partially explain patient and family

centred decision making practices: Emotional support for nurses F(1, 362) = 4.48, p

< .05, partial η2 = .01, palliative values F(1, 362) = 12.72, p < .001, partial η2 = .03,

patient and family preferences F(1, 362) = 21.90, p < .001, partial η2 = .06,

knowledge F(1, 362) = 13.33, p < .001, partial η2 = .04, preparedness F(1, 362) =

19.20, p < .001, partial η2 = .05, death anxiety F(1, 362) = 5.09, p < .05, partial η2 =

.01, work location F(1, 362) = 4.74, p < .05, partial η2 = .01 and qualification F(1,

362) = 4.86, p < .05, partial η2 = .01.

The parameter estimates for each variable included in the patient and family centred

decision making practices model is provided in Table 9–17. The model identified a

statistically significant difference in patient and family decision making practices

between regional areas and major cities. Major cities had a difference of –.77 in

patient and family decision making practices compared with regional areas. In

addition, participants without postgraduate qualifications had a difference in patient

and family centred decision making of –.94 compared with those with postgraduate

qualifications. A statistically significant positive relationship was identified between

160

the explanatory variables of palliative values, patient and family preferences,

knowledge and preparedness and the outcome variable of patient and family centred

decision making. A statistically significant negative relationship was noted between

the explanatory variables of emotional support for nurses and DAS and the outcome

variable of patient and family centred decision making.

Table 9-17: Parameter estimates for patient and family centred decision making

model

Parameter B SE B t 95% CI Partial η2

Work location—Major city –0.77* 0.35 –2.18 [–1.46, –0.08] .013

Work location—

Regional/remote

0a

.

. .

.

Hospital certificate/

Undergraduate

–0.94*

0.43

–2.21 [–1.78, –0.10]

.013

Postgraduate 0a . . . .

Emotional support for nurses –0.09* 0.04 –2.12 [–0.17, -0.01] .012

Palliative values 0.27*** 0.08 3.57 [0.12, 0.42] .034

Patient and family preferences 0.25*** 0.05 4.68 [0.15, 0.36] .057

Knowledge 0.20*** 0.06 3.65 [0.09, 0.31] .036

Preparedness 0.34*** 0.08 4.38 [0.19, 0.49] .050

Death anxiety scale –0.05* 0.02 –2.26 [–0.10, –0.01] .014

a. This parameter is set to zero because it is redundant

*p < .05 **p < .01 ***p < .001

9.7.5 Symptom management practices model

A univariate general linear model was developed to identify the factors associated

with critical care nurses’ symptom management practices. The overall model was

significant F(8, 376) = 7.10, p < .001) and explained (adjusted R2) 11.3% of the

variation in symptom management practices. The overall model consisted of four

variables that partially explain symptom management practices: Palliative values

161

F(1, 376) = 15.79, p < .001, partial η2 = .04, patient and family preferences F(1, 376)

= 4.49, p < .05, partial η2 = .01, age F(1, 376) = 12.65, p < .001, partial η2 = .03 and

workplace F(5, 376) = 3.17, p < .01, partial η2 = .04. Post-hoc pairwise comparisons,

with Bonferroni correction for multiple comparisons, identified no significant mean

difference between the workplace categories. The parameter estimates for the

symptom management model are provided in Table 9–18. A significant positive

relationship was identified between the explanatory variables of palliative values,

patient and family preferences and age and the outcome variable of symptom

management practices.

Table 9-18: Parameter estimates for symptom management model

Parameter B SE B t 95% CI Partial η2

Other non-clinical 1.01 0.59 1.71 [–0.15, 2.17] .008

Other clinical 2.59** 0.88 2.95 [0.86, 4.32] .023

Emergency department –0.62 0.84 –0.75 [–2.27, 1.02] .001

Paediatric ICU 1.41* 0.68 2.08 [0.08, 2.74] .011

ICU/CCU/HDU 0.08 0.41 0.19 [–0.73, 0.89] .000

General/Specialty ICU 0a . . . .

Palliative values 0.30*** 0.08 3.97 [0.15, 0.46] .040

Patient and family preferences 0.12* 0.06 2.12 [0.01, 0.23] .012

Age 0.07*** 0.02 3.56 [0.03, 0.10] .033

Note. ICU = intensive care unit; CCU = coronary care unit; HDU = high dependency

unit.

a. This parameter is set to zero because it is redundant.

*p < .05 **p < .01 ***p < .001

162

9.7.6 Spiritual support practices model

A univariate general linear model was developed to identify the factors associated

with critical care nurses’ spiritual support practices. The overall model was

significant F(9, 367) = 14.66, p < .001 and explained (adjusted R2) 24.6% of the

variation in spiritual support practices. The overall model consisted of six variables

that partially explain spiritual support practices: Palliative values F(1, 367) = 18.64,

p < .001, partial η2 = .05, patient and family preferences F(1, 367) = 17.16, p < .001,

partial η2 = .05, preparedness F(1, 367) = 16.83, p < .001, partial η2 = .04, gender

F(1, 367) = 11.84, p < .01, partial η2 = .03, nursing experience F(4, 367) = 3.11, p <

.05, partial η2 = .03 and qualification F(1, 367) = 4.24, p < .05, partial η2 = .01. Post-

hoc pairwise comparison with Bonferroni correction revealed a significant mean

difference between 0–5 years and 11–15 years of nursing experience and between 0–

5 years and more than 21 years of nursing experience. Participants with 0–5 years of

experience had lower mean scores on the spiritual support summated scale.

Table 9–19 provides the parameter estimates for the variables included in the

spiritual support practices model. It is noted that male gender was associated with a

change in spiritual support practices of –1.78 compared with females. The

explanatory variables of palliative values, patient and family preferences and

preparedness all had a positive relationship with the outcome variable of spiritual

support.

Table 9-19: Parameter estimates for spiritual support model

Parameter B SE B t 95% CI Partial η2

Nursing experience 0–5 years –2.02** 0.69 –2.93 [–3.37, –0.66] .023

Nursing experience 6–10 years –0.71 0.52 –1.36 [–1.73, 0.32] .005

Nursing experience 11–15 years 0.81 0.65 1.25 [–0.46, 2.09] .004

Nursing experience 16–20 years –0.18 0.51 –0.35 [–1.17, 0.82] .000

Nursing experience >21 years 0a . . . .

Male –1.78** 0.52 –3.44 [–2.80, –0.77] .031

Female 0a . . . .

163

Hospital certificate/

Undergraduate

–0.99*

0.48

–2.06 [–1.93, –0.04]

.011

Postgraduate 0a . . . .

Palliative values 0.36*** 0.08 4.32 [0.20, 0.53] .048

Patient and family preferences 0.24*** 0.06 4.14 [0.12, 0.35] .045

Preparedness 0.34*** 0.08 4.10 [0.18, 0.50] .044

a. This parameter is set to zero because it is redundant

*p < 0.05 **p < 0.01 ***p < 0.001

A summary of the significant explanatory variables in each practice model is

provided in Table 9–20. The explanatory variable of palliative values was significant

in each practice model. Patient and family preferences was a significant explanatory

variable in all models except the information sharing practices model. The

explanatory variables representing the construct context were only significant in the

environmental modification practices model and were not identified as significant

explanatory variables in the other practice models.

164

Table 9-20: Comparison of practice models and contributing parameters

Information

sharing

Environmental

modification

Emotional support Patient and family centred decision

making

Symptom

management

Spiritual support

Context

Organisational culture X

Resources X

Evidence

Patient and family preferences X X X X X

Knowledge X X

Preparedness X X X X

Facilitation

Palliative values X X X X X X

Emotional support for nurses X

Care planning X

Death Anxiety Scale X

Demographics

Age X

Gender X X

Nursing experience X

Critical care experience

Work location X X

Workplace X X

165

Qualification X X

Critical care education

Primary nursing role

Time since direct patient care

Time since provision of EOL care

Note. EOL = end of life.

166

9.8 Assumption testing

The assumptions of normality, linearity, homoscedasticity/homogeneity and

multicollinearity were assessed for each model. Prior to bivariate analysis, review of

the histogram and Shapiro Wilk statistic for each outcome variable indicated that the

data was not normally distributed. However, both bivariate and multivariate statistics

are generally robust against violations of this assumption, particularly when sample

sizes exceed 200 participants (Hair et al., 2006), as they did in this phase of the

study.

Residual plots of observed-by-predicted-by-standardised values for each of the

models were also obtained to assess the assumptions of multivariate normality,

linearity and equality of variance. Review of residual plots for equal variance of the

residuals, with the residuals randomly distributed around zero, indicates the

assumptions are met within the model. Figure 9–8 is an example of the residual plot

for the emotional support practices model.

Figure 9-8: Residual plot of emotional support practices model

Homoscedasticity/homogeneity of variances was also assessed using Levene’s Test

of Equality of Error Variances. A non-significant result on this test indicates that the

167

assumption is not violated and that variances are equal. Levene’s test was significant

for the spiritual support model F(18, 358) = 2.30, p < .01), indicating violation of this

assumption. Similarly to tests of normality, Levene’s test is sensitive to large sample

sizes, with small deviations in equality of variances producing a significant result on

this test (Field, 2009; Hair et al., 2006). In light of the large sample used in phase

three of the study and the assumption of homoscedasticity being met in five of the six

practice models, no further action was considered necessary in regard to the

significant result on Levene’s test for the spiritual support model.

9.9 Summary

In summary, analysis of data obtained from survey responses of 392 critical care

nurses was undertaken in this third phase of the study. Key results of this phase

included the identification of six end-of-life care practice areas: information sharing,

environmental modification, emotional support, patient and family centred decision

making, symptom management and spiritual support. Using a backward elimination

approach, a general linear model for each of the six practice areas was built,

identifying the significant explanatory variables of each end-of-life care practice

area. The explanatory variable of palliative values was a significant contributor in

each practice model, and patient and family preferences was a significant

explanatory variable in all models except the information sharing practices model.

However, the explanatory variables representing the construct context were only

significant to environmental modification practices and not identified as significant

explanatory variables in the other practice models.

168

10 DISCUSSION

10.1 Introduction

The purpose of this three-phase exploratory sequential mixed methods study was to

identify the practices undertaken by critical care nurses in the provision of end-of-life

care and the factors associated with engagement in these practices. Three key

research questions guided this study: What are the most frequently reported end-of-

life care practices of Australian critical care nurses? What are the factors influencing

the provision of end-of-life care as perceived by Australian critical care nurses?

What factors are associated with Australian critical care nurses’ end-of-life

practices? The findings of this study depict a context of practice in which critical

care nurses actively engage in many important practices to facilitate quality care at

the end of life. The findings do, however, highlight some variation in the extent to

which these practices are reported by nurses. Importantly, the study has identified

key factors that contribute to this variation. This chapter will discuss the key findings

of the study. The chapter concludes with a discussion of the strengths and limitations

of the study.

10.2 What are the most frequently reported end-of-life care practices of

Australian critical care nurses?

Six end-of-life care practice areas were identified in phase three of this study:

information sharing, environmental modification, emotional support, patient and

family centred decision making, symptom management, and spiritual support. The

practices most frequently identified as always undertaken were within the

environmental modification and information sharing practice areas. The practices

least frequently identified as always undertaken included items from the emotional

support and symptom management practice areas.

The majority of nurses in this study reported always engaging in practices to modify

the environment to support family presence and create a peaceful setting for the

patient and family to experience end-of-life care. Specifically, permitting the family

169

to visit outside of regular visiting hours and placing chairs around the bed for the

family were the most frequently reported end-of-life care practices, with 75% of

participants indicating that they always engage in these practices. These types of

unrestricted visiting practice have been identified in a number of previous studies of

intensive care nurses and acute medical ward nurses as being important elements of

quality end-of-life care (Latour et al., 2009; Thompson et al., 2006).

Drawing curtains to provide privacy for the patient and family was also among the

most frequently reported practices by nurses in phase three of this study, with 69.1%

of participants identifying that they always undertake this practice. Such findings

were consistent with comments by participants in phase one of this study, which

indicated that interventions to maintain privacy were considered important for

patients at the end of life and their families, as well as for other critically ill patients

and their families. Providing time and space to enable families to be with their dying

relative at end of life is thus an important element of care. However, findings also

revealed that moving the patient to a more private area of the unit was among the

least frequently reported practices, with only 19.9% of participants indicating that

they always undertake this practice. Indeed, critical care unit design and/or

availability of private rooms can limit opportunities for the provision of end-of-life

care in private rooms. Lack of privacy has been identified as a barrier to good end-

of-life care in previous qualitative studies with intensive care nurses (Bloomer et al.,

in press; Calvin et al., 2009; Fridh et al., 2009; McKeown et al., 2010), and in other

studies involving critical care nurses in emergency departments (Jackson et al.,

2010). As such, while transferring patients to private rooms may not always be

possible in modern critical care units, the priority given by nurses in this study to

privacy emphasises the importance they place on modifying the environment, when

possible, to ensure quality end-of-life care.

The findings of phase three of this study also revealed more frequent engagement in

practices from the information sharing practice area, and less frequent engagement in

practices from the emotional support practice area. Specifically, information sharing

practices frequently identified as always undertaken included: introducing the

nursing staff on the oncoming shift to the patient and family (75.8%), handing over

the family’s preference for the patient’s end-of-life care to the oncoming nurse (73%)

170

and answering the family’s questions about the patient’s condition (65.8%). The type

of support for family members through information sharing that was reflected in

these responses has been identified in previous studies as contributing to good end-

of-life care (Borbasi et al., 2005; Fridh et al., 2009). In contrast, while practices

related to emotional support were used by many participants in this study, such

practices were among the practices least frequently identified as always undertaken

in the provision of end-of-life care. For example, only 18.4% of participants

identified that they always encourage the family to reminisce about the patient. This

is despite studies that have highlighted that encouraging the family to talk about the

patient can be important to humanising the patient for the nurse (Calvin et al., 2009;

Fridh et al., 2009; Hov et al., 2007). In addition, less than one-third of participants

(32.9%) always provide opportunities for the family to create positive memories.

Although few other studies have evaluated the benefits of memory-making practices

in the provision of end-of-life care in critical care settings, creating good memories

has been suggested as contributing to a good end-of-life care experience for families

(Espinosa et al., 2010; Fridh et al., 2009; Vanderspank-Wright et al., 2011).

Variation in the frequency of reported engagement in practices in the symptom

management practice area was also identified in the findings of phase three of this

study. For instance, providing care to maintain patient hygiene was a frequently

reported practice, always undertaken by 75.8% of participants. However, only 20–

24% of participants identified that they always regularly obtain and document patient

vital signs during the provision of end-of-life care, initiate (or implement) weaning

from mechanical ventilation and administer fluids to the patient to maintain

hydration. In contrast, in a survey of European intensive care nurses, 71.7% agreed

that ventilated patients should have their oxygen level reduced to room air and 74%

agreed that patients at the end of life should receive fluids to maintain hydration

(Latour et al., 2009). Other qualitative studies have also identified end-of-life care

practices in the provision of care to maintain patient hygiene (McClement & Degner,

1995) and reduce technical equipment and monitoring (Fridh et al., 2009; Halcomb

et al., 2004; Hov et al., 2007). Differences in item wording and the measurement

scale between the survey instruments and differences between the intensive care

nurse and critical care nurse samples could explain some of the variation in the

findings between studies in this field. However, it is also likely that this variation

171

reflects the general lack of clear evidence and guidance relating to symptom

management practices for critical care patients at end of life that has previously been

noted in the literature (Coombs & Long, 2008).

There are a number of explanations for the findings reported in this study regarding

the end-of-life care practices of critical care nurses. First, some authors have argued

that the ambivalent relationship that modern society has with death and dying is

reflected in the way in which health care professionals today practice end-of-life care

(McNamara, 2001; Stevens, McFarlane, & Stirling, 2000; Timmermans, 2005). For

example, the environmental modification practices frequently reported as always

undertaken in phase three of this study are important in the context of modern health

care, in which the institutionalisation of death and dying and control of these

processes by health care professionals have become increasingly common

(McNamara, 2001; Strazzari, 2005; Timmermans, 2005). Whereas once a person

would die at home surrounded by the loving care of his or her family and friends,

advances in medicine and technology during the second half of the twentieth century

have enhanced the ability to prolong life and delay death (McNamara, 2001). The

majority of deaths now occur within health care institutions, in which use of

technical interventions in the care of the patient and institutional practices such as

restricted visiting hours create a position of greater control over the dying process by

health care professionals (McNamara, 2001). In response to these trends, family

centred care during the provision of end-of-life care in critical care settings has been

identified as a ‘comprehensive ideal’ (Truog et al., 2008, p. 953). Critical care

nurses’ engagement in environmental modification practices arguably reflect

attempts to share control and support inclusion of families experiencing death and

dying in critical care settings. For example, flexibility in visiting hours and placing

chairs around the bed are modifications that can be easily undertaken by nursing staff

to support the family being with the patient at the end of life. However, while these

environmental modification practices that focus on securing privacy within the

constraints of an environment designed for curative care do reflect nurses’

appreciation of the personal and private nature of death and dying, other writers

suggest that such practices can also reflect an attempt to sequester death from the

public arena. That is, some authors have suggested that a culture of avoidance of

death exists in modern society, whereby the dying patient in critical care is hidden

172

from view to minimise disruption and the discomfort that human beings experience

when faced with death (Byock, 2002; Hart et al., 1998; McNamara, 2001; Strazzari,

2005; Timmermans, 2005). These types of contrasting explanation for the common

end-of-life care practices of nurses in critical care reflect what has been referred to as

the ambivalent location of death in modern society (Byock, 2002; Exley, 2004;

McNamara, 2001; Strazzari, 2005; Timmermans, 2005).

Secondly, environmental modification practices as described by critical care nurses

in this study are designed to create a calm, peaceful environment. Similarly, the

symptom management practices commonly described by nurses emphasise a key

focus on comfort. These goals are consistent with modern notions of a good death.

More specifically, although variation in the concept of good death has been identified

in a conceptual analysis that included both nursing and sociological literature (Kehl,

2006), a good death or quality end-of-life care as identified in empirical studies

involving critical care nurses is often associated with maintaining patient dignity

(Beckstrand et al., 2006; Halcomb et al., 2004; Kirchhoff et al., 2000). The responses

of nurses in this study suggest that provision of care to maintain patient hygiene is

perceived by critical care nurses as contributing to patient dignity at the end of life,

while other technical-medical interventions pertaining to the physical care of the

patient are not important at this time. Importantly, the social construction of death

and dying, including the labelling of death as good or bad, is to some extent

influenced by those with social control over these processes. As such, death and

dying that conforms to the rules of the health care institution and the expected norms

of behaviour from the perspective of health care professionals may not be consistent

with a good death from the perspective of the individual patient and family (Hart et

al., 1998; McNamara, 2001). This study did not explicitly seek to explore the concept

of good death and, although what a good death entails, from the perspective of

critical care nurses, is not explicitly articulated, it is arguably embedded within the

selection of items in survey development and participant responses to the survey.

It is also important to examine reasons for the findings of this study that suggest that,

while there is a high level of engagement in information sharing, emotional support

practices were less frequently reported. Interpersonal engagement in end-of-life care

can be emotionally confronting and may pose difficulties for critical care nurses

173

lacking experience in the provision of end-of-life care as well as for those frequently

exposed to death and human suffering. One possible reason for emotional support

practices being less frequently reported than information sharing practices could thus

be intentional or unintentional avoidance behaviours designed to minimise emotional

involvement and associated stress and burnout (Byock, 2002). Indeed, advances in

medicine and technology have delayed death until older age and contributed to the

institutionalisation of death and dying. These changes to the frequency and location

of death and dying have reduced opportunities for personal experience with and

learning from death and dying (Hockey, 2007; Stevens et al., 2000). Alternatively,

the argument that death is denied in contemporary Western society has been largely

contested. Instead, a view of death as socially awkward, where people are concerned

about saying or doing the right thing and thus avoiding meaningful interpersonal

interactions, has been presented (Byock, 2002; Exley, 2004; McNamara, 2001).

Participants in phase one of this research study alluded to the challenges of

interpersonal care at the end of life, referring to explaining and providing information

as a safe area of practice.

Providing information and answering questions about the patient’s condition are

aspects of practice that arguably require less emotional engagement by the nurse.

These practices, frequently noted by nurses in this study, could thus be less

confronting than other interpersonal practices related to emotional support of patients

at the end of life and their families. Emotional support practices may be especially

difficult for nurses in critical care units because patients at the end of life are often

cognitively impaired as a result of their injury/illness or administration of sedatives

used in their medical management (Truog et al., 2008). The patient’s cognitive

impairment limits nurses’ ability to establish a relationship and interact with the

patient. Related to this is the reality that critical care nurses’ preparation for end-of-

life care has to date been limited. Several authors have identified a lack of end-of-life

content in formal critical care nurse education (Campbell, 2002; Efstathiou &

Clifford, 2011; Kirchhoff & Kowalkowski, 2010). Thus, critical care nurses may

engage in information sharing practices, protecting their own vulnerability within the

emotionally confronting area of end-of-life care.

174

10.3 What are the factors influencing the provision of end-of-life care as

perceived by Australian critical care nurses?

In the previous section, critical care nurses’ reports of their end-of-life care practices

have been described, and some potential explanations for the findings of this study

have been presented. An important contribution of this study is that data were

collected to allow an examination of relationships between various contextual and

personal factors, to further understand the factors that might explain critical care

nurses’ end-of-life care practices. Specifically, the PARIHS framework provided the

framework for the current study to identify the factors that can influence the

provision of end-of-life care in critical care settings. The three key elements of the

PARIHS framework—context, evidence and facilitation—were considered to capture

the complexity of factors that can influence nursing practice. In phase three of the

current study, eight latent constructs were identified during EFA, and these

constructs conceptually fit within the three elements of the PARIHS framework. The

elements of context, evidence and facilitation will provide an organising framework

in this section to discuss key findings pertaining to the research question: What are

the factors influencing the provision of end-of-life care as perceived by Australian

critical care nurses?

10.3.1 Context

The element context, within the PARIHS framework, is defined as all aspects of the

setting in which practice takes place (McCormack et al., 2002, p. 94). Conceptual

analysis of the element of context by the authors of the PARIHS framework

identified culture, the physical environment, leadership and evaluation of

performance as factors contributing to the context of the practice setting

(McCormack et al., 2002; Rycroft-Malone et al., 2002). In phase three of the current

study, EFA of participant responses to the survey instrument identified the latent

constructs organisational culture and resources. These constructs were considered to

represent contextual factors that potentially influence the provision of end-of-life

care in critical care settings.

175

The items in the latent construct organisational culture reflected a context of nurse-

led practice during the provision of end-of-life care and did not include items

capturing the attitudes of the medical team towards end-of-life care or the broader

curative culture. In phase three of this study, 55.6% of participants agreed or strongly

agreed that less time is spent with the patient by the critical care medical team and

56% of participants agreed or strongly agreed that nurses are left to manage care of

the patient. In addition, most participants (80.1%) disagreed or strongly disagreed

with the statement after a decision is made to withdraw treatment, I spend less time

with the patient. These findings provide evidence of a practice context of nurse-led

care after a decision is made to withhold or withdraw life-sustaining treatment. The

perception that the medical team withdraw from participating in the care of the

patient following the decision to withdraw life-sustaining treatment was also shared

by participants during interviews in phase one of this study.

The findings of the current study are consistent with other qualitative studies

undertaken in Australia that have reported the perception that nurses are left to

manage the process of withdrawal of treatment and the provision of care to the

patient at the end of life (Halcomb et al., 2004; Heland, 2006). The perception that

the medical team determines treatment and critical care nurses are then left to

implement treatment decisions and manage the provision of end-of-life care has also

been reported internationally (Coombs & Long, 2008; Espinosa et al., 2010;

Kirchhoff & Kowalkowski, 2010; Seymour, 2001). In a recent survey of critical care

nurses undertaken in the United States, the presence of medical staff at the patient’s

bedside following withdrawal of life-sustaining treatment was reported to occur

sometimes by 30.3% of critical care nurses, but rarely or never by 45% of critical

care nurses (Kirchhoff & Kowalkowski, 2010).

Previous literature has highlighted the curative culture of critical care settings as a

barrier to end-of-life care (Crump et al., 2010; Nelson, 2006; Davidson et al., 2002).

In particular, the perception of death as failure can pose a barrier to end-of-life care

in the critical care setting (Beckstrand et al., 2006; McKeown et al., 2010). Although

participants in phase one of this study acknowledged the perception of death as

failure, they themselves stated that they valued their participation in the provision of

176

end-of-life care. Similarly, the perception of death as failure was not shared by

90.8% of participants in phase three of the current study. Instead, participants valued

their participation in the provision of end-of-life care.

A second important feature of the context of critical care settings that was identified

as being important to how end-of-life care was provided related to the physical

environment and resources. Specifically, resources represented the availability of

space and equipment, as well as adequate staffing to enable the provision of end-of-

life care. In phase three of this study, 60% of participants disagreed or strongly

disagreed that the physical environment of their unit is ideal for providing end-of-life

care. However, greater variation was noted in access to resources with 37.5%

agreeing or strongly agreeing that their unit is adequately equipped to support the

comfort needs of the family during end-of-life care and 47.7% disagreeing or

strongly disagreeing with this statement. The responses of some participants in the

current study were consistent with previous research that has identified the

perception that critical care and acute care environments are not adequately designed

or resourced for the provision of end-of-life care (Beckstrand et al., 2008; Bloomer et

al., in press; Davidson et al., 2003; Jackson et al., 2010; McKeown et al., 2010;

Nordgren & Olsson, 2004; Wotton et al., 2005).

An organisational culture reflecting nurse-led care at the end of life and the

availability of resources were factors contributing to the practice context of the

critical care setting and influencing the provision of end-of-life care. The broader

context of critical care, which reflects a curative culture whereby medical

practitioners withdraw from end-of-life care provision, highlights an important shift

in practice, with practice largely nurse led and more family centred than care for

patients who are being treated with curative intent. Several factors can contribute to

nurse-led practice in the provision of end-of-life care. Critical care nurses typically

have a high level of autonomy and tend to practice alongside their medical

colleagues (Seymour, 2001). Registered nurses are also generally highly skilled in

providing comfort and have a close relationship with the patient and family

(Seymour, 2001) through always being at the bedside. Such features of the critical

care context can strengthen the nurse’s position and possibly limit the participation

of their medical colleagues in situations in which comfort and support are priorities

177

of care. That is, although a one-to-one nurse patient ratio exists in critical care

settings, the critical care medical team are responsible for the care of multiple

patients and the allocation of finite resources (Seymour, 2001). Thus, following a

decision to withhold or withdraw life-sustaining treatment, the medical team may

redirect their attention to other critically ill patients to achieve the aim of critical care

in curing illness and preserving life.

Critical care environments are noisy, fast-paced environments designed for

monitoring critically ill patients and resourced to support curative care. Although

variation in participant responses regarding resource availability was identified,

nurses can use their initiative to create a context to positively influence the provision

of end-of-life care in critical care settings. Thus, it is perhaps not surprising that the

provision of end-of-life care by nurses can occur in isolation to the broader practice

context of curative care and environmental constraints.

10.3.2 Evidence

In the PARIHS framework, evidence is broadly defined to include all sources of

evidence that inform practice. In a conceptual analysis of the element of evidence,

research, clinician practice knowledge and patient’s experiences and preferences

were all identified as sources of evidence (Rycroft-Malone et al., 2002). Similarly, in

phase three of the current study, EFA of participant responses to the survey identified

three latent constructs—knowledge, preparedness and patient and family

preferences—which conceptually fit within the element of evidence as defined in the

PARIHS framework.

The items comprising the latent construct of knowledge did not assess nurses’ actual

knowledge related to end-of-life care but rather assessed opportunities for knowledge

acquisition through inclusion of end-of-life content in workplace in-services and

postgraduate nursing education, and use of evidence from the literature. In phase

three of this study, variation in response was noted for the items concerning end-of-

life content in in-service education and in undergraduate and postgraduate nursing

178

education programmes. Just over half of all participants disagreed or strongly

disagreed that they had received in-service education that assists them to provide

care and manage patient symptoms at the end of life (54.7%) or in-service education

that assists them to support and communicate with families of dying patients

(53.8%). Approximately one-third of participants disagreed or strongly disagreed that

their undergraduate nursing education included end-of-life care content (36.2%) or

that their postgraduate nursing education included end-of-life care content relevant to

the critical care context (33.6%). Similarly, Bloomer et al. (in press) conducted focus

groups across two Australian metropolitan ICUs and reported the perception of

inadequate training in the provision of end-of-life care among critical care nurse

participants. In addition, a survey of critical care nurses in the United States reported

that 44% of participants had attended an in-service on withdrawal of life-sustaining

treatment and 78.7% reported no content related to withdrawal of life-sustaining

treatment during their initial nursing education (Kirchhoff & Kowalkowski, 2010).

Lack of end-of-life content in nursing education (Campbell, 2002; Efstathiou &

Clifford, 2011) and lack of knowledge as barriers to end-of-life care (Nordgren &

Olsson, 2004) have previously been noted in the literature.

Despite identifying limited end-of-life content in in-service education and formal

nursing education, 78.9% of participants agreed or strongly agreed that they felt

adequately prepared to care for patients at the end of life. In addition, 72.6%

disagreed or strongly disagreed that they don’t know what to say to the family of a

patient at the end of life and 80.6% disagreed or strongly disagreed that they have

limited knowledge of symptom management for patients at the end of life. Thus,

participants in this study generally agreed that they are adequately prepared, despite

having limited access to formal opportunities for knowledge acquisition regarding

end-of-life care through workplace in-services and formal nursing education. It

should be noted that the majority of critical care nurses responding to the survey

were experienced critical care nurses, with 64% reporting more than 10 years of

experience in a critical care setting. Kirchhoff and Kowalkowski (2010) also reported

that almost half (49.7%) of critical care nurses in their study were very confident in

their own ability to provide end-of-life care, despite limited education and training.

However, their participants were also experienced critical care nurses (median 18

years of experience). Evidence for practice is not only obtained through education

179

and training opportunities. Experience can contribute to nurse practice knowledge

and nurses’ perceptions of their personal preparedness to provide end-of-life care.

Previous research supports this view, suggesting that actual experience in the

provision of end-of-life care contributes to nurse practice knowledge in this area

(Fridh et al., 2009; Heland, 2006; Zomorodi & Lynn, 2010b).

Patient and family preferences are also factors contributing to evidence influencing

the provision of end-of-life care. In the current study, the majority of participants

agreed or strongly agreed with statements reflecting inclusion of patient and family

preference in their unit, including families are involved in decisions about the dying

patient (90.3%), the patient’s preference (or their likely preference according to

family) for end-of-life care are considered (86%) and family meetings with the health

care team are held to resolve differing views about a patient’s treatment (87.5%). The

inclusion of patients and families in end-of-life decision making is supported in the

position statement on withholding and withdrawing treatment published by ANZICS

(2003). In addition, the inclusion of families in end-of-life decision making and the

actual provision of end-of-life care has been perceived to contribute to a good death

(Borbasi et al., 2005).

Almost all participants in phase three of this study agreed or strongly agreed that

withdrawal of life-sustaining treatment is delayed while waiting for relatives of the

patient to arrive (91%) and that families are given adequate time to consider

decisions to withhold or withdraw life-sustaining treatment for the patient (85.5%).

Delay following a decision to withdraw life-sustaining treatment was also raised by

participants during interviews in phase one of this study. Other qualitative studies

have similarly acknowledged that withdrawal of treatment can be delayed to enable

time for family to accept the change in treatment plan and/or gather at the bedside

(Bloomer et al., in press; Calvin et al., 2009; McMillen, 2008). Family presence is

associated with good death in critical care settings and families have reported

dissatisfaction with their end-of-life experience when they were not present at the

time of their relative’s death (Kjerulf, Regehr, Popova, & Baker, 2005). However,

the continuation of aggressive treatment or prolongation of inevitable death while

waiting for the family may cause unnecessary discomfort and suffering for the

180

patient (Crump et al., 2010) and thus, the needs of the patient must be balanced with

the need for family presence.

Opportunities for knowledge acquisition, personal preparedness and patient and

family preferences contribute to the element of evidence, influencing the provision of

end-of-life care for Australian critical care nurses in this study. Specifically, the lack

of education and training, including end-of-life care content, perception of

preparedness and the inclusion of patient and family in end-of-life care, can be partly

explained by the medicalisation and institutionalisation of death and dying in modern

society. The focus of critical care is on curative care, and finite resources are

distributed to achieve this aim. This curative focus is reflected in the content of

training and education, with evidence suggesting that knowledge pertaining to

curative treatment and care is prioritised and end-of-life content has historically been

minimal or absent (Campbell, 2002; Efstathiou & Clifford, 2011).

The ageing population, medical advances postponing death and the

institutionalisation of death and dying have created a situation in which experience of

death and dying will often not be encountered until later in life (Exley, 2004;

Hockey, 2007). Many nurses begin their professional lives having little or no formal

education or training in end-of-life care and no personal experience with death or

loss (Hockey, 2007; Stevens et al., 2000). Thus, they may be inadequately prepared

for the provision of end-of-life care. Experience in the provision of end-of-life care

in the workplace and reflection on this experience contribute to nurses’ practice

knowledge and perception of their personal preparedness. The majority of

participants in phase three of the current study were experienced critical care nurses,

who felt prepared for the provision of end-of-life care despite limited or no formal

education or professional development in end-of-life care.

The admission of critically ill and dying patients to acute care hospitals has resulted

in the institutionalisation of death and dying, under the control of the medical

professions (Exley, 2004; Hart et al., 1998; McNamara, 2001). Although critical care

settings are focused on curative treatment, when a decision is made to withdraw or

withhold life-sustaining treatment, critical care nurses are placed in a position in

which they obtain increased control over the dying process and death. Attitudes and

181

practices that invite patient and family participation are important in sharing control

of the processes of dying and death. For example, in the critical care setting, the

critical care team has the ability to control death and delay withdrawal of treatment

to enable time for family to gather at the bedside (Crump et al., 2010; Fridh et al.,

2009) or accept that the patient is not responding to treatment (McMillen, 2008). The

timing of removal of treatment and dying trajectory may be controlled to create a

natural death, whereby the demise of the patient is disconnected from the removal of

treatment (Seymour, 2001). Thus, death occurs naturally as opposed to as a result of

action or inaction by medical professionals directly involved (Seymour, 2001;

Timmermans, 2005). Although this position of control may be used to create a good

death, the concept of a good death has also been criticised as an implicit form of

control (Hart et al., 1998; McNamara, 2001). For instance, a good death may be

considered one in which the patient and family conform to the rules of the institution

and expected behavioural norms from the perspective of the nurses and doctors

controlling the process of death and dying. Obtaining a shared view of a good death

by seeking and supporting patient and family preferences for end-of-life care can

assist in creating a good end-of-life care experience for the patient and family

(McNamara, 2001). The findings of this study provide further evidence that the

institutionalisation of death and dying affords critical care clinicians increased social

control over the process of death and dying. This privilege needs to be understood

and appropriately used by health professionals.

10.3.3 Facilitation

A conceptual analysis of facilitation identified variation in description and use of the

term in the literature (Harvey et al., 2002). In the current study, facilitation was

broadly defined as the process of making things easier, to capture the variety of

facilitative factors that can influence the provision of end-of-life care in critical care

settings. EFA of participant responses to the survey instrument in phase three of this

study identified the latent constructs of palliative values, emotional support for

nurses and care planning. In the current study, these constructs, along with the DAS,

182

were considered facilitators of the provision of end-of-life care in critical care

settings.

Nurses’ personal attitudes and beliefs can influence the provision of end-of-life care

for patients and their families. There was strong agreement by participants in phase

three of this study with the items in the survey instrument that loaded on the latent

construct of palliative values. The items in this latent construct reflect values

consistent with a palliative approach. Almost all participants (99.5%) agreed or

strongly agreed that end-of-life care should include care of the patient’s family.

Strong agreement was also found with the items: Analgesia should be titrated to keep

the patient comfortable even if this hastens death (97%), End-of-life care is as

important as curative care in the critical care environment (93.8%), It is a privilege

to care for a patient at the end of life and their family (91%) and I can influence the

patient and family’s experience of end-of-life care (95.1%). In addition, 91% of

participants in the current study disagreed or strongly disagreed that patients at the

end of life require little nursing care. The value placed on end-of-life care and the

importance of care of the family were clearly articulated by participants during

interviews in phase one of this study. These findings indicate that participants in this

study have a strong commitment to the inclusion of families in end-of-life care and

the value of this care within the critical care setting. A shift in the focus of care at the

end of life to the family (Fridh et al., 2009; Halcomb et al., 2004) and reports of end-

of-life care being considered a privilege in the critical care setting (Calvin et al.,

2007; Fridh et al., 2009; Vanderspank-Wright et al., 2011) have been identified in

previous studies.

The items in the latent construct of emotional support for nurses captured nurses’

perceptions of the support and assistance that they receive from their colleagues and

managers in the workplace when providing end-of-life care. Although two-thirds of

participants in phase three of this study agreed or strongly agreed that, when a patient

dies in their unit, counselling is available if they need it (67.7%), more than 80% of

participants agreed or strongly agreed that, after caring for a patient at the end of life,

their colleagues will ask them if they are OK (81.1%). Similar findings were

identified in the experiences of emotional support during and following the provision

of end-of-life care by phase one participants in this study. Critical care nurses share a

183

common bond with their colleagues through the experience of caring for critically ill

and dying patients within their workplace. They may find it difficult to relate to

outsiders about these experiences (Kirchhoff et al., 2000; Shorter & Stayt, 2010).

Informal support through talking with colleagues has been widely reported in

previous research (Halcomb et al., 2004; Jones & FitzGerald, 1998; Kirchhoff et al.,

2000; Shorter & Stayt, 2010). Similarly, the value of collegial support networks and

a shared understanding of experience with death and dying among hospice nurses has

also been reported in the literature (McNamara, 2001).

Care planning can facilitate the provision of end-of-life care in critical care settings.

In the current study, the items in the care planning latent construct reflected support

for end-of-life care and agreement on a plan of care among the health care team. The

majority of participants in phase three of the current study agreed or strongly agreed

that, in their unit, all members of the health care team agree with and support end-of-

life care when it is implemented (76%) and that a plan of care for the patient is

clearly documented by the medical team after a decision to withhold or withdraw

life-sustaining treatment is made (59.7%). However, only 42.9% of participants

agreed or strongly agreed that there are policies or guidelines to assist in the delivery

of end-of-life care in their unit. In comparison, in their survey of critical care nurses

in the United States, Kirchhoff & Kowalkowski (2010) reported that the provision of

end-of-life care by critical care nurses is guided by orders from the medical team

(63.8%), care plans (20%) and standing orders (11.8%). No use of clinical practice

guidelines was reported by participants in their study (Kirchhoff & Kowalkowski,

2010). Adequate guidance in the form of a written plan of care or standardised order

forms may facilitate the provision of good end-of-life care (Truog et al., 2008),

particularly for nurses with less experience in the provision of end-of-life care in this

setting.

In the current study, participants expressed comfort with death, with 84.7% agreeing

or strongly agreeing that it doesn’t make them nervous when people talk about death

and 94.9% disagreeing or strongly disagreeing that the sight of a dead body is

horrifying to them. In addition, a significant negative correlation was found between

the categories of years of nursing experience and death anxiety. These findings

indicate that, as critical care nurses gain more professional experience, their anxiety

184

towards death decreases. Similar findings of the relationship between nursing

experience and death anxiety have been reported in the literature (Deffner & Bell,

2005).

Critical care nurses are frequently exposed to human suffering and death in their

daily practice. The provision of care to a critically ill patient in the hope of achieving

cure can result in immense suffering. Nurses’ accounts of continuing curative

treatment as torturing and flogging the patient when they perceive treatment to be

futile were reported in previous studies (Badger, 2005a; Halcomb et al., 2004;

Kirchhoff et al., 2000; McKeown et al, 2010) and in the findings of phase one of the

current study. The decision to withhold or withdraw life-sustaining treatment signals

impending death of the patient and grief and loss for the surviving family members.

Although the perception of death as a professional failure has been reported

elsewhere (Beckstrand et al., 2006; Cartwright et al., 1997; Halcomb et al., 2004;

McKeown et al., 2010), the findings of the current study support an alternate view,

with participation in the provision of end-of-life care reported by the majority of

nurses as being a privilege. It may be that the preciousness of life is raised to

conscious awareness through experiences of suffering and the inevitability of death

(Byock, 2002). Nurses are in a privileged position of accompanying the patient and

the patient’s family during their final journey in life together, and the provision of

end-of-life care presents an opportunity for critical care nurses to achieve a good

death for the patient and a good experience and memories of end-of-life care for the

family. Critical care nurses in this study believed they make a difference to the

experience of end-of-life care for patients and their families, and this could be

contributing to the nurses’ personal satisfaction with their work.

However, although death is frequently encountered by critical care nurses, the

sequestration of death to health care institutions has reduced personal opportunities

for experience with and learning from death and dying in the wider community

(Exley, 2004; Stevens et al., 2000). Nurses may refrain from discussing their

experiences of death and dying with those outside the health care institution, where

such conversations may be socially awkward, instead relying on collegial support

from those who have a shared experience with death and dying (McNamara, 2001;

Shorter & Stayt, 2010). In addition, the perception that critical care nurses are

185

expected to cope with death and dying was identified in phase one of this study and

has previously been reported in the literature (McNamara, 2001). Nurses can

maintain a resilient façade by sharing their personal experiences among their chosen

support network rather than publically acknowledging the need for emotional support

by accessing external debriefing and counselling services.

The findings of this study indicate that more exposure to death and dying through

experiences as a nurse could assist nurses to become more comfortable and

experience less anxiety towards death. Previous qualitative research has alluded to

the normalisation of death and dying in critical care settings, where nurses are

frequently exposed to patient death (Badger, 2005b; Shorter & Stayt, 2010). Critical

care nurses bear witness to suffering associated with attempts to cure critically ill

patients and, in some situations, death can be perceived as a preferential option

(Badger, 2005b). In addition, the provision of end-of-life care to a patient and the

patient’s family can achieve a good death—a description of death as a positive

experience.

In summary, the provision of end-of-life care occurs in a practice context of nurse-

led care. Nurses engage in the provision of this care, within the constraints of an

environment designed for curative care using available resources. Critical care

nurses’ agreement with items in the patient and family preferences scale indicated

their support for principles of patient and family centred care during the provision of

end-of-life care. Despite limited opportunities for knowledge acquisition, nurses

report feeling adequately prepared for the provision of end-of-life care. Critical care

nurses in this study expressed agreement with items suggesting a high value is placed

on end-of-life care and that emotional support is available through collegial support

networks. Moreover, the majority of participants expressed comfort with death and

the provision of end-of-life care in critical care settings.

186

10.4 What factors are associated with Australian critical care nurses’ end-of-

life care practices?

The third key question examined in this study sought to identify relationships

between the context, facilitation and evidence factors described in the preceding

section and the end-of-life care practices reported by critical care nurses in phase

three of this study. Separate univariate general linear models for each of the six end-

of-life care practice areas were built to identify statistically significant associations

between context, facilitation and evidence factors with each of the six practice areas.

The findings of these analyses thus identified key factors that are associated with

engagement in end-of-life care practices by critical care nurses. In summary, findings

revealed that more positive scores on scales reflecting palliative values and support

for patient and family preferences, which conceptually fit within the elements of

facilitation and evidence, were associated with more frequent engagement in the end-

of-life care practices examined in this study. Importantly, findings also revealed that

contextual factors had a limited influence as explanatory variables in the end-of-life

care practice models. The elements of the PARIHS framework are used as an

organising framework in the sections that follow, to discuss the key findings relating

to this final study question.

10.4.1 Facilitation

The latent construct of palliative values was identified as an explanatory variable

significantly associated with each of the six end-of-life care practice areas identified

in phase three of this study. The nature of this association reflected that critical care

nurses’ personal values are at the core of their practice intentions, with higher

agreement on statements reflecting values that are more consistent with the palliative

approach associated with more frequent engagement in each of the six end-of-life

care practice areas. Although positive regard for the value of end-of-life care has

been reported elsewhere (Calvin et al., 2007; Fridh et al., 2009; Vanderspank-Wright

et al., 2011), the findings that these values influence engagement in end-of-life care

practices has not previously been reported and contributes new knowledge of end-of-

187

life care in critical care settings. Thus, critical care nurses with a positive attitude

towards the value of end-of-life care and the inclusion of family are more likely to

report more frequent engagement in end-of-life care practices.

The latent constructs of emotional support for nurses, death anxiety and care

planning were also facilitation factors significantly associated with nurses’

engagement in specific areas of end-of-life care practice. Specifically, a negative

relationship was identified between the explanatory variables of emotional support

for nurses and death anxiety and the outcome variable of patient and family centred

decision making practices. Critical care nurses reporting higher scores on the

emotional support for nurses’ scale, measuring agreement with the support and

assistance they receive, and/or higher scores on the DAS, were associated with lower

levels of reported engagement in patient and family centred decision making

practices.

There are several possible explanations for these findings. In the first instance, such

findings could indicate that critical care nurses who report higher levels of perceived

support and assistance from colleagues and managers actually require more

emotional support than other critical care nurses. These nurses may thus avoid

interpersonal interactions with patients who are dying and their families. In addition,

critical care nurses who are more anxious about death may find conversations with

the family experiencing grief and loss difficult and avoid practices that require

interpersonal interactions and relationship with the patient and family, instead

focusing on other practices in which they can maintain a more superficial presence,

such as physical tasks and technical interventions. Previous qualitative studies have

reported that, for a nurse to be there at the end of life for a patient and the patient’s

family, they need to feel comfortable with being there themselves (Thompson et al.,

2006). However, other studies have reported that nurses may engage in distancing

strategies, such as focusing on technical aspects of care, as a method of avoiding

interpersonal interactions with family members (Badger, 2005a). Such avoidance

strategies may be utilised by critical care nurses frequently exposed to death to

protect their own vulnerability and minimise burnout. The relationships observed in

this study between the explanatory variables of emotional support for nurses and

death anxiety and engagement in patient and family centred decision making could

188

also be influenced by the nurses’ skills in this area. The findings that death anxiety

decreases with experience in nursing supports the notion that lack of experience with

death and dying and concomitant lack of preparation for end-of-life care can explain

decreased participation in patient and family centred decision making practices. In

addition, critical care nurses who report higher agreement with the support and

assistance from colleagues in the provision of end-of-life care may be reflecting less

confidence in their ability to provide those domains of end-of-life care that are

especially complex, such as involving families in end-of-life decision making.

A further finding of this study was that the latent construct of care planning was

significantly associated with information sharing practices, with those who reported

stronger agreement with support for end-of-life care and consensus on a plan of care

among the health care team more likely to engage in information sharing practices.

Identifying the relationship between these factors is important, as previous research

has identified lack of a plan of care (Crump et al., 2010) and lack of consensus

among the health care team and/or family as barriers to the provision of end-of-life

care (Kirchhoff & Beckstrand, 2000). Additionally, in the consensus statement on

end-of-life care by the American College of Critical Care Medicine (Truog et al.,

2008), the use of a plan of care is recommended to guide clinician practice in the

provision of end-of-life care, including interventions in the withdrawal of life-

sustaining treatment. Consensus and documentation of a plan of care can facilitate

the provision of end-of-life care and the communication of this care to families

through information sharing practices.

10.4.2 Evidence

The latent constructs of patient and family preferences, knowledge and preparedness,

which conceptually fit within the element of evidence, were associated with nurses’

engagement in select end-of-life care practices. Specifically, the stronger agreement

with items reflecting inclusion of patient and family preferences in end-of-life care

was significantly associated with higher levels of engagement in all practice areas

except information sharing practices. That is, support for patient and family

189

preferences was associated with critical care nurses’ engagement in practices to

support family presence through environmental modification, increase family

participation in decision making and support the family emotionally and spiritually.

A patient and family centred care approach to the provision of end-of-life care in

critical care settings has been recommended (Seymour, 2001; Truog et al., 2008).

Empirical studies have identified care of the family as important to quality end-of-

life care (Halcomb et al., 2004; Kirchhoff & Beckstrand, 2000; Puntillo et al., 2001),

and that good end-of-life care includes patient and family preferences for care

(Borbasi et al., 2005) and the involvement of family members in care of the patient

(Kirchhoff et al., 2000). The current study provides additional evidence that critical

care nurses’ support for patient and family centred care during the provision of end-

of-life care in critical care units can influence their engagement in practices that

share control with and support inclusion of families at this time.

This study also identified that the latent construct of knowledge was positively

associated with greater engagement in emotional support and patient and family

centred decision making practices in the provision of end-of-life care. Critical care

nurses who had received in-service education pertaining to end-of-life care,

postgraduate education including end-of-life care content and/or draw on evidence

from the literature were more likely to engage in the interpersonal practices of

providing emotional support and patient and family centred decision making. These

findings are consistent with studies that have reported that lack of knowledge is a

barrier to good end-of-life care (Nordgren & Olsson, 2004).

Importantly, over half of all participants who completed the survey instrument in

phase three had not received in-service education about end-of-life care in the

workplace and approximately one-third of participants reported that their

postgraduate education did not include end-of-life content. These reduced

opportunities for exposure to knowledge may limit critical care nurses’ evidence base

for practice, ability and/or confidence in engaging with the patient’s family on an

interpersonal level. The finding that knowledge was not an explanatory variable of

significance in the other practice models, including information sharing and

symptom management, could reflect that nurses engage in these practice areas using

knowledge obtained from educational opportunities not specific to end-of-life care or

190

from previous practice experiences. Information sharing and symptom management

practices may also not require the same level of interpersonal engagement that is

required by critical care nurses as the provision of emotional support and supporting

patient and family centred decision making.

Higher levels of agreement on the preparedness latent construct was also associated

with higher levels of engagement in end-of-life care practices related to care of the

family through information sharing, emotional support, patient and family centred

decision making and spiritual support. That is, critical care nurses who felt more

prepared to provide end-of-life care were more likely to engage in practices that

required communicating with and interpersonal care of the family. Several authors

have identified that nurses learn from experience in the provision of end-of-life care

(Fridh et al., 2009; Heland, 2006; Zomorodi & Lynn, 2010b), rather than through

propositional knowledge alone (Fridh et al., 2009). Nurses inexperienced in and

unprepared for the provision of end-of-life care may thus become overwhelmingly

concerned with saying and doing the right thing, which can result in not doing

anything (Exley, 2004; McNamara, 2001). The findings of this study provide support

for this claim.

10.4.3 Context

In this study, contextual factors had a limited influence as explanatory variables of

critical care nurses’ end-of-life care practices. The latent constructs representing

contextual factors, organisational culture and resources were significantly associated

with environmental modification practices. The organisational culture latent

construct reflected a practice context of nurse-led care and the resources latent

construct referred to the availability of space, equipment and adequate time. Thus,

the findings of this study indicate that given the nurses’ personal initiative and the

availability of resources, critical care nurses typically engage in practices to modify

the environment during the provision of end-of-life care.

191

No significant relationship was found between the contextual factors and the other

end-of-life care practice areas identified in the current study, suggesting that the

provision of end-of-life care is not greatly influenced by contextual features

measured in this study. According to the PARIHS framework, when an element is

rated as low, the influence of this element may be overcome if the other elements are

rated highly (Kitson et al., 1998). Thus, in critical care units where the context is not

satisfactory as a result of a dominant curative culture or insufficient resources,

nurses’ engagement in positive end-of-life care practices could be achieved if

evidence and facilitation factors that influence practice are present. For example, the

findings of this study could suggest that evidence (including patient and family

preferences) and facilitation (such as the personal values of the individual nurse)

factors are most important in influencing critical care nurses’ engagement in end-of-

life care practice, while the broader context could be less significant in the intimate

setting of end-of-life care. That is, the provision of nursing care is greater than just

the physical context in which it is provided, and nurses can engage in positive end-

of-life care practices regardless of the broader practice context of critical care. A

large number of the end-of-life care practices identified in this study consist of items

related to the interpersonal care of the family and communication between the nurse

and the family, and the broader practice context could have less of a bearing on such

practices than the values of the individual nurses themselves and the preferences of

the patient and family at this time.

10.4.4 Demographic explanatory variables

The demographic variables of gender, nursing experience, postgraduate

qualifications, work location and workplace each contributed to at least one end-of-

life care practice model as a significant explanatory variable. The findings of this

study indicate that critical care nurse engagement in the interpersonal caring

practices of emotional support and spiritual support differed according to gender,

with male critical care nurses being more likely to report less frequent engagement in

these practice areas. Nilsson and Larsson (2005) identified the perception that male

nurses have a more technical focus than female nurses in the findings of their study.

192

Thus, this argument would suggest that male nurses could be attracted to the

medicalised and technologically focused practice of critical care. Indeed, workforce

statistics demonstrate that a higher proportion of male nurses are employed in critical

care areas than in other areas of nursing practice (AIHW, 2011). Moreover, in

Western society, several authors argue that males are socialised to be strong and

maintain emotional composure (Brown, 2009; Silverberg, 1985). Kellehear (2000),

for example, notes the masculinity of the Australian culture is such that grief is

expressed by women, and not by men. Such views thus propose that socialisation

does influence the practices in which critical care nurses will engage when faced

with caring for people who are dying. In contrast, Rosenberg (2009) suggests that

traditional images of masculinity in Australian society are changing and new images

are emerging with males engaging in secular rituals to express grief and loss, while

retaining their masculinity. The role of gender in responses to grief and distress, and

in the way in which male and female nurses’ practise their discipline, requires further

investigation. It is possible that male nurses support the emotional and spiritual needs

of patients at the end of life and their families by engaging in practices not captured

in the survey instrument.

Nursing experience and postgraduate qualification were both significant explanatory

variables of spiritual support practices. Postgraduate qualifications were also a

significant explanatory variable of patient and family centred decision making

practices. Specifically, those with higher levels of education were more likely to

engage in practices in both of these domains, suggesting that education and

experience can increase nurses’ ability to engage in these interpersonal caring

practices. However, critical care education and critical care experience were not

significant explanatory variables in any of the end-of-life care practice models. The

limited inclusion of end-of-life content in critical care curricula and workplace in-

service programmes identified by participants in this study could contribute to the

limited influence of these variables to end-of-life care practices.

Work location was significantly associated with environmental modification

practices, with critical care nurses employed in regional or remote areas reporting

more frequent engagement in environmental modification practices than critical care

nurses working in major cities. The size or workload of the critical care units in

193

regional and remote areas may be different from those in major cities, and such areas

may be more accommodating or flexible in how care is organised. Therefore, such

contextual variables may be important in determining the nurses’ abilities to engage

in environmental modification practices.

Similarly, the demographic explanatory variable of workplace was associated with

emotional support practices and symptom management practices, with employment

in an emergency department associated with less frequent engagement in the

emotional support and symptom management practice areas. Emergency departments

provide a different context from other critical care practice settings such as ICUs. Le

Conte et al. (2010) suggest that physicians in emergency departments have limited

time to make decisions, lack previous knowledge of patients’ prior health status and

lack long-term relationships with the patients. Nurses in emergency departments also

have different nurse-patient ratios from those of their nursing colleagues in intensive

care, and patients are in the emergency department for a shorter period of time,

precluding them from forming closer relationships with the patient and family.

Beckstrand et al. (2008) agree, acknowledging that while the items in their study

related to communicating with the patient and family regarding an anticipated death

and talking with patients about their feelings occurred in ICUs, they did not

frequently occurr in emergency departments. These authors suggest that this is

because the emergency department is a place of transition with inadequate time for

nurses to form relationships with patients (Beckstrand et al., 2008). Indeed, in the

current study, the items in the survey instrument pertaining to emotional support and

symptom management practices may not be as applicable in emergency department

settings. This is not surprising given the diversity of patient diagnoses and illness

trajectories between emergency departments and other critical care settings. Instead,

emergency department nurses may engage in other practices not captured by items in

the survey instrument to provide emotional support and symptom management.

194

10.5 Strengths and limitations

There are some important strengths and limitations of the study that should be

considered at this point. First, in the current study, end-of-life care was defined as the

phase of care that occurs following a decision to withdraw or withhold life-sustaining

treatment, with death of the patient the expected outcome. Although this definition

was consistent with current practice and provided a focus for the study that was

considered within the scope of a PhD research project, the view that palliative care

is applicable only in the last days of life is a barrier to the broader role of palliative

care for all patients in the critical care setting (Nelson & Meier, 1999; Rodriguez et

al., 2007, p. 107). Further research to explore the earlier application of a palliative

approach in critical care settings is needed.

Additionally, while the PARIHS framework was used to guide this study, some

strengths and criticisms of the framework and their implications for the present study

should also be considered. First, the PARIHS framework was originally developed to

represent the complexity of factors influencing successful implementation of

research evidence into practice. However, the elements of context, evidence and

facilitation identified in the framework were considered to represent the complexity

of factors that can also influence nursing practice. Thus, the PARIHS framework was

selected as an organising framework to identify the factors that may influence critical

care nurses’ end-of-life care practices.

Nevertheless, a major criticism of the PARIHS framework is the conceptual overlap

between subelements and the lack of clarity regarding the relationships between the

elements and subelements (Helfrich et al., 2010). The subelements in the PARIHS

framework were identified through conceptual analysis of each element (Harvey et

al., 2002; McCormack et al., 2002; Rycroft-Malone et al., 2002). Although the key

elements of the framework were thought to capture the complexity of factors

influencing critical care nurses’ end-of-life care practices, some of the subelements

were not considered relevant to the provision of end-of-life care by critical care

nurses. Instead, a range of factors influencing practice for each key element was

proposed a priori. In addition, the eight factors influencing the provision of end-of-

life care identified through EFA in phase three of this study were considered to have

195

conceptual and theoretical fit within the three key elements of the PARIHS

framework. The current study also considered the relationship between each factor

and critical care nurse engagement in end-of-life care practices. Thus, the current

study contributes to nursing knowledge by systematically documenting critical care

nurses’ end-of-life care practices, and identifying facilitation, evidence and

contextual factors influencing engagement in these practices. The findings of this

study revealed that factors within the elements of facilitation and evidence were

associated with more frequent engagement in end-of-life care practices, while

contextual factors had a limited influence. Moreover, in the PARIHS framework, the

element of facilitation is defined as a role (Helfrich et al., 2010). However, a broader

definition of facilitation was utilised in this study to capture the range of facilitative

factors, including individual values, beliefs and attitudes, which may influence the

provision of end-of-life care by critical care nurses.

There are also a number of methodological strengths and limitations associated with

the current study that should be noted. Despite an extensive search of the literature,

no validated tools were identified that could be used to address the research purpose

and research questions in this study. Although other survey instruments to measure

end-of-life care in critical care settings have been developed since the current study’s

inception (Kinoshita & Miyashita, 2011; Zomorodi & Lynn, 2010a), the survey

instrument developed in the current study includes items to identify nurses’ self-

report of frequency of actual end-of-life care practices and their agreement with

factors that can influence end-of-life care. A rigorous process of instrument

development was undertaken in the current study. The development of a robust

survey instrument that is reliable and valid is a time-consuming process. However,

investment of time in developing the instrument by conceptualising the domains of

interest and generating items contributes to the content validity of the instrument

(Polit & Beck, 2006; Polit et al., 2007).

Review of the survey instrument by an expert panel and collection of relevance

ratings by experts for each item in the survey instrument permitted the calculation

and assessment of the content validity of individual items (I-CVI) as well as the

entire instrument (S-CVI) (Polit et al., 2007). Content validity index (CVI) is widely

used by nurse researchers and is easily calculated (Beckstead, 2009; Polit et al.,

196

2007). Another strength of the CVI is that it measures agreement amongst experts

regarding the relevance of items as opposed to the internal consistency between

ratings provided (Polit et al., 2007). It has been noted in the literature that

computation methods of CVI do not account for chance agreements (Beckstead,

2009; Lynn, 1986; Polit & Beck, 2006; Polit et al., 2007). However Polit et al. (2007)

note that members of the expert panel are selected for their expertise and it is

anticipated that they provide considered responses regarding the relevance of each

item, not a random response. In the current study, six experts were included on the

panel ensuring an adequate sample size to appropriately identify items as relevant

(Lynn, 1986; Polit & Beck, 2006). The collapse of relevance rating categories into

dichotomous categories during calculation of CVI has been criticised for loss of

information (Beckstead, 2009; Polit & Beck, 2006; Polit et al., 2007). However this

method of data management eases computation of CVI and the initial data collected

on the 4-point rating scale can be reviewed for identification of items that may need

revision (Polit et al., 2007). Lastly, Polit and Beck (2006) identify a lack of detail in

nursing literature regarding computational methods used in CVI obscuring the results

provided. Consistent with Polit and Beck’s (2006) recommendations, the methods

used to calculate both I-CVI and S-CVI have been clearly articulated in this thesis.

In addition to review of the survey instrument by an expert panel, completion of a

pilot study enabled preliminary assessment of the psychometric properties of the

survey instrument. The pilot test also tested the feasibility of some of the methods

used in phase three of this study, including data collection, management and analysis.

The results of the expert review and pilot study informed revisions to the survey

instrument and study methods.

Regression models are derived from the variables included in the analysis (Hair et

al., 2006). The latent constructs identified following EFA of participant responses to

the survey instrument in the current study are derived from the items included in the

content of the survey instrument. Practice is a complex, multidimensional

phenomenon and only a limited number of items could be included in the survey

instrument. Therefore, there is a potential for selection bias influencing the

identification of latent constructs. Although a rigorous process of item selection,

instrument development and EFA was undertaken to create the summated scales

197

included in the multiple regression model, it is possible that aspects of practice or

factors influencing practice may not be captured within the models generated in this

study, potentially biasing the models and regression coefficients obtained (Hair et al.,

2006). Additional latent constructs of end-of-life care practices and factors

influencing these practices may exist and may not have been identified. For example,

the items encourage the family to touch the patient and encourage the family to talk

to the patient loaded together during EFA, but were not accepted as a factor because

no other items loaded on the factor. The items are conceptually similar and perhaps

represent a latent construct not adequately covered by the items in the instrument.

Convenience samples were utilised in all phases of this study, and the strengths and

limitations of this sampling approach as well as the samples obtained must be

acknowledged and considered in the interpretation of the findings of the study.

Convenience samples are a time-efficient method of recruiting potential participants.

However, participant self-selection may potentially bias the sample because those

that choose to participate may be characteristically different from those who choose

not to participate. Thus, a convenience sample may not reflect the wider population

that they are recruited to represent (Schneider, Elliott, LoBiondo-Wood, & Haber,

2003).

Although phase one of this study used a small sample of participants, the aim of

phase one was not to generalise the findings, but to provide evidence of the current

experience and practices of critical care nurses in the provision of end-of-life care, to

inform the development of a survey instrument of critical care nurses’ end-of-life

care practices and factors influencing these practices. Thus, the sampling strategy

was deemed appropriate for the purpose of this phase of the study. However, the

sample was limited to critical care nurses working in one ICU. Inclusion of critical

care nurses employed in other settings, such as emergency departments, could

provide other insights into critical care nurses’ end-of-life care experiences and

practices not identified in phase one of this study. The variation noted in participant

responses to items in the emotional support and symptom management practice areas

according to workplace may not reflect a lack of engagement in these practices, but

rather engagement in other practices not captured in the current survey instrument.

198

Comparison of the demographic profile of participants in phase three of this study

with the demographic information available from the Nursing and Midwifery Labour

Force 2009 (AIHW, 2011) survey and the Intensive Care Resources & Activity:

Australia & New Zealand 2007/2008 (Drennan et al., 2010) report assisted in

identifying similarities and differences between the participants in phase three of this

study and the national intensive care nurse and registered nurse populations. In

addition, comprehensive description of the sample assists in determining the

applicability and generalisability of the findings to other populations, and identifying

areas for further research using new samples. The average age of participants in

phase three of this study was similar to the national average (43.1 years v. 44.3

years) (AIHW, 2011). Although the proportion of males in this survey was larger

than the national proportion of males registered as a nurse (13.5% v. 9.6%), it was

consistent with current workforce figures of males employed in critical care and

emergency settings (13.3%) (AIHW, 2011). The phase three sample had a high

proportion of participants (90.3%) who had completed or were currently undertaking

formal education in critical care. These figures are much larger than the proportion of

registered nurses with critical care qualifications (56.1%) in the Intensive Care

Resources & Activity: Australia & New Zealand 2007/2008 (Drennan et al., 2010)

report. In addition, in the current study, only 55.1% of participants identified clinical

work as their primary role compared with the findings of the Nursing and Midwifery

Labour Force 2009 (AIHW, 2011) survey in which 77.7% of nurses identify their

main nursing job as a clinical position. These variations in critical care qualifications

and primary nursing role of participants in the current study compared with the

national workforce statistics for general and critical care nurses (AIHW, 2011;

Drennan et al., 2010) is likely to reflect membership of the professional association

for critical care nurses, whose members are more interested in professional issues

and education than non-members.

Interestingly, although only 55.1% of participants in the current study identified

clinical work as their primary role, 74.5% of participants reported providing direct

patient care within the last week, indicating that most of the sample is currently

engaged in clinical work. In addition, of participants that had provided direct patient

care in the past 12 months, 95.5% reported providing end-of-life care. Thus, the

majority of participants had recent experience in the provision of end-of-life care in

199

their workplace and their responses to the survey are less likely to be influenced by

recall bias.

Overall, there were some similarities and differences between the demographic

profile of critical care nurse participants in phase three of this study and the available

comparison population of Australian intensive care nurses (Drennan et al., 2010) that

need to be considered in interpreting the results of this study. However, the

differences may be insignificant as few demographic variables were identified as

explanatory variables in the end-of-life care practice models. For example, although

variation between the sample and comparable population was identified in regard to

formal education in critical care, this variable was not a significant explanatory

variable in any of the end-of-life care practice models.

In this study, data were collected via interviews in phase one and a web-based survey

in phases two and three. It has been acknowledged that an interview is a social

interaction and, as such, the researcher as interviewer contributes to and influences

the interaction (data collection) and the interpretation of the interaction (data

analysis) (Fontana & Frey, 2005). The interview is a mutually created, contextually

based account between the interviewer and interviewee (Fontana & Frey, 2005;

Silverman, 2006). Thus, the researcher cannot claim to be impartial or unbiased. The

researcher is a former employee of the study setting. The capacity of this

employment was as a registered nurse providing direct patient care. The position of

the researcher as a former registered nurse in the ICU assists in understanding the

context of the participants’ work. The researcher was aware of personal experience

as a registered nurse in intensive care and, when it was recognised that an assumption

was being made, a question was asked to clarify the meaning for the participant. In

addition, check-coding of transcripts by the supervisory team and presenting the

findings back to nurses in the study setting were specific measures undertaken to

enhance the dependability of the findings (Miles & Huberman, 1994).

Web-based surveys are both time- and cost-efficient methods of collecting data from

large numbers of participants. Web surveys are also time effective in eliminating data

entry requirements through the ability to export data from the survey software

directly into SPSS for data analysis (Sue & Ritter, 2007). Unfortunately, the exact

200

response rate to the survey in phase three of this study could not be ascertained. Key

Survey version 7.3 was used in phase three of this study and the rules specified

during the development of the survey did not allow incomplete survey responses to

be recorded. Thus, the number of critical care nurses who accessed the survey or

only partially completed the survey is unknown. In addition, due to privacy

restrictions, the researcher was unable to track undelivered emails. However,

responses to the survey were maximised by careful consideration of the wording of

recruitment materials and the use of a reminder email to encourage participation. The

estimated response rate was 25%. This was comparable to a recent national postal

survey of pain assessment and management by Canadian critical care nurses in which

a response rate of only 24.5% was achieved despite three rounds of contact with

participants that included an initial invitation and two reminders (Luk & Rose, 2011).

Specific factors that may have contributed to the low response rate in this research

study include the length of the survey instrument and the topic under investigation.

The survey instrument was approximately four pages in length, and some

participants in the pilot test in phase two commented that the survey was quite

lengthy. However, the length of the survey instrument was necessary to permit

adequate coverage of the variables of interest to the study. End-of-life care may be a

confronting and emotive topic of nursing practice. This topic may engender interest

and participation for some and, conversely, may reduce participation by others.

A considered and rigorous approach to data analysis was maintained throughout this

study. The backward elimination approach to regression model building utilised in

phase three has been criticised for the potential to obtain a model that fits the sample

data rather than the broader population, thus limiting generalisability (Tabachnick &

Fidell, 2007). However, this approach is considered satisfactory for the purpose of

identifying the predictors of critical care nurses’ end-of-life care practices. Although

other plausible results may be obtained through alternate methods of data analysis,

the findings of this study were considered to have conceptual and/or theoretical

resonance. Replication of the current study using a new sample and CFA would

provide additional evidence of the reliability and validity of the survey instrument.

201

The current study only considered the self-reported practices and views of critical

care nurses. Use of alternate data collection approaches may obtain differing results.

For example, analysis of actual practice through ethnographic methods may uncover

hidden practices not articulated or readily identified by critical care nurses in the

provision of end-of-life care. Overall, the findings of this study have contributed new

insight into critical care nurses’ engagement in the provision of end-of-life care and

new knowledge of the factors associated with nurse engagement in end-of-life care

practices, as well as preliminary evidence of the psychometric properties of a new

survey instrument.

10.6 Summary

In summary, practices related to environmental modification and information sharing

were the most frequently reported end-of-life care practices of Australian critical care

nurses in this study. Environmental modification and information sharing practices

are simple interventions for nurses to undertake in the provision of end-of-life care.

Practices related to the emotional support of patients at the end of life and their

relatives were the least frequently reported as always undertaken in the provision of

end-of-life care. The provision of end-of-life care may be emotionally confronting

for nurses, creating difficulties for interpersonal engagement of nurses with patients

at the end of life and their families.

The findings of this study revealed that the provision of end-of-life care is influenced

by the personal values of critical care nurses, with those reporting values consistent

with the palliative approach engaging in positive end-of-life care practices. The

latent construct of patient and family preferences was also a significant explanatory

in five of the six end-of-life care practice areas. Critical care nurses perceived that

they were prepared to provide end-of-life care despite limited opportunities for

knowledge acquisition. Knowledge and preparedness were associated with critical

care nurses engaging in the interpersonal practices of patient and family centred

decision making and emotional support of the patient and family. In contrast, critical

care nurses requiring emotional support and those with higher death anxiety reported

202

lower engagement practices pertaining to care of the family. Contextual factors had

limited influence as explanatory variables of end-of-life care practices areas.

However, a context of nurse-led practice was identified, with critical care nurses

drawing on their personal practice knowledge and initiative to construct a good

experience of end-of-life care for the patient and surviving family members.

A rigorous research approach and careful consideration of the strengths and

limitations of this study have been undertaken. The findings of this study contribute

to knowledge of critical care nurses’ end-of-life care practices. In addition, the

findings of this study identified variation in critical care nurse engagement in select

practices and the factors that contribute to explaining this variation.

203

11 RECOMMENDATIONS AND CONCLUSIONS

11.1 Introduction

The findings of this three-phase exploratory sequential mixed methods study have

informed the development of a number of recommendations for practice, education

and further research to improve care of patients at the end of life and their families,

and support critical care nurses in the provision of this care. Overall, the findings of

this study provide new insight into the end-of-life care practices of critical care

nurses and contribute new knowledge of the specific factors associated with nurse

engagement in these practices.

11.2 Recommendations for practice

Six end-of-life care practice areas were identified in this study, with variation noted

in the frequency with which critical care nurses engage in end-of-life care practices.

The identification of strategies to both assist critical care nurses to engage in the

provision of end-of-life care and support them in this challenging role are essential in

providing optimal care to patients at the end of life and their families and

maintaining the emotional health and wellbeing of critical care nurses.

With less than half of all phase three participants reporting the availability of

guidelines or policy related to the provision of end-of-life care and the variation

identified in critical care nurse engagement in end-of-life care practices, it is

recommended that guidelines be developed that address each of the six practice areas

identified in this study. These end-of-life care guidelines could raise awareness of

interventions that could be implemented from all practice areas. In addition,

guidelines could assist less experienced and/or prompt busy nursing staff to provide

comprehensive nurse-led care to patients at the end of life and their families (Clarke

et al., 2004; Rose & Gidman, 2010). For example, inclusion of interventions to

provide emotional support to patients at the end of life and their families in end-of-

204

life care guidelines could prompt increased engagement in these practices. However,

it is important to individualise care to meet the needs and preferences of the patients

and their families. Thus, guidelines must not be implemented in a prescriptive

manner whereby patients and families are expected to comply (Rose & Gidman,

2010). Instead, these guidelines should provide a platform for discussion of care

needs and preferences between the critical care nurse and the patient and family,

facilitating patient and family centred care.

Patient and family centred care during the provision of end-of-life care in critical

care settings has been recommended (Seymour, 2001; Truog et al., 2008). In the

current study, nurses with higher levels of agreement on the patient and family

preferences scale more frequently engaged in five out of the six end-of-life care

practice areas. Therefore, it is recommended that strategies to assist critical care

nurses to identify and support patient and family preferences could increase

engagement in end-of-life care practices. For example, following admission to a

critical care unit, early and ongoing regular communication is needed between the

patient, family and all members of the health care team regarding the patient’s

condition, possible outcomes and preferences for current and future care . These

discussions can smooth transitions if treatment plans change and assist in formulating

a shared position of what constitutes good end-of-life care (Boyle, Miller, & Forbes-

Thompson, 2005). Use of a family diary could also facilitate communication between

the family and health care team and across the health care team. In addition, within

individual critical care units, standardisation of documentation of patient and family

preferences on a patient care plan or flow chart would facilitate continuity of care

incorporating patient and family preferences between care providers.

It is recommended that palliative care nurses and/or critical care nurses with

expertise in the provision of end-of-life care be identified and available to provide

consultation to critical care nurses engaged in the provision of end-of-life care. These

expert nurses could share their practice knowledge and experiences and act as role

models in the practice of providing end-of-life care, promoting positive end-of-life

care attitudes and values among critical care nurses, and supporting less experienced

nurses in this care. The inclusion of palliative care consults in the critical care setting

has been reported in the literature (Campbell & Guzman, 2003; Hansen et al., 2009;

205

O’Mahony et al., 2010). Evidence of the effectiveness of this intervention is limited

to reductions in time between identification of poor prognosis and implementation of

end-of-life care and decreased use of unnecessary treatment interventions, with

further research needed to ascertain improvements in overall quality of care.

The findings of this study indicate that environmental modification practices are

simple interventions for nurses to implement given the nurses’ personal initiative and

the availability of resources, including time and equipment. Variation in the

availability of resources for end-of-life care was also noted in this study. Although

critical care environments are designed for optimal surveillance of critically ill

patients and resourced to support curative care, strategies to optimise patient privacy

and the availability of appropriate resources to support the comfort of patients at the

end of life and their families are needed for the significant numbers of patients that

die in these settings. It is recommended that the availability of facilities for the

provision of end-of-life care, such as private rooms and family rooms, be considered

in the planning of new and redevelopment of existing critical care units. Lack of

privacy in critical care settings is a well-documented barrier to end-of-life care

(Bloomer et al., in press; Calvin et al., 2009; Fridh et al., 2009; Jackson et al., 2010;

McKeown et al., 2010). In addition, it is recommended that critical care units

develop resource kits to support patient and family comfort. Participants in phase one

of this study described a resource kit, developed by nursing staff, that contained

linen, lamps and cushions available for use during the provision of end-of-life care.

Hansen et al. (2009, p. 266) also reported use of a resource kit that included a quilt,

music compact discs, an envelope for a lock of hair and materials for creating a hand

print. These kits could also contain information for families such as bereavement

materials and available support services.

It is essential that emotional support be provided to critical care nurses experiencing

death and loss in the workplace. Without appropriate support, nurses may experience

stress and burnout (Shorter & Stayt, 2010). In particular, strategies to foster collegial

support networks are recommended as essential in critical care settings. The

importance of collegial support networks was identified by participants during phase

one of the current study. The value of these networks was further emphasised, with

the majority of participants in the national survey agreeing with items related to

206

collegial support. Strategies to strengthen these informal support networks, relied

upon by so many critical care nurses for their emotional support needs, are essential

to ensure that the stress associated with the provision of end-of-life care does not

accumulate and lead to burnout. Strategies to formalise collegial support networks

could include rostering patterns and support for regular shared meal breaks to

support existing networks between colleagues. In addition, the development of buddy

systems for new and less experienced critical care nurses with more experienced

critical care nurses can provide interim support for these nurses until they establish

their own collegial support network in the workplace.

11.3 Recommendations for education

In this study, critical care nurses reporting higher agreement for values consistent

with a palliative approach were more likely to engage in all end-of-life care practice

areas. If critical care nurses’ personal values are at the core of their practice

intentions, then strategies to support the development of positive beliefs and attitudes

about the provision of end-of-life care can improve practice. Therefore, inclusion of

end-of-life content in critical care curricula and professional development activities

using a values-based education approach is recommended. Values-based education

and practice was originally developed and used in mental health practice (Fulford,

2011; McLean, 2012). The potential applicability of values-based practice in other

areas of health care has been acknowledged (Fulford, 2011). Values-based practice

recognises the centrality of the patient and family in care and aims to create

partnerships between health care providers and the patient and family through shared

understanding of the diversity of values present at any given time (McCarthy &

Grumley, 2010). Values-based education aims to develop awareness of one’s own

values through activities that encourage personal reflection, skills to identify and

consider the values of others and communication skills that can be used to support

participation in decision making and manage conflict (Fulford, 2011). A values-

based education approach could support the development of positive values towards

end-of-life care and communication skills, contributing to critical care nurses’

engagement in end-of-life care practices.

207

The findings of the current study suggest that the provision of end-of-life care in the

context of critical care setting is largely nurse led. Yet, limited opportunities for

acquisition of knowledge to support the provision of end-of-life care were identified

by participants in this study. The frequency with which end-of-life care is provided

in critical care settings dictates that opportunities for education and discussion of

end-of-life care issues be provided. The inclusion of end-of-life care content in

critical care curricula and professional development activities is needed to address

the complexity of this area of practice, including the care and inclusion of family, the

physical care and emotional support of the patient and self-care for the nurse. A suite

of resources, similar to those developed for the Palliative Care Curriculum for

Undergraduates (PCC4U) project, that includes case studies and experiential learning

activities could support integration of end-of-life content in critical care curricula

(PCC4U, 2011).

Inclusion of in-services within the workplace and the opportunity for attendance at

workshops specifically designed to address the challenges and opportunities for the

provision of end-of-life care in critical care settings is needed. For example, the

findings of the current study indicated the centrality of the family to the provision of

end-of-life care, yet items related to emotional support of the patient and family were

practices least frequently identified as always undertaken by critical care nurses. In

addition, opportunities for knowledge acquisition and preparedness were associated

with critical care nurses’ engagement in interpersonal caring practices. Thus,

strategies to support nurses to gain experience in the provision of end-of-life care

may enhance nurses’ preparation and knowledge for practice and increase their

engagement in interpersonal care practices. For example, emotional support practices

may be enhanced through experiential learning strategies, such as facilitated practice

development workshops in which critical care nurses have an opportunity to practice

communication skills and role play difficult conversations that could be considered

socially awkward in other contexts. Practising these interventions in a supportive

environment may increase critical care nurses’ confidence and future engagement in

emotional support practices.

208

Specialist palliative care nurses could also be invited into critical care settings to

share their knowledge and experiences, which may be translated into critical care

practice. Critical care nurses could also participate in supernumerary clinical

experiences in palliative care or hospice settings to increase their experience with

death and dying. For example, the Program of Experience in the Palliative Approach

(PEPA) provides opportunities for nurses to attend workshops and gain professional

experience in a palliative approach through supervised clinical placement

experiences in palliative care settings (PEPA, 2003). Experiences gained in palliative

care settings may be translated into practice in the critical care context. Inclusion of

end-of-life content in both formal education and professional development may raise

awareness of the value and importance of this area of practice and contribute to

nurses’ practice knowledge and perception of their personal preparedness.

In addition, critical care nurses should be encouraged to share their personal

experiences of providing end-of-life care to raise awareness of this area of practice

and the value and contribution that can be made to patients at the end of life and their

families within their own workplace. Opportunities for critical care nurses to share

their experiences of end-of-life care with the broader community should also be

sought. These personal narratives of critical care nurses may facilitate wider

acceptance and meaning of death and dying in critical care settings and wider society

(Byock, 2002; Hart et al., 1998).

11.4 Recommendations for research

The palliative values expressed by critical care nurses in this study are applicable in

the management of all critically ill patients and not just those patients identified as at

the end of life. Further research to ascertain the attitudes of critical care nurses and

investigate the potential for earlier integration of a palliative approach in the

management of critically ill patients would make an important contribution to

improving knowledge of this area of nursing practice in critical care settings. Earlier

inclusion of a palliative approach could enhance care for patients experiencing

209

critical illness and smooth transitions when a decision to withhold or withdraw life-

sustaining treatment is made (NHMRC, 2011; Stroud, 2002).

Evaluative studies of values-based education should be undertaken to obtain

evidence of the effectiveness of this educational approach in supporting the

development of positive values towards end-of-life care among critical care nurses

and improving the quality of care experienced by patients at the end of life and their

families. In addition, content mapping of end-of-life care in critical care curricula in

higher education institutions can assist in appropriately targeting the development of

resources needed to integrate end-of-life content in curricula. In addition, further

research to explore critical care nurses’ perception of their preparedness for the

provision of end-of-life care and the factors that contribute to this preparedness is

needed. This research may inform the development of strategies to prepare nurses for

the provision of end-of-life care in critical care and other settings.

Although the findings of this study suggest nurse-led care, critical care nurses

practice within a broader multidisciplinary team. Further research to identify the role

and contribution of other members of the multidisciplinary team from the perspective

of each of the different professions is needed to obtain a more complete picture of

this area of practice in critical care settings. In addition, the perspective of patients

and families is important, and further research should consider the patient and family

perception of what constitutes good death. The experience of end-of-life care from

the family’s perspective could provide new insight and identify areas for practice

development.

Engagement in practices related to the emotional support of the patient and family

and symptom management were also found to vary across the different critical care

workplaces. Critical care is a broad term encompassing diverse settings of nursing

practice. Further in-depth research in specific areas of critical care, such as intensive

care and emergency departments, may contribute to nursing knowledge of end-of-life

care in the different critical care workplaces in which end-of-life care is provided.

For example, research into the emotional support practices utilised by emergency

nurses in the provision of end-of-life care in emergency departments would enhance

knowledge of this area of practice. In addition, the experience of end-of-life care in

210

emergency departments for bereaved families would assist in identifying areas for

practice improvement.

In addition, further research to explore the finding that males differ in their

engagement in interpersonal caring practices is needed. The perspective of recipients

of this care, such as bereaved family members could be assessed to investigate

whether emotional and spiritual support needs are adequately addressed. In addition,

the emotional support needs of male critical care nurses and strategies required to

address these needs may be different from those of female critical care nurses.

The survey instrument was specifically developed to address the research questions

posed in the current study. Preliminary evidence of the reliability and validity of the

survey instrument has been obtained. Further development and testing of the survey

instrument is required. Specifically, the symptom management scale requires further

development. Although the internal consistency of the items in the symptom

management scale was deemed satisfactory in pilot testing, varied feedback on this

scale was obtained from the expert panel and multivariable analysis revealed low

variation explained by predictors in the regression model. Perhaps this is a difficult

area to assess using a self-report survey instrument. Other methods of research such

as chart audit or completion of a checklist style instrument immediately post

provision of end-of-life care may eliminate recall biases and provide more insight

into the symptom management practices of critical care nurses. In addition, further

testing of the entire survey instrument using CFA or a structural equation modelling

approach in a new sample would provide more evidence of the reliability and validity

of this new tool.

11.5 Conclusions

This study has contributed new insights into critical care nurses’ engagement in end-

of-life care practices. Importantly, this study has also contributed new knowledge of

the factors associated with critical nurse engagement in these practices. Critical care

nurses engage in many important practices in the provision of end-of-life care.

211

However, variation was noted in critical care nurses’ engagement in some practice

areas and key factors associated with this variation were identified.

First, critical care nurses’ responses to the survey reflected values consistent with a

palliative approach. These palliative values were associated with engagement in all

end-of-life care practice areas identified in this study. Thus, the personal values of

the critical care nurse were at the core of their practice intentions. Development of

positive attitudes and values among nurses towards end-of-life care in critical care

settings can thus contribute to quality care for patients at the end of life and their

families.

Consistent with recommendations in the literature (Seymour, 2001; Truog et al.,

2008), the responses of participants in this study indicate agreement with and support

for patient and family centred care. In addition, higher agreement on items regarding

patient and family preferences was associated with critical care nurse engagement in

practices that share control with and support inclusion of families. Strategies that

assist critical care nurses to identify and support patient and family preferences will

contribute to increased participation in the end-of-life care practice areas identified in

this study.

Despite limited opportunities for knowledge acquisition, critical care nurses felt

adequately prepared for the provision of end-of-life care. These factors were

associated with variation in critical care nurses’ engagement in interpersonal caring

practices such as those pertaining to patient and family centred decision making and

emotional support for patients and families. In addition, critical care nurses reporting

higher scores on the emotional support for nurses’ scale, measuring agreement with

the support and assistance they receive, and/or higher scores on the DAS were

associated with lower levels of reported engagement in patient and family centred

decision making practices. Collectively, these findings indicate that critical care

nurses who have limited opportunities for knowledge acquisition, who perceive that

they are not prepared, and/or who are less comfortable with the provision of end-of-

life care and death report lower engagement in end-of-life care practices that reflect

interpersonal care.

212

In addition, it is argued that the practices most frequently identified as always

undertaken in the provision of end-of-life care, specifically, information sharing and

environmental modification practices, require less emotional engagement by the

nurse and are less confronting than other interpersonal caring practices, such as the

emotional support practices least frequently identified as always undertaken.

Participation in professional development opportunities incorporating experiential

learning strategies and a values-based approach as well as supported clinical

experiences in the provision of end-of-life care can equip nurses with the knowledge,

skills and ability to engage in interpersonal caring practices.

An end-of-life care practice context of nurse-led care with variation in availability of

resources was identified. Contextual factors had a limited influence as explanatory

variables and were only associated with environmental modification practices. These

findings suggest that the broader practice context and culture of critical care are less

significant in the intimate settings of end-of-life care, where interpersonal care by the

critical care nurse comprises a large portion of practice. Further research that

considers the perspectives of families and other members of the multidisciplinary

team would make a valuable contribution to knowledge of end-of-life care in critical

care settings.

The findings of this study informed a number of recommendations for practice and

education and further research. The provision of end-of-life care in critical care

settings represents a significant component of nursing work that is highly valued by

nurses engaged in this work. Critical care nurses must be adequately prepared and

supported to provide comprehensive care in all areas of end-of-life care practice to

optimise care for patients at the end of life and their families, and to maintain their

own health and wellbeing in this often emotionally demanding role.

213

12 REFERENCES

Aitken, L. M., Currey, J., Marshall, A., & Elliott, D. (2006). The diversity of critical care nursing education in Australian universities. Australian Critical Care, 19(2), 46–52.

Australian and New Zealand Intensive Care Society. (2003). Statement on withholding and withdrawing treatment. Version 1. Melbourne, Vic: Author. Retrieved from the Australian and New Zealand Intensive Care Society (ANZICS) website: http://www.anzics.com.au/core/reports

Australian College of Critical Care Nurses. (2003). ICU staffing position statement (2003) on intensive care nursing staffing. Version January 2003. Melbourne, Vic: Author. Retrieved from http://www.acccn.com.au/images/stories/downloads/staffing_intensive_care_nursing.pdf

Australian Council on Healthcare Standards. (1997). Guidelines for intensive care units. Sydney, NSW: Author.

Australian Government Department of Health and Ageing. (2010). Supporting Australians to live well at the end of life: National Palliative Care Strategy 2010. Retrieved from the Australian Government Department of Health and Ageing (DoHA) website: http://www.health.gov.au/internet/main/publishing.nsf/Content/palliativecare-strategy.htm

Australian Government National Health and Hospitals Reform Commission. (2009). A healthier future for all Australians: Final report of the National Health and Hospitals Reform Commission June 2009. Retrived from the Australian Government National Health and Hospitals Reform Commission (NHHRC) website: http://www.health.gov.au/internet/nhhrc/publishing.nsf/Content/home-1

Australian Institute of Health and Welfare (2011). Nursing and midwifery labour force 2009. AIHW bulletin no. 90. Cat. no. AUS 139. Canberra, ACT: Author. Retrieved from http://www.aihw.gov.au/publication-detail/?id=10737419682

214

Badger, J. M. (2005a). A descriptive study of coping strategies used by medical intensive care unit nurses during transitions from cure- to comfort-oriented care. Heart & Lung, 34(1), 63–68.

Badger, J. M. (2005b). Factors that enable or complicate end-of-life transitions in critical care. American Journal of Critical Care, 14(6), 513–521.

Bandalos, D. L., & Finney, S. J. (2010). Factor analysis: Exploratory and confirmatory. In G. R. Hancock and R. O. Mueller (Eds.), The reviewers guide to quantitative methods in the social sciences (pp. 93–114). Hoboken, NJ: Routledge.

Beckstead, J.W. (2009). Content validity is naught. International Journal of Nursing Studies, 46, 1274–1283.

Beckstrand, R. L., Callister, L. C., & Kirchhoff, K. T. (2006). Providing a ‘good death’: Critical care nurses’ suggestions for improving end-of-life care. American Journal of Critical Care, 15(1), 38–45.

Beckstrand, R. L., & Kirchhoff, K. T. (2005). Providing end-of-life care to patients: Critical care nurses’ perceived obstacles and supportive behaviors. American Journal of Critical Care, 14(5), 395–403.

Beckstrand, R. L., Smith, M. D., Heaston, S., & Bond, A. E. (2008). Emergency nurses’ perceptions of size, frequency, and magnitude of obstacles and supportive behaviours in end-of-life care. Journal of Emergency Nursing, 34(4), 290–300.

Billings, J. A., Keeley, A., Bauman, J., Cist, A., Coakley, E., Dahlin, C., . . . Wise, M. (2006). Merging cultures: Palliative care specialists in the medical intensive care unit. Critical Care Medicine, 34(11), S388–S393.

Bloomer, M., Lee, S., & O’Connor, M. (2010). End of life clinician-family communication in ICU: A retrospective observational study—implications for nursing. Australian Journal of Advanced Nursing, 28(2), 17–23.

215

Bloomer, M. J., Morphet, J., O’Connor, M., Lee, S., & Griffiths, D. (in press). Nursing care of the family before and after a death in the ICU—An exploratory pilot study. Australian Critical Care.

Bloomer, M. J., Triuvoipati, R., Tsiripillis, M., & Botha, J. A. (2010). End of life management of adult patients in an Australian metropolitan intensive care unit: A retrospective observational study. Australian Critical Care, 23(1), 13–19.

Borbasi, S., Wotton, K., Redden, M., & Chapman, Y. (2005) Letting go: A qualitative study of acute care and community nurses’ perceptions of a ‘good’ versus a ‘bad’ death. Australian Critical Care, 18(3), 104–113.

Boyle, D. K., Miller, P. A., & Forbes-Thompson, S. A. (2005). Communication and end-of-life care in the intensive care unit: Patient, family and clinician outcomes. Critical Care Nursing Quarterly, 28(4), 302–316.

Brown, B. (2009). Men in nursing: Re-evaluating masculinities, re-evaluating gender. Contemporary Nurse, 33(2), 120–129.

Bucknall, T., & Thomas, S. (1997) Nurses’ reflections on problems associated with decision-making in critical care settings. Journal of Advanced Nursing, 25(2), 229–237.

Burns, K. J., Jacobs, B. B., & Jacobs, L. M. (2011). A time for trauma end-of-life optimum support: The TELOS best-practice model. Journal of Trauma Nursing, 18(2), 97–101.

Byock, I. (2002). The meaning and value of death. Journal of Palliative Medicine, 5(2), 279–288.

Calvin, A. O., Kite-Powell, D. M., & Hickey, J. V. (2007). The neuroscience ICU nurse’s perceptions about end-of-life care. Journal of Neuroscience Nursing, 39(3), 143–150.

216

Calvin, A. O., Lindy, C. M., & Clingon, S. L. (2009). The cardiovascular intensive care unit nurse’s experience with end-of-life care: A qualitative descriptive study. Intensive and Critical Care Nursing, 25(4), 214–220.

Campbell, M. L. (2002). End of life care in the ICU: Current practice and future hopes. Critical Care Nursing Clinics of North America, 14(2), 197–200.

Campbell, M. L. (2006). Palliative care consultation in the intensive care unit. Critical Care Medicine, 34(11), S355–S358.

Campbell, M. L., & Guzman, J. A. (2003). Impact of a proactive approach to improve end-of-life care in a medical ICU. Chest, 123(1), 266–271.

Cartwright, C., Steinberg, M., Williams, G., Najman, J., & Williams, G. (1997). Issues of death and dying: The perspective of critical care nurses. Australian Critical Care, 10(3), 81–87.

Christenson, M. (2007). Noise levels in a general intensive care unit: A descriptive study. Nursing in Critical Care, 12(4), 188–197.

Ciccarello, G. P. (2003). Strategies to improve end-of-life care in the intensive care unit. Dimensions of Critical Care Nursing, 22(5), 216–222.

Clarke, E. B., Curtis, J. R., Luce, J. M., Levy, M., Danis, M., Nelson, J., & Solomon, M. Z. (2003). Quality indicators for end-of-life care in the intensive care unit. Critical Care Medicine, 31(9), 2255–2262.

Clarke, E. B., Luce, J. M., Curtis, J. R., Danis, M., Levy, M., Nelson, J., & Solomon, M. Z. (2004). A content analysis of forms, guidelines and other materials documenting end-of-life care in intensive care units. Journal of Critical Care, 19(2), 108–117.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

217

Coombs, M. A. (2010). The mourning before: Can anticipatory grief theory inform family care in adult intensive care? International Journal of Palliative Nursing, 16(12), 580–584.

Coombs, M., & Long, T. (2008). Managing a good death in critical care: Can health policy help? Nursing in Critical Care, 13(4), 208–214.

Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment Research & Evaluation, 10(7), 1–9.

Creswell, J. W., & Plano Clark, V. L. (2007). Designing and conducting mixed methods research. Thousand Oaks, CA: Sage.

Creswell, J. W., & Plano Clark, V. L. (2011). Designing and conducting mixed methods research (2nd ed.). Thousand Oaks, CA: Sage.

Crump, S. K., Schaffer, M. A., & Schulte, E. (2010). Critical care nurses’ perceptions of obstacles, supports and knowledge needed in providing quality end-of-life care. Dimensions of Critical Care Nursing, 29(6), 297–306.

Davidson, P. M., Introna, K., Cockburn, J., Daly, J., Dunford, M., Paull, G., & Dracup, K. (2002). Synergising acute care and palliative care to optimise nursing care in end-stage cardiorespiratory disease. Australian Critical Care, 15(2), 64–69.

Davidson, P. M., Introna, K., Daly, J., Paull, G., Jarvis, R., Angus, J., . . . Dracup, K. (2003). Cardiorespiratory nurses’ perceptions of palliative care in nonmalignant disease: Data for the development of clinical practice. American Journal of Critical Care, 12(1), 47–53.

Davis, L. L. (1992). Instrument review: Getting the most from a panel of experts. Applied Nursing Research, 5(4), 194–197.

Deffner, J. M. & Bell, S. K. (2005). Nurses’ death anxiety, comfort level during communication with patients and families regarding death, and exposure to

218

communication education: A quantitative study. Journal for Nurses in Staff Development, 21(1), 19–23.

DeVellis, R. F. (2003). Scale development: Theory and applications (2nd ed.). Thousand Oaks, CA: SAGE.

Dillman, D. A., Smyth, J. D., & Christian, L. M. (2009). Internet, mail and mixed-mode surveys: The tailored design method (3rd ed.). Hoboken, NJ: Wiley & Sons.

Drennan, K., Hicks P., & Hart G. K. (2010). Intensive care resources & activity: Australia & New Zealand 2007/2008. Retrieved from the Australian and New Zealand Intensive Care Society (ANZICS) website: http://www.anzics.com.au/core/reports

Efstathiou, N., & Clifford, C. (2011). The critical care nurse’s role in end-of-life care: Issues and challenges. Nursing in Critical Care, 16(3), 116–123.

Elliott, D., Aitken, L., & Chaboyer, W. (2012). ACCCN’s Critical Care Nursing (2nd ed.). Chatswood, NSW: Mosby Elsevier Australia.

Espinosa, L., Young, A., Symes, L., Haile, B., & Walsh, T. (2010). ICU nurses’ experiences in providing terminal care. Critical Care Nursing Quarterly, 33(3), 273–281.

Exley, C. (2004). Review article: The sociology of dying, death and bereavement. Sociology of Health & Illness, 26(1), 110–122.

Faas, A. (2004). The intimacy of dying. Dimensions of Critical Care Nursing, 23(3), 122–124.

Fabrigar, L. R., Wegener D. T., MacCallum R. C., & Strahan E. J. (1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4(3), 272–299.

219

Fairweather, C., & Gardner, G. (2000). Specialist nurse: An investigation of common and distinct aspects of practice. Collegian, 7(2), 26–33.

Field, A. (2009). Discovering statistics using SPSS (3rd ed.). London, UK: SAGE.

Floyd, F. J., & Widaman, K. F. (1995). Factor analysis in the development and refinement of clinical assessment instruments. Psychological Assessment, 7(3), 286–299.

Fontana, A., & Frey J. H. (2005). The interview: From neutral stance to political involvement. In N. K. Denzin & Y. S. Lincoln (Eds.), The Sage handbook of qualitative research (3rd ed.; pp. 695–727). California: Sage Publications.

Fridh, I., Forsberg, A., & Bergbom, I. (2009). Doing one’s utmost: Nurses’ descriptions of caring for dying patients in an intensive care environment. Intensive and Critical Care Nursing, 25(5), 233–241.

Fulford, K. W. M. (2011). The value of evidence and evidence of values: Bringing together values-based and evidence-based practice in policy and service development in mental health. Journal of Evaluation in Clinical Practice, 17(5), 976–987.

Grant, J. S., & Davis, L. L. (1997). Selection and use of content experts for instrument development. Research in Nursing & Health, 20(3), 269–274.

Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate data analysis (6th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.

Halcomb, E., Daly, J., Jackson, D., & Davidson, P. (2004). An insight into Australia nurses’ experience of withdrawal/withholding of treatment in the ICU. Intensive and Critical Care Nursing, 20(4), 214–222.

Hall, R. I., Rocker, G. M., & Murray, D. (2004). Simple changes can improve conduct of end-of-life care in the intensive care unit. Canadian Journal of Anesthesia, 51(6), 631–636.

220

Hansen, L., Goodell, T. T., DeHaven, J., & Smith, M. D. (2009). Nurses’ perceptions of end-of-life care after multiple interventions for improvement. American Journal of Critical Care, 18(3), 263–272.

Hart, B., Sainsbury, P., & Short, S. (1998). Whose dying? A sociological critique of the ‘good death’. Mortality, 3(1), 65–77.

Harvey, G., Loftus-Hills, A., Rycroft-Malone, J., Titchen, A., Kitson, A., McCormack, B., & Seers, K. (2002). Getting evidence into practice: The role and function of facilitation. Journal of Advanced Nursing, 37(6), 577–588.

Heland, M. (2006). Fruitful or futile: Intensive care nurses’ experiences and perceptions of medical futility. Australian Critical Care, 19(1), 25–31.

Helfrich, C. D., Damschroder, L. J., Hagedorn, H. J., Daggett, G. S., Sahay, A., Ritchie, M., . . . Stetler, C. B. (2010). A critical synthesis of literature on the promoting action on research implementation in health services (PARIHS) framework. Implementation Science, 5(82). Retrieved from http://www.implementationscience.com/content/5/1/82

Hockey, J. (2007). Closing in on death? Reflections on research and researchers in the field of death and dying. Health Sociology Review, 16(5), 436–446.

Hodde, N. M., Engelberg, R. A., Treece, P. D., Steinberg, K. P., & Curtis, J. R. (2004). Factors associated with nurse assessment of the quality of dying and death in the intensive care unit. Critical Care Medicine, 32(8), 1648–1653.

Hov, R., Hedelin, B., & Athlin E. (2007). Good nursing care to ICU patients on the edge of life. Intensive and Critical Care Nursing, 23, 331–341.

Jackson, A., Purkis, J., Burnham, E., Hundt, G. L., & Blaxter, L. (2010). Views of relatives, carers and staff on end-of-life care pathways. Emergency Nurse, 17(10), 22–26.

221

Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher, 33(7), 14–26.

Johnson, R. B., Onwuegbuzie, A. J., & Turner, L. A. (2007). Toward a definition of mixed methods research. Journal of Mixed Methods Research, 1(2), 112–133.

Joint Faculty of Intensive Care Medicine. (2003) Minimum standards for intensive care units. IC-1. Melbourne, Vic: Author.

Jones, T., & FitzGerald M. (1998). Withdrawal of life-support treatment: The experience of critical care nurses. Australian Critical Care, 11(4), 117–121.

Judson, J. A., & Fisher, M. M. (2006). Intensive care in Australia and New Zealand. Critical Care Clinics, 22(3), 407–423.

Kain, V., Gardner, G., & Yates, P. (2009). Neonatal Palliative Care Attitude Scale: Development of an instrument to measure the barriers to and facilitators of palliative care in neonatal nursing. Pediatrics, 123(2), e207–e213.

Kehl, K. A. (2006). Moving towards peace: An analysis of the concept of a good death. American Journal of Hospice and Palliative Medicine, 23(4), 277–286.

Kellehear, A. (2000). The Australian way of death: Formative historical and social influences. In A. Kellehear (Ed.), Death & Dying in Australia (pp. 173–189). South Melbourne, Vic: Oxford University Press.

Kellehear, A. (2008). Modern death: Who cares? Soundings, 38, 82–88.

Kim, H-J. (2008). Common factor analysis versus principal component analysis: Choice for symptom cluster research. Asian Nursing Research, 2(1), 17–24.

Kinoshita, S., & Miyashita, M. (2011). Development of a scale for ‘difficulties felt by ICU nurses providing end-of-life care’ (DFINE): A survey study. Intensive and Critical Care Nursing, 27(4), 202–210.

222

Kirchhoff, K. T., Anumandla, P. R., Foth, K. T., Lues, S. N., & Gilbertson-White, S. H. (2004). Documentation on withdrawal of life support in adult patients in the intensive care unit. American Journal of Critical Care, 13(4), 328–334.

Kirchhoff, K. T., & Beckstrand, R. L. (2000). Critical care nurses’ perceptions of obstacles and helpful behaviours in providing end-of-life care to dying patients. American Journal of Critical Care, 9(2), 96–105.

Kirchhoff, K. T., Beckstrand, R. L., & Anumandla, P. R. (2003). Analysis of end-of-life content in critical care nursing textbooks. Journal of Professional Nursing, 19(6), 372–381.

Kirchhoff, K. T., & Kowalkowski, J. A. (2010). Current practices for withdrawal of life support in intensive care units. American Journal of Critical Care, 19(6), 532–541.

Kirchhoff, K. T., Spuhler, V., Walker, L., Hutton, A., Vaughan Cole, B., & Clemmer, T. (2000). Intensive Care Nurses’ experiences with end-of-life care. American Journal of Critical Care, 9(1), 36–42.

Kitson, A., Harvey, G., & McCormack, B. (1998). Enabling the implementation of evidence based practice: A conceptual framework. Quality in Health Care, 7(3), 149–158.

Kitson, A. L., Rycroft-Malone, J., Harvey, G., McCormack, B., Seers, K. & Titchen, A. (2008). Evaluating the successful implementation of evidence into practice using the PARIHS framework: Theoretical and practical challenges. Implementation Science, 3(1), 1–12.

Kjerulf, M., Regehr, C., Popova, S. R., & Baker, A. J. (2005). Family perceptions of end-of-life care in an urban ICU. Dynamics, 16(3), 22–25.

Latour, J. M., Fulbrook, P., & Albarran, J. W. (2009). EfCCNa survey: European intensive care nurses’ attitudes and beliefs towards end-of-life care. Nursing in Critical Care, 14(3), 110–121.

223

Le Conte, P., Riochet, D., Batard, E., Volteau, C., Giradudeau, B., Arnaudet, I., . . . Potel, G. (2010). Death in emergency departments: A multicentre cross-sectional survey with analysis of withholding and withdrawing life support. Intensive Care Medicine, 36(5), 765–772.

Luce, J. M. (1997). Making decisions about the forgoing of life-sustaining therapy. American Journal of Respiratory Critical Care Medicine, 156(6), 1715–1718.

Luk, E. K., & Rose, L. (2011). Conducting a large survey of critical care nurses in Canada: Lessons learnt. Intensive and Critical Care Nursing, 27(4), 173–179.

Lynn, M. R. (1986). Determination and quantification of content validity. Nursing Research, 35(6), 382–385.

Marx, R. G., Menezes, A., Horovitz, L., Jones, E. C., & Warren, R. F. (2003). A comparison of two time intervals for test-retest reliability of health status instruments. Journal of Clinical Epidemiology, 56(8), 730–735.

McCarthy, J., & Grumley, S. (2010). A values-based era in professional caring. In J. McCarthy & P. Rose (Eds.), Values-based health & social care (pp. 142–164). London, UK: SAGE.

McClement, S. E., & Degner, L. F. (1995). Expert nursing behaviors in care of the dying adult in the intensive care unit. Heart & Lung, 24(5), 408–419.

McCormack, B., Kitson, A., Harvey, G., Rycroft-Malone, J., Titchen, A., & Seers, K. (2002). Getting evidence into practice: The meaning of ‘context’. Journal of Advanced Nursing, 38(1), 94–104.

McKeown, A., Cairns, C., Cornbleet, M., & Longmate, A. (2010). Palliative care in the intensive care unit: An interview-based study of the team perspective. International Journal of Palliative Nursing, 16(7), 334–338.

McLean, C. (2012). The yellow brick road: A values based curriculum model. Nurse Education in Practice, 12(3), 159–163.

224

McMillen, R. E. (2008). End-of-life decisions: Nurses perceptions, feelings and experiences. Intensive and Critical Care Nursing, 24(4), 251–259.

McMordie, W. R. (1979). Improving measurement of death anxiety. Psychological Reports, 44(3 Pt 1), 975–980.

McNamara, B. (2001). Fragile lives: Death, dying and care. Crows Nest, NSW: Allen & Unwin.

Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage.

Monsen, M. G., & Edell-Gustafsson, U. M. (2005). Noise and sleep disturbance factors before and after implementation of a behavioural modification programme. Intensive & Critical Care Nursing, 21(4), 208–219.

Moran, J. L., Bristow, P., Solomon, P. J., George, C., & Hart, G. K. (2008). Mortality and length-of-stay outcomes, 1993–2003, in the binational Australian and New Zealand intensive care adult patient database. Critical Care Medicine, 36(1), 46–61.

Morgan, D. L. (2007). Paradigms lost and pragmatism regained: Methodological implications of combining qualitative and quantitative methods. Journal of Mixed Methods Research, 1(1), 48–76.

Morrison, A. L., Beckmann, U., Durie, M., Carless, R., & Gillies, D. M. (2001). The effects of nursing staff inexperience (NSI) on the occurrence of adverse patient experiences in ICUs. Australian Critical Care, 14(3), 116–121.

Mulligan, A. (2005). Should dying patients be monitored? A reflective analysis of a critical incident. Nursing in Critical Care, 10(3), 122–126.

National Health and Medical Research Council. (2007a). Australian code for the responsible conduct of research. Retrieved from http://www.nhmrc.gov.au/_files_nhmrc/publications/attachments/r39.pdf

225

National Health and Medical Research Council. (2007b). National statement on ethical conduct in human research. Retrieved from http://www.nhmrc.gov.au/_files_nhmrc/publications/attachments/e72.pdf

National Health and Medical Research Council. (2011). An ethical framework for integrating palliative care principles into the management of advanced chronic or terminal conditions. Retrieved from http://www.nhmrc.gov.au/_files_nhmrc/publications/attachments/rec31_ethical_framework_palliative_care_terminal_110908.pdf

Nelson, J. E. (2006). Identifying and overcoming the barriers to high-quality palliative care in the intensive care unit. Critical Care Medicine, 34(11), S324–S331.

Nelson, J. E., Angus, D. C., Weissfeld, L. A., Puntillo, K. A., Danis, M., Deal, D., . . . Cook, D. J. (2006) End-of-life care for the critically ill: A national intensive care unit survey. Critical Care Medicine, 34(10), 2547–2553.

Nelson, J. E., & Meier, D. E. (1999). Palliative care in the intensive care unit: Part I. Journal of Intensive Care Medicine, 14(3), 130–139.

Nilsson, K., & Larsson, U. S. (2005). Conceptions of gender: A study of female and male head nurses’ statements. Journal of Nursing Management, 13(2), 179–186.

Nordgren, L., & Olsson, H. (2004). Palliative care in a coronary care unit: A qualitative study of physicians’ and nurses’ perceptions. Journal of Clinical Nursing, 13(2), 185–193.

Norton, S. A., Tilden, V. P., Tolle, S. W., Nelson, C. A., & Eggman, S. T. (2003). Life support withdrawal: Communication and conflict. American Journal of Critical Care, 12(6), 548–555.

NSW Department of Health. (2005). Guidelines for end-of-life care and decision-making. Retrieved from http://www.health.nsw.gov.au/policies/gl/2005/pdf/GL2005_057.pdf

226

Ogle, R., Copley, D., Bethune, E., & Parkin, N. (2004). Intensive care workforce and workflow: Literature review. Melbourne, Vic: Deakin University, School of Nursing and Midwifery.

O’Mahony, S., McHenry, J., Blank, A. E., Snow, D., Karakas, S. E., Santoro, G., Selwyn, P., & Kvetan, V. (2010). Preliminary report of the integration of a palliative care team into an intensive care unit. Palliative Medicine, 24(2), 154–165.

Paice, J. A., Ferrell, B. R., Virani, R., Grant, M., Malloy, P., & Rhome, A. (2006). Graduate nursing education regarding end-of-life care. Nursing Outlook, 54(1), 46–52.

Palliative Care Australia. (2005). Standards for providing quality palliative care for all Australians. Retrieved from http://www.palliativecare.org.au/Portals/46/Standards%20for%20providing%20quality%20palliative%20care%20for%20all%20Australians.pdf

Palliative Care Australia. (2008). Palliative and end of life care—Glossary of terms. Retrieved from http://www.palliativecare.org.au/Portals/46/PCA%20glossary.pdf

Palliative Care Curriculum for Undergraduates (PCC4U) Project Team, with funding from the Australian Government Department of Health and Ageing. (2011). Palliative Care Curriculum for Undergraduates (PCC4U). Retrieved from www.pcc4u.org

Polit, D. F., & Beck, C. T. (2006). The content validity index: Are you sure you know what’s being reported? Critique and recommendations. Research in Nursing & Health, 29(5), 489–497.

Polit, D. F., & Beck, C. T. (2008). Nursing research: Generating and accessing evidence for nursing practice (8th ed.). Philadelphia, PA: Wolters Kluwer/Lippincott, Williams & Wilkins.

Polit, D. F., & Beck, C. T. (2010). Essentials of nursing research: Appraising evidence for nursing practice (7th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.

227

Polit, D. F., Beck, C. T., & Owen, S. V. (2007). Is the CVI an acceptable indicator of content validity? Appraisal and recommendations. Research in Nursing & Health, 30(4), 459–467.

Program of Experience in the Palliative Approach (PEPA) Project Team, with funding from the Australian Government Department of Health and Ageing. (2003). Program of Experience in the Palliative Approach (PEPA). Retrieved from www.pepaeducation.com

Puntillo, K. A., Benner, P., Drought, T., Drew, B., Stotts, N., Stannard, D., . . . White, C. (2001). End-of-life issues in intensive care units: A national random survey of nurses’ knowledge and beliefs. American Journal of Critical Care, 10(4), 216–229.

Rodriguez, K. L., Barnato, A. E., & Arnold, R. M. (2007) Perceptions and utilization of palliative care services in acute care hospitals. Journal of Palliative Medicine, 10(1), 99–110.

Rogers, A., Karlsen, S., & Addington-Hall, J. (2000). ‘All the services were excellent. It is when the human element comes in that things go wrong’: Dissatisfaction with hospital care in the last year of life. Journal of Advanced Nursing, 31(4), 768–774.

Rose, P., & Gidman, J. (2010) Evidence-based practice within values-based care. In J. McCarthy & P. Rose (Eds.), Values-based health & social care (pp. 25–47). London, UK: SAGE.

Rosenberg, J. P. (2009). Circles in the surf: Australian masculinity, mortality and grief. Critical Public Health, 19(3–4), 417–426.

Ross, M. M., McDonald, B., & McGuinness, J., (1996). The palliative care quiz for nursing (PCQN): The development of an instrument to measure nurses’ knowledge of palliative care. Journal of Advanced Nursing, 23(1), 126–137.

Rushton, C. H., Williams, M. A., & Sabatier, K. H. (2002). The integration of palliative care and critical care: One vision, one voice. Critical Care Nursing Clinics of North America, 14(2), 133–140.

228

Rycroft-Malone, J., Harvey, G., Seers, K., Kitson, A., McCormack, B., & Titchen, A. (2004). An exploration of factors that influence the implementation of evidence into practice. Journal of Clinical Nursing, 13(8), 913–924.

Rycroft-Malone, J., Kitson, A., Harvey, G., McCormack, B., Seers, K., Titchen, A., & Estabrooks, C. (2002) Ingredients for change: Revisiting a conceptual framework. Quality & Safety in Health Care, 11(2), 174–180.

Rycroft-Malone, J., Seers, K., Titchen, A., Harvey, G., Kitson, A., & McCormack, B. (2004). What counts as evidence in evidence-based practice? Journal of Advanced Nursing, 47(1), 81–90.

Schneider, Z., Elliott, D., LoBiondo-Wood, G., & Haber, J. (2003). Nursing research: Methods, critical appraisal and utilisation (2nd ed.). Chatswood, NSW: Elsevier Australia.

Seymour, J. (2001). Critical moments: Death and dying in intensive care. Buckingham, UK: Open University Press.

Shorter, M., & Stayt, L. C. (2010). Critical care nurses’ experiences of grief in an adult intensive care unit. Journal of Advanced Nursing, 66(1), 159–167.

Silverberg, R. A. (1985). Men confronting death: Management versus self determination. Clinical Social Work Journal, 13(2), 157–169.

Silverman, D. (2006). Interpreting qualitative data: Methods for analysing talk, text and interaction (3rd ed.). London, UK: Sage.

Skerman, H. M., Yates, P. M., & Battistutta, D. (2009). Multivariate methods to identify cancer-related symptom clusters. Research in Nursing & Health, 32(3), 345–360.

Stevens, J., McFarlane, J., & Stirling, K. (2000). Ageing and dying. In A. Kellehear (Ed.), Death & dying in Australia (pp. 173–189). South Melbourne, Vic: Oxford University Press.

229

Strazzari, M. (2005). Ageing, dying and death in the twenty-first century. In J. Germov (Ed.), Second opinion: An introduction to health sociology (pp. 244–264). South Melbourne, Vic: Oxford University Press.

Stroud, R. (2002). The withdrawal of life support in adult intensive care: An evaluative review of the literature. Nursing in Critical Care, 7(4), 176–184.

Sue, V. M., & Ritter, L. A. (2007). Conducting online surveys. Thousand Oaks, CA: Sage.

Tabachnick, B. G., & Fidell L. S. (2007). Using multivariate statistics (5th ed.). Boston, MA: Pearson Education.

Tashakkori, A., & Teddlie, C. (2010). Putting the human back in ‘human research methodology’: The researcher in mixed methods research. Journal of Mixed Methods Research, 4: 271–277.

Teddlie, C., & Tashakkori, A. (2010). Overview of contemporary issues in mixed methods research. In A. Tashakkori & C. Teddlie (Eds.), SAGE handbook of mixed methods in social & behavioural research (2nd ed.). Thousand Oaks CA: SAGE.

Templer, D. I. (1970). The construction and validation of a death anxiety scale. The Journal of General Psychology, 82, 165–177.

Thelan, M. (2005). End-of-life decision making in intensive care. Critical Care Nurse, 25(6), 28–37.

Thompson, G., McClement, S., & Daeninck, P. (2006) Nurses’ perceptions of quality end-of-life care on an acute medical ward. Journal of Advanced Nursing, 53(2), 169–177.

Timmermans, S. (2005). Death brokering: Constructing culturally appropriate deaths. Sociology of Health & Illness, 27(7), 993–1013.

230

Treece, P. D., Engelberg, R. A., Crowley, L., Chan, J. D., Rubenfeld, G. D., Steinberg, K. P., & Curtis, J. R. (2004) Evaluation of a standardized order form for the withdrawal of life support in the intensive care unit. Critical Care Medicine, 32(5), 1141–1148.

Treece, P. D., Engelberg, R. A., Shannon, S. E., Nielsen, E. L. Braungardt, T., Rubenfeld, G. D., Steinberg, K. P., & Curtis, J. R. (2006). Integrating palliative and critical care: Description of an intervention. Critical Care Medicine, 34(11), S380–S387.

Truog, R. D., Campbell, M. L., Curtis, J. R., Haas, C. E., Luce, J. M., Rubenfeld, G. D., . . . Kaufman, D. C. (2008). Recommendations for end-of-life care in the intensive care unit: A consensus statement by the American College of Critical Care Medicine. Critical Care Medicine, 36(3), 953–963.

Vanderspank-Wright, B., Fothergill-Bourbonnais, F., Brajtman, S., & Gagnon, P. (2011). Caring for patients and families at end of life: The experiences of nurses during withdrawal of life-sustaining treatment. Dynamics, 22(4), 31–35.

WorldAPP Key Survey (2009). Key Survey. (version 7.2). Braintree, MA: WorldAPP. Retrieved from http://survey.qut.edu.au/site/

WorldAPP Key Survey (2011). Key Survey. (version 7.3). Braintree, MA: WorldAPP. Retrieved from http://survey.qut.edu.au/site/

Wotton, K., Borbasi, S., & Redden, M. (2005). When all else has failed: Nurses’ perception of factors influencing palliative care for patients with end-stage heart failure. Journal of Cardiovascular Nursing, 20(1), 18–25.

Wunsch, H., Harrison, D. A., Harvey, S., & Rowan, K. (2005). End-of-life decisions: A cohort study of the withdrawal of all active treatment in intensive care units in the United Kingdom. Intensive Care Medicine, 31(6), 823–831.

Zomorodi, M., & Lynn, M. R. (2010a). Instrument development measuring critical care nurses’ attitudes and behaviors with end-of-life care. Nursing Research, 59(4), 234–240.

231

Zomorodi, M., & Lynn, M. R. (2010b). Critical care nurses’ values and behaviours with end-of-life care: Perceptions and challenges. Journal of Hospice and Palliative Nursing, 12(2), 89–96.

232

13 APPENDICES

13.1 QUT research ethics approval—Phase one

From: Research Ethics [mailto:ethicscontact@qut.edu.au]

Sent: Wed 8/10/2008 12:53 PM

To: Ms Kristen Louise Ranse

Cc: Ms Janette Lamb

Subject: Ethics Application Approval -- 0800000660

Dear Ms Kristen Ranse

Re: End-of-life care in the intensive care setting: Australian nurses’ practice

behaviours and factors influencing these practice behaviours

This email is to advise that your application has been reviewed and confirmed as

meeting the requirements of the National Statement on Ethical Conduct in Human

Research. Your ethics approval number is 0800000660. Please quote this number in

all future correspondence.

Whilst the data collection of your project has received ethical clearance, the decision

to commence and authority to commence may be dependant on factors beyond the

remit of the ethics review process. For example, your research may need ethics

clearance from other organisations or permissions from other organisations to access

staff. Therefore the proposed data collection should not commence until you have

satisfied these requirements.

If you require a formal approval certificate, please respond via reply email and one

will be issued.

Decisions related to Low Risk ethical review are subject to ratification at the next

available Committee meeting. You will only be contacted again in relation to this

matter if the Committee raises any additional questions or concerns.

233

This project has been awarded ethical clearance until 8/10/2011 and a progress report

must be submitted for an active ethical clearance at least once every twelve months.

Researchers who fail to submit an appropriate progress report may have their ethical

clearance revoked and/or the ethical clearances of other projects suspended. When

your project has been completed please advise us by email at your earliest

convenience.

Please do not hesitate to contact the unit if you have any queries.

Regards

Research Ethics Unit | Office of Research

O Block Podium | Gardens Point Campus

p +61 7 3138 5123 | f +61 7 3138 1304

e ethicscontact@qut.edu.au

w http://www.research.qut.edu.au/ethics/

234

13.2 QUT research ethics approval—Phases two and three

From: Research Ethics [ethicscontact@qut.edu.au]

Sent: Monday, 14 December 2009 10:14 AM

To: KRISTEN RANSE; Patsy Yates; Fiona Coyer

Cc: Janette Lamb

Subject: Ethics Application Approval -- 0900001370

Dear Ms Kristen Ranse

Project Title:

End-of-life care in the intensive care setting: Australian nurses'

practices and factors affecting these practices

Approval Number: 0900001370

Clearance Until: 14/12/2012

Ethics Category: Human

This email is to advise that your application has been reviewed by the

Chair, University Human Research Ethics Committee, and confirmed as meeting

the requirements of the National Statement on Ethical Conduct in Human

Research.

***Please make particular note of:

1. The QUT Code of Practice requirements for data storage (resend link

to Statement of Authorship and Location of Data Form):

http://www.mopp.qut.edu.au/D/D_02_06.jsp#D_02_06.08.mdoc

2. Data must NOT be stored on USB devices as these are insecure and

easily stolen. They should be used for data transfer between machines only.

***

Whilst the data collection of your project has received ethical clearance,

the decision to commence and authority to commence may be dependant on

235

factors beyond the remit of the ethics review process. For example, your

research may need ethics clearance from other organisations or permissions

from other organisations to access staff. Therefore the proposed data

collection should not commence until you have satisfied these requirements.

If you require a formal approval certificate, please respond via reply

email and one will be issued.

Decisions related to low risk ethical review are subject to ratification at

the next available Committee meeting. You will only be contacted again in

relation to this matter if the Committee raises any additional questions or

concerns.

This project has been awarded ethical clearance until 14/12/2012 and a

progress report must be submitted for an active ethical clearance at least

once every twelve months. Researchers who fail to submit an appropriate

progress report may have their ethical clearance revoked and/or the ethical

clearances of other projects suspended. When your project has been

completed please advise us by email at your earliest convenience.

For variations, please complete and submit an online variation form:

http://www.research.qut.edu.au/ethics/forms/hum/var/variation.jsp

Please do not hesitate to contact the unit if you have any queries.

Regards

Research Ethics Unit | Office of Research

Level 4 | 88 Musk Ave | Kelvin Grove

p: +61 7 3138 5123 | f: +61 7 3138 4543

e: ethicscontact@qut.edu.au | w: http://www.research.qut.edu.au/ethics/

236

13.3 ACT health ethics approval

237

238

13.4 Email indicating support for phase three of the study from the Australian

College of Critical Care Nurses

From: ACCCN [acccn@acccn.com.au]

Sent: Friday, 16 October 2009 3:26 PM

To: KRISTEN RANSE

Subject: RE: Recruitment of ACCCN members as participants in research

Hi Kristen,

Your research request has been approved. We will distribute an e-mail in March

2010 (or when you inform us), to invite members who have nominated 'to be

contacted for research' to participate in your research project. Keep me posted.

Have a good weekend.

Lynn Herson

Australian College of Critical Care Nurses Ltd

Tel:(03) 9347 8577 Fax: (03) 9347 8522

Free Call: 1800 357 968

National Office: acccn@acccn.com.au

Website: http://www.acccn.com.au/

239

13.5 Phase one bulletin board advertisement

PARTICIPATE IN RESEARCH

Information for Prospective Participants The following research activity has been reviewed via QUT arrangements for the conduct of research involving human participation and the ACT Health Human Research Ethics Committee. If you choose

to participate, you will be provided with more detailed participant information, including who you can contact if you have any concerns.

End-of-life care in the intensive care setting: Australian nurses’ practice behaviours and factors influencing these practice behaviours.

Research Team Contacts Kristen Ranse, PhD student Professor Patsy Yates, Principal supervisor

Phone: 0409 034 015 Phone: (07) 3138 3835 Email: k.ranse@student.qut.edu.au Email: p.yates@qut.edu.au

Please contact the researcher team members to have any questions answered or if you require further information about the project

What is the purpose of the research? The purpose of this research is to identify and describe Intensive Care Nurses’ experiences and practices in relation to providing end-of-life care to patients and their families in the intensive care setting.

Are you looking for people like me? The research team is looking for Registered Nurses currently employed in a clinical position within the Intensive Care Unit who have provided care to a dying patient and their family in the Intensive Care Unit.

What will you ask me to do? Your participation will involve an interview of approximately 1 hour in length to discuss your experiences and practices in the provision of end-of-life care.

Are there any risks for me in taking part? The research team does not believe there are any risks for you if you choose to participate in this research.

It should be noted that if you do agree to participate, you can withdraw from participation at any time during the project without comment or penalty.

Are there any benefits for me in taking part? It is expected that this project will not benefit you directly. However, the findings of this project will contribute to nursing knowledge of current end-of-life care practice and the future evaluation of end-of-life care practice in intensive care settings. Recommendations for practice development to prepare and support intensive care nurses in the provision of end-of-life care will be made.

Will I be compensated for my time? This research has been endorsed by the Clinical Nurse Consultant (ICU) and Acting Director of Nursing (Surgical Services) and approval has been granted for participants to attend the interview during work time.

I am interested—what should I do next? If you would like to participate in this study, please contact the research team for details of the next step.

You will be provided with further information to ensure that your decision and consent to participate is fully informed.

Thank You!

240

13.6 Phase one information sheet and consent form

PARTICIPANT INFORMATION for QUT RESEARCH PROJECT

End-of-life care in the intensive care setting: Australian nurses’ practice behaviours

and factors influencing these practice behaviours.

Research Team Contacts Kristen Ranse, PhD student Prof. Patsy Yates, Principal supervisor

Phone: 0409 034 015 Phone: (07) 3138 3835 Email: k.ranse@student.qut.edu.au Email: p.yates@qut.edu.au

Description This project is being undertaken as part of a PhD program of research by Kristen Ranse. The purpose of this project is to identify and describe Intensive Care Nurses’ experiences and practices in relation to providing end-of-life care to patients and their families in the intensive care setting. The findings of this research will inform future research and practice development to prepare and support nurses in the provision of end-of-life care. Participation Your participation in this project is voluntary. If you do agree to participate, you can withdraw from participation at any time during the project without comment or penalty. Your decision to participate will in no way impact upon your current or future relationship with ACT Health or QUT. Your participation will involve an individual interview of approximately 1 hour in length at The Canberra Hospital. During this interview, you will be encouraged to discuss your experiences and practices in relation to the provision of end-of-life care. A digital voice recorder will be used to record each interview to enable a transcription of the discussion to be produced and permit the researcher to return to the discussion at a later time. Expected benefits It is expected that this project will not benefit you directly. However, the knowledge gained from this project will contribute to nursing knowledge of current end-of-life care practice in the intensive care setting and the future evaluation of end-of-life care for patients and their families. Risks The risks associated with your participation in this project are minimal. Reflection on experience is encouraged in professional practice; however, reflection can sometimes result in vulnerability and distress. You may withdraw your participation at any time during the project. Employees of ACT Health, and their family members, can access an employee assistance program to discuss matters relating to employment, such as stress. IPS Worldwide is the provider of the ACT Health Employee Assistance Program and can be contacted on 1300 366 789 or at www.eap.com.au Confidentiality All comments and responses are anonymous and will be treated confidentially. The names of individual persons are not required in any of the responses. Access to the digital audio recording and interview transcription will be restricted to the principal researcher and supervisory team. Consent to Participate We would like to ask you to sign a written consent form to confirm your agreement to participate. Questions / further information about the project

241

Please contact the researcher team members named above to have any questions answered or if you require further information about the project. Concerns / complaints regarding the conduct of the project ACT Health and QUT are committed to researcher integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project you may contact the ACT Health Human Research Ethics Committee: Secretariat, ACT Health Human Research Ethics Committee, Executive Coordination, Level 3, 11 Moore Street, Canberra City, ACT 2061 or phone (02) 6205 0846. Alternatively, you may contact the QUT Research Ethics Officer on (07) 3138 2340 or ethicscontact@qut.edu.au. The Research Ethics Officer is not connected with the research project and can facilitate a resolution to your concern in an impartial manner.

CONSENT FORM for QUT RESEARCH PROJECT

End-of-life care in the intensive care setting: Australian nurses’ practice behaviours

and factors influencing these practice behaviours.

Statement of consent By signing below, you are indicating that you:

• have read and understood the information document regarding this project

• have had any questions answered to your satisfaction

• understand that if you have any additional questions you can contact the research team

• understand that you are free to withdraw at any time, without comment or penalty

• understand that you can contact the ACT Health Human Research Ethics Committee on (02) 6205 0846 or the Research Ethics Officer at Queensland University of Technology (QUT) on (07) 3138 2340 or ethicscontact@qut.edu.au if you have concerns about the ethical conduct of the project .

• agree to participate in the project

• understand that the project will include audio recording

Name

Signature

Date / /

242

13.7 Email invitation to experts to participate in the survey review panel

Dear [NAME],

My name is Kristen Ranse and I am undertaking a research project on intensive care nurses’ end-of-life care practices and the factors influencing these practices as part of a PhD program of research at Queensland University of Technology. My supervisors are Prof Patsy Yates and Dr Fiona Coyer.

This research project involves a three-phase exploratory sequential mixed method design. Phase one used a qualitative approach to interview five intensive care nurses about their experiences and practices of end-of-life care. Phase two involves the development of a survey instrument, based on a review of the literature and the findings of phase one. The instrument will then be used in phase three to conduct a national survey of Australian intensive care nurse end-of-life care practices and factors that may influence these practices.

I am writing to you to invite you to participate as a content expert reviewer of the survey instrument. As a reviewer of the instrument, you will be asked to rate the relevance of each item on the instrument. You will also be asked to comment on the readability of the survey instrument, the survey content in regard to coverage of the variables of interest and suggest the addition, deletion and revision of individual items.

To assist you in reviewing the instrument, please find attached an information sheet outlining the research project including the research questions, definitions of key terms and an overview of each section of the survey instrument.

If you would like to participate as a content reviewer, please click [S] or copy and paste the URL below to complete the review online:

[URL]

I would appreciate your feedback on the instrument by Friday 4 December 2009. If you are unable to participate on this occasion, please advise me by e-mail (k.ranse@student.qut.edu.au). Please do not hesitate to contact me by e-mail or phone (0431 690 114) if you have any queries.

Your support for this project is valued.

Kind regards

Kristen Ranse

243

13.8 Information sheet for expert panel

INFORMATION SHEET for QUT RESEARCH

PROJECT

End-of-life care in the intensive care setting: Australian nurses’ practices and factors

affecting these practices.

Research Team Contacts

Kristen Ranse, PhD student Professor Patsy Yates, Principal supervisor

Phone: 0431 690 114 Phone: (07) 3138 3835

Email: k.ranse@student.qut.edu.au Email: p.yates@qut.edu.au

Project Description Although end-of-life care is a significant component of nursing work in the intensive care setting, few studies have examined nursing practice in relation to care of the dying patient and their family in this setting. Inadequate preparation and support of nurses providing end-of-life care, in addition to the stress associated with the provision of end-of-life care, may contribute to burnout and poor retention of intensive care nurses. To prepare and support nurses in the provision of optimal end-of-life care and to evaluate and improve the care that dying patients and their families receive, it is imperative that the current end-of-life care practices of intensive care nurses are identified and described. In addition, identification of factors influencing the end-of-life care practices of intensive care nurses can inform further research and development. This research project involves a three-phase exploratory sequential mixed method design. Phase one used a qualitative approach to interview five intensive care nurses to describe their experiences and practices of end-of-life care. Phase two involves the development of a survey instrument, based on a review of the literature and the findings of phase one. This phase also includes a review of the survey instrument by a panel of experts and a pilot test of the survey instrument to determine the validity and reliability of the instrument. The refined instrument will then be used in phase three to conduct a national survey of intensive care nurse end-of-life care practices and factors that may influence these practices. Research Questions The research questions of interest in this research project are:

1. What are the most frequently reported end-of-life care practices of Australian intensive care nurses?

2. Which factors predict intensive care nurse end-of-life care practices? a. What are the select content and facilitation factors influencing end-of-life care as

perceived by Australian intensive care nurses? b. What is the relationship between nurse characteristics and select context and

facilitation factors influencing end-of-life care? c. What is the relationship between select context and facilitation factors influencing

end-of-life care and end-of-life practices of intensive care nurses?

244

Variables Practices: the actions (and inactions) undertaken by intensive care nurses in the provision of end-of-life care of a patient and their family, where end-of-life care is defined as the care undertaken after a decision is made to withhold or withdraw treatment, with death of the patient the expected outcome. Practices will be measured on the survey instrument using a 5-point rating scale where 1 = never, 2 = rarely, 3 = sometimes, 4 = often and 5 = always. Context: all aspects of the setting where practice takes place (McCormack et al. 2002). Empirically derived contextual factors considered relevant to this research project include culture, physical environment, staffing and the availability of resources. Contextual factors will be measured on the survey instrument using a 5-point Likert scale where 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree and 5 = strongly agree. Facilitation: the process of making things easier. Facilitative factors include the attitudes and beliefs of the nurse and communication issues and the fragmentation of care delivery. Facilitation factors will be measured on the survey instrument using a 5-point Likert scale where 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree and 5 = strongly agree. Survey description The survey consists of three sections: demographic information, end-of-life care practices and factors influencing end-of-life care practices. The instructions to be provided to participants are in bold type under each section heading. Instructions specific to you as a member of the expert panel are indicated after the text EXPERT PANEL within the survey instrument. Section two of the instrument requires respondents to rate how often they undertake select practices when caring for a patient at the end-of-life. This section of the survey instrument aims to address the research question ‘What are the end-of-life care practices of Australian intensive care nurses?’. The items in section two (end-of-life care practices) on the survey instrument will be measured using a 5-point rating scale where 1 = never, 2 = rarely, 3 = sometimes, 4 = often and 5 = always. This rating scale will provide evidence of the frequency that specified practices are undertaken in the provision of end-of-life care by respondents to the survey. Section three of the survey instrument aims to address the research question: What factors predict Australian intensive care nurses’ end-of-life care practices?. Section three of the survey instrument includes a modified version of Kain et al.’s (2009) Neonatal Palliative Care Attitude Scale (NiPCAS) and additional items related to beliefs about end-of-life care and organisational and emotional support for clinicians that were emerged from the findings of phase one and the literature review of this research project. The items in section three (factors influencing end-of-life care) on the survey instrument will be measured using a 5-point Likert scale where 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree and 5 = strongly agree. References Kain, V., Gardner, G. & Yates, P. (2009). Neonatal Palliative Care Attitude Scale: Development of an instrument to measure the barriers to and facilitators of palliative care in neonatal nursing. Pediatrics, 123, e207–e213. McCormack, B., Kitson, A., Harvey, G., Rycroft-Malone, J., Titchen, A. & Seers, K. (2002). Getting evidence into practice: The meaning of ‘context’. Journal of Advanced Nursing, 38(1), 94–104.

245

13.9 Phase two pilot test email invitation, email reminder and email retest

invitation

Email invitation—Pilot test

Hi,

My name is Kristen Ranse and I’m undertaking a research project on end-of-life care in intensive care settings as part of a PhD program of research at Queensland University of Technology. An information sheet regarding this study is attached to this e-mail.

I am writing to you to invite you to participate in a pilot study by completing an online survey. In two to four weeks time, you will again be requested to complete the same survey. This will enable the survey instrument to be tested for reliability prior to the survey being distributed to a national sample.

If you would like to participate, please click [S] or copy and paste the URL below to complete the review online:

[URL]

Your participation in this research will contribute to the development of a survey that will be distributed to a national sample of critical care nurses. The knowledge gained from this PhD research project may inform the development of strategies to assist in the care of the dying patient and their family in critical care units and support nurses in the provision of this care.

Many thanks for your consideration in this request.

Kristen

Reminder invitation

Hi,

I recently sent you an invitation to participate in a research project on end-of-life care in critical care settings by completing an online survey. An information sheet regarding this study is attached to this e-mail.

246

If you would like to participate, please click [S] or copy and paste the URL below to complete the review online:

[URL]

Your participation in this research will contribute to the development of a survey that will be distributed to a national sample of critical care nurses. The knowledge gained from this PhD research project may inform the development of strategies to assist in the care of the dying patient and their family in critical care units and support nurses in the provision of this care.

Many thanks for your consideration in this request.

Kristen Ranse PhD Candidate Queensland University of Technology

Re-test email invitation

Hi,

Thankyou for participating in the pilot study for the research project on end-of-life care in intensive care settings by completing the online survey. I am writing to you to invite you to complete the survey again, to enable the survey instrument to be tested for reliability prior to the survey being distributed to a national sample. If you would like to participate, please click [S] or copy and paste the URL below and follow the instructions to complete the survey: [URL]

Your participation in this research will contribute to the development of a survey that will be distributed to a national sample of critical care nurses. The knowledge gained from this PhD research project may inform the development of strategies to assist in the care of the dying patient and their family in critical care units and support nurses in the provision of this care.

Many thanks for your consideration in this request.

Kristen

247

13.10 Phase three email invitation and email reminder

Email invitation

Subject: End-of-life care in critical care settings: A National survey

Dear ACCCN member,

My name is Kristen Ranse and I’m undertaking a research project on end-of-life care in critical care settings as part of a PhD program of research at Queensland University of Technology. I am writing to you to invite you to participate in this study by completing an online survey. An information sheet regarding this study is attached to this e-mail.

If you would like to participate, please click on the link below and follow the instructions to complete the survey:

[URL]

Your participation in this research will contribute to nursing knowledge of end-of-life care in critical care settings and the development of strategies to assist in the care of the dying patient and their family.

Many thanks for your consideration in this request.

Kristen

Email reminder

Subject: End-of-life care survey reminder

Dear ACCCN member,

My name is Kristen Ranse and I’m undertaking a research project on end-of-life care in critical care settings as part of a PhD program of research at Queensland University of Technology. I recently sent you an invitation to participate in this study by completing an online survey. An information sheet regarding this study is attached to this e-mail.

Thank you if you have already taken the time to participate in this survey. If you have not yet participated and would like to participate, please click on the link below

248

and follow the instructions to complete the survey by 5pm EST Wednesday 8 June :

[URL]

Your participation in this research will contribute to nursing knowledge of end-of-life care in critical care settings and the development of strategies to assist in the care of the dying patient and their family.

Many thanks for your consideration in this request.

Kristen Ranse

249

13.11 Information sheet attached to email invitation to participate (phase two

pilot test and phase three)

PARTICIPANT INFORMATION for QUT RESEARCH

PROJECT

End-of-life care in the intensive care setting: Australian nurses’

practices and factors affecting these practices.

Research Team Contacts

Kristen Ranse, PhD student Professor Patsy Yates, Principal supervisor

Phone: 0431 690 114 Phone: (07) 3138 3835

Email: k.ranse@student.qut.edu.au Email: p.yates@qut.edu.au

Description This project is being undertaken as part of a PhD program of research by Kristen Ranse. The purpose of this project is to identify the end-of-life care practices of Australian Intensive Care Nurses and the factors that might influence these practices. The findings of this research will inform future research and practice development to prepare and support nurses in the provision of end-of-life care. The project is funded by an ACT Nursing & Midwifery Board Research Grant. The ACT Nursing & Midwifery Board will not have access to the data obtained during the project. Participation Your participation in this project is voluntary. If you do agree to participate, you can withdraw from participation at any time during the project without comment or penalty. Your decision to participate will in no way impact upon your current or future relationship with QUT (for example your grades) or with the Australian College of Critical Care Nurses (ACCCN). As this project involves the submission of an anonymous survey, it will not be possible to withdraw from the project once submitted. Your participation will involve completion of an online survey. It is anticipated that completion of the survey will take approximately 20 minutes. Surveys will maintain participant anonymity by not requesting your name or contact details. The content of the survey will include demographic information, nurses end-of-life care practices and factors influencing nursing end-of-life care practice. Expected benefits It is expected that this project will not benefit you directly. However, this research will contribute to nursing knowledge of end-of-life care in the intensive care setting and inform the development of strategies to support nurses in the provision of end-of-life care. This research may inform future research to evaluate and improve the care that dying patients and their families receive. Risks The risks associated with your participation in this project are minimal. Reflection on experience is encouraged in professional practice, however, reflection can sometimes result in vulnerability or discomfort. If you do encounter any discomfort whilst completing the survey you can of course stop completing the survey and you may wish to consider accessing counselling services. QUT provides for limited free counselling for research participants of QUT projects, who may

250

experience discomfort or distress as a result of their participation in the research. Should you wish to access this service please contact the Clinic Receptionist of the QUT Psychology Clinic on +61 7 3138 0999. Please indicate to the receptionist that you are a research participant. Confidentiality All comments and responses will be treated confidentially. The names of individual persons are not required in any of the responses. Consent to Participate The submission an online survey is accepted as an indication of your consent to participate in this project. Questions / further information about the project Please contact the researcher team members named above to have any questions answered or if you require further information about the project. Concerns / complaints regarding the conduct of the project QUT is committed to researcher integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project you may contact the QUT Research Ethics Coordinator on +61 7 3138 2091 or ethicscontact@qut.edu.au. The Research Ethics Coordinator is not connected with the research project and can facilitate a resolution to your concern in an impartial manner.

251

13.12 Final version of survey instrument (Word document version)

PHASE THREE SURVEY INSTRUMENT TO BE CIRCULATED TO

ACCCN MEMBERS

Please complete the following demographic information

1. Age _ _ (years)

2. Sex Male/Female

3. Where do you live? ACT/NSW/QLD/VIC/TAS/WA/SA/NT/NZ/Other

(specify country)

4. How many years of experience do you have working as a nurse? _ _

5. How many years of experience do you have working as a nurse in an intensive or

critical care unit? _ _

6. What is your highest qualification in nursing?

Hospital certificate, undergraduate, postgraduate

7. Have you undertaken further formal education (graduate certificate/diploma etc.)

in intensive or critical care nursing?

No, Yes, Currently studying

252

8. How would you describe your current workplace:

General ICU (integrated medical/surgical including ICU managed HDU)

ICU/CCU/HDU (integrated intensive care/coronary care/high dependency unit)

Paediatric ICU

Neonatal ICU

High Dependency/Step Down/Special care unit

Emergency Department

Staff development/education unit

Research Centre

University

Other (specify)

9. Please specify the location of your current workplace:

Major city, regional area, remote area

10. The majority of your work in your current employment is:

Clinical (direct patient care)/Nursing management (team leader / supervision of

clinicians) / Nursing management (administration)/nursing education/nursing

research/Other

11. Please specify the last time you provided direct patient care in a critical care unit

Within the last week

Within the last month

In last 3 months

253

In last 6 months

In last 12 months

In the last 1–5 years

In the last 6–10 years

More than 10 years

Never

In this project, end-of-life care is defined as the care that is provided after a

decision has been made to withhold or withdraw treatment, with death of the

patient the expected outcome.

12. Please specify the last time that you provided end-of-life care to a patient in a

critical care unit

Within the last week

Within the last month

In last 3 months

In last 6 months

In last 12 months

In the last 1–5 years

In the last 6–10 years

More than 10 years

Never

13. Approximately how many patients have you cared for at the end-of-life in the last month?

254

14. Approximately how many patients have you cared for at the end-of-life in the last 12 months?

END-OF-LIFE CARE IN THE CRITICAL CARE CONTEXT

We would like to know about the provision of end-of-life care in critical care

units. In this section, please rate the extent to which you agree or disagree with

the following statements about end-of-life care in your unit.

Please note that there are no right or wrong answers.

Scale

Strongly

Agree

Agree Neutral Disagree Strongly

Disagree

The medical staff support end-of-life care for dying patients in my unit

In my Unit, the team expresses its opinions, values and beliefs about providing care

to dying patients

All members of the healthcare team in my Unit agree with and support end-of-life

care when it is implemented

After a decision is made to withdraw life-sustaining treatment, less time is spent with

the patient by the critical care medical team

After a decision is made to withdraw life-sustaining treatment, nurses are left to

manage the care of the patient

In my Unit, end-of-life care is mostly allocated to junior nurses

In my Unit, nurses view the provision of end-of-life care as an important component

of critical care nursing

The physical environment of my Unit is ideal for providing end-of-life care

255

Private rooms are available to care for the patient at the end-of-life

My Unit is adequately equipped (chairs/recliners/waiting room etc.) to support the

comfort needs of the family during end-of-life care

My Unit is adequately staffed for providing the needs of dying patients and their

families

When a patient dies in my Unit, nurses have sufficient time to spend with the family

There are policies/guidelines to assist in the delivery of end-of-life care in my Unit

Lack of staff results in no ‘down time’ between the death of a patient and the

admission of the next patient

When a patient dies in my Unit, families have sufficient time to spend with the

patient

In my Unit, families are involved in decisions about the dying patient

In my Unit, when a diagnosis with a likely poor outcome is made, families are

informed of end-of-life care options

In my Unit, withdrawal of treatment is delayed whilst waiting for relatives of the

patient to arrive

In my Unit, the patient’s preferences (or their likely preference according to family)

for end-of-life care are considered

In my Unit, families are given adequate time to consider decisions to

withhold/withdraw treatment for the patient (their relative)

In my Unit, family meetings with the health care team are held to resolve differing

views about a patient’s treatment

I have received in-service education that assists me to support and communicate with

families of dying patients

I have received in-service education that assists me to provide care and manage

patient symptoms at the end-of-life

256

My undergraduate nursing education included end-of-life care content

My postgraduate nursing education included end-of-life care content relevant to the

critical care context

I draw on evidence from the literature to provide end-of-life care to patients and their

families

I feel adequately prepared to care for patients at the end-of-life

I don’t know what to say to the family of a patient at the end-of-life

I have limited knowledge of symptom management for patient’s at the end-of-life

I am often exposed to death in the critical care environment

I have learnt from other nurses how to provide end-of-life care

Analgesia should be titrated to keep the patient comfortable even if this hastens death

Pressure area care should continue to be provided for a patient at the end-of-life

End-of-life care should include care of the patient’s family

Abrupt withdrawal of mechanical ventilation may contribute to dyspnoea

When withdrawing mechanical ventilation at the end-of-life, ventilator weaning

(such as reductions in FiO2 and PEEP) should occur prior to extubating the patient or

placing a patient on a T-piece

Intravenous fluids should be provided to maintain hydration for the patient at the

end-of-life

The provision of end-of-life care requires emotional detachment

It is OK to cry when a patient dies

When a patient dies in my Unit, counselling is available if I need it

In my Unit, nurses are expected to cope with death and dying

I feel supported when caring for a patient at the end-of-life

257

After caring for a patient at the end-of-life, my colleagues will ask me if I am OK

My supervisors/managers ensure staff caring for patients at the end-of-life are

supported

My colleagues appreciate the stressors associated with caring for patients at the end-

of-life

My supervisors/managers provide guidance that assists me to provide end-of-life

care

I feel a sense of personal failure when a patient dies

It is a privilege to care for a patient at the end-of-life and their family

I can influence the patient and family’s experience of end-of-life care

My own personal experiences of death have influenced the care I provide to patients

at the end-of-life

End-of-life care is as important as curative care in the critical care environment

After a decision is made to withdraw treatment, I spend less time with the patient

Patients at the end-of-life require little nursing care

End-of-life care requires advanced nursing skills

In my unit, nurses that know the patient and family are allocated to care for the

patient to maintain continuity of care

In my unit, a plan of care for the patient is clearly documented by the medical team

after a decision to withhold or withdraw treatment is made

In my unit, the critical care team and the patient’s primary medical or surgical team

agree on a plan of care

Following a decision to withhold or withdraw treatment, patients remain in my Unit

for end-of-life care

I am very much afraid to die

The thought of death seldom enters my mind

258

It doesn’t make me nervous when people talk about death

I dread to think about having an operation

I am not at all afraid to die

I am not particularly afraid of getting cancer

The thought of death never bothers me

I am often distressed by the way that time flies so very rapidly

I fear dying a painful death

The subject of life after death troubles me greatly

I am really scared of having a heart attack

I often think about how short life really is

I shudder when I hear people talking about world war III

The sight of a dead body is horrifying to me

I feel that the future holds nothing for me to fear

END-OF-LIFE CARE PRACTICES

We would like to know what activities YOU undertake when caring for a

patient at the end-of-life in an intensive care or critical care unit. Please note, in

this section we are interested in your own practices when caring for people at

end-of-life, not those of your unit more generally. There are no right or wrong

answers.

For each statement listed below, please rate how often YOU engage in the

following practices when providing end-of-life care in the intensive/critical care

setting.

259

Scale

Always Often Sometimes Rarely Never

Provide the family with options about the care of the patient

Ask the family if they would like to be involved in care of the patient

Ask the family their preferences for the patient’s care

Attend family meetings with the intensive care team

Ask the family about their knowledge of the patient’s wishes for end-of-life care

Answer the family’s questions about the patient’s condition

Fully inform the family about all aspects of the plan of care

Explain all interventions undertaken in the care of the patient

Provide information to the family about the dying process

Explain to the family what is happening to the patient

Document the family’s preference for the care of the patient

Introduce the nursing staff on the oncoming shift to the patient and family

Handover the family’s preferences for the patient’s end-of-life care to the oncoming

nurse

Ask the intensive care team to document the plan of care/treatment orders

Liaise with other members of the health care team to meet identified needs of the

patient and family

Encourage the family to reminisce about the patient

Listen to the family reminisce about the patient

Permit family to visit outside of regular visiting hours

Encourage the family to talk to the patient

260

Encourage the family to touch the patient

Provide the family with access to a private room to congregate

Ask family members if there is anyone they would like to be there

Arrange for a counsellor or social worker to meet with the family

Stay in the room with the patient and family, if the patient and family wants

Ask the family about their feelings

Ask the family how they are coping

Provide opportunities for the family to create positive memories

Move the patient to a more private area of the unit

Create a calm, peaceful environment

Turn off monitoring equipment

Silence alarms on all equipment

Adjust (dim or turn off) the lighting around the patient’s bed space

Play music

Put up photographs of the patient, family members and/or pets

Cluster/coordinate care to minimise interruptions and provide privacy for the patient

and family

Draw curtains to provide privacy for the patient and family

Place chairs around the bed for the family

Administer fluids to the patient to maintain hydration

Keep the patient sedated, if appropriate

Administer analgesia via intravenous infusion

Provide care to maintain patient hygiene

261

Provide pressure area care

Maintain the patient’s airway using oro/endotracheal suction

Remove the endotracheal tube if the patient is breathing spontaneously

Initiate (or implement) weaning from mechanical ventilation once the decision to

withdraw mechanical ventilation has been made (i.e. reduce FiO2, rate and/or tidal

volume)

Administer medications to prevent/minimise dyspnoea and respiratory secretions

Regularly obtain and document patient vital signs during the provision of end-of-life

care

Ask the family about spiritual needs

Offer to arrange for a spiritual advisor or pastoral care representative to visit the

patient and family

Ask the family about cultural needs

Support the family to participate in cultural practices related to dying

Open-ended question: List the 5 most important practices in the provision of end-

of-life care.

Thank you for completing this survey. In the space below, please add any

additional comments about end-of-life care

262

13.13 Participant responses as a percentage of the sample, means and standard deviations for each item in the Critical care nurses’ end-

of-life care practices section of the survey instrument (n = 392) Item Never Rarely Sometimes Often Always M (SD)

Provide the family with options about the care of the patient 0.3 5.4 21.4 44.4 28.6 4.0 (0.9)

Ask the family if they would like to be involved in the care of the patient 0.5 4.3 19.6 36.0 39.5 4.1 (0.9)

Ask the family their preferences for the patient’s care 0.5 5.1 19.4 40.1 34.9 4.0 (0.9)

Attend family meetings with the medical team 0.5 3.8 15.8 32.1 47.7 4.2 (0.9)

Ask the family about their knowledge of the patient’s wishes for EOLC 0.8 3.8 16.6 38.0 40.8 4.1 (0.9)

Answer the family’s questions about the patient’s condition 0 0.5 3.3 30.4 65.8 4.6 (0.6)

Fully inform the family about all aspects of the plan of care 0 1.0 4.8 32.9 61.2 4.5 (0.6)

Explain all interventions undertaken in the care of the patient 0 0.5 4.6 34.7 60.2 4.6 (0.6)

Provide information to the family about the dying process 1.0 3.1 16.3 43.4 36.2 4.1 (0.9)

Explain to the family what is happening to the patient 0 1.8 4.8 39.3 54.1 4.5 (0.7)

Document the family’s preference for the care of the patient 1.3 5.4 14.3 33.9 45.2 4.2 (1.0)

Introduce the nursing staff on the oncoming shift to the patient and family 0 0.5 2.3 21.4 75.8 4.7 (0.5)

Handover the family’s preference for the patient’s EOLC to the oncoming nurse 0.3 1.0 4.1 21.7 73.0 4.7 (0.6)

Ask the intensive care team to document the plan of care/treatment orders 0 4.1 9.2 29.1 57.7 4.4 (0.8)

Liaise with other members of the health care team to meet identified needs of the patient and family 0 1.3 9.2 38.3 51.3 4.4 (0.7)

Encourage the family to reminisce about the patient 1.0 6.4 30.9 43.4 18.4 3.7 (0.9)

Listen to the family reminisce about the patient 0 2.0 15.6 44.4 38.0 4.2 (0.8)

Permit the family to visit outside of regular visiting hours 0.3 0 2.3 21.2 76.3 4.7 (0.5)

Encourage the family to talk to the patient 0 0.5 5.9 25.0 68.6 4.6 (0.6)

Encourage the family to touch the patient 0 0.3 3.3 24.2 72.2 4.7 (0.6)

Provide the family with access to a private room to congregate 0.3 2.6 13.8 39.3 44.1 4.2 (0.8)

Ask family members if there is anyone they would like to be there 0 0.8 4.1 31.1 64.0 4.6 (0.6)

263

Item Never Rarely Sometimes Often Always M (SD)

Arrange for a counsellor or social worker to meet with the family 0.5 2.6 14.0 37.8 45.2 4.2 (0.8)

Stay in the room with the patient and family, if the patient and family wants 0 1.5 11.2 33.9 53.3 4.4 (0.8)

Ask the family about their feelings 0.8 8.2 26.3 41.8 23.0 3.8 (0.9)

Ask the family how they are coping 0.3 2.0 14.5 48.5 34.7 4.2 (0.8)

Provide opportunities for the family to create positive memories 0 5.9 21.4 39.8 32.9 4.0 (0.9)

Move the patient to a more private area of the unit 1.0 8.2 23.0 48.0 19.9 3.8 (0.9)

Create a calm, peaceful environment 0.5 1.0 7.7 49.7 41.1 4.3 (0.7)

Turn off monitoring equipment 0.8 2.3 20.7 43.9 32.4 4.1 (0.8)

Silence alarms on all equipment 0 0.5 5.6 26.5 67.3 4.6 (0.6)

Adjust (dim or turn off) the lighting around the patient’s bed space 0.5 3.1 12.2 49.7 34.4 4.2 (0.8)

Play music 6.1 13.0 46.7 28.1 6.1 3.2 (0.9)

Put up photographs of the patient, family members and/or pets 2.6 4.6 26.0 49.0 17.9 3.8 (0.9)

Cluster/coordinate care to minimise interruptions and provide privacy for the patient and family 0.3 1.5 3.8 45.4 49.0 4.4 (0.7)

Draw curtains to provide privacy for the patient and family 0 0.5 4.1 26.3 69.1 4.6 (0.6)

Place chairs around the bed for the family 0 0.5 1.3 23.2 75.0 4.7 (0.5)

Administer fluids to the patient to maintain hydration 2.8 8.2 30.6 38.3 20.2 3.7 (1.0)

Keep the patient sedated, if appropriate 0 0.5 9.2 43.1 47.2 4.4 (0.7)

Administer analgesia via intravenous infusion 0 0.8 6.1 46.7 46.4 4.4 (0.6)

Provide care to maintain patient hygiene 0 0.3 2.8 21.2 75.8 4.7 (0.5)

Provide pressure area care 0.3 1.3 6.4 32.1 59.9 4.5 (0.7)

Maintain the patient’s airway using oro/endotracheal suction 1.5 2.8 16.6 40.1 39.0 4.1 (0.9)

Remove the endotracheal tube if the patient is breathing spontaneously 1.8 1.5 19.6 44.9 32.1 4.0 (0.9)

Initiate (or implement) weaning from mechanical ventilation once the decision to withdraw

mechanical ventilation has been made (i.e. reduce FiO2, rate and/or tidal volume) 5.1 8.2 26.5 38.3 21.9 3.6 (1.1)

Administer medications to prevent/minimise dyspnoea and respiratory sections 1.3 2.6 16.6 44.4 35.2 4.1 (0.9)

264

Item Never Rarely Sometimes Often Always M (SD)

Regularly obtain and document patient vital signs during the provision of EOLC 3.1 13.3 28.3 31.4 24.0 3.6 (1.1)

Ask the family about spiritual needs 0.5 2.8 9.4 29.3 57.9 4.4 (0.8)

Offer to arrange for a spiritual advisor or pastoral care representative to visit the patient and family 0.8 1.0 7.1 31.1 59.9 4.5 (0.7)

Ask the family about cultural needs 0.8 4.3 16.3 29.3 49.2 4.2 (0.9)

Support the family to participate in cultural practices related to dying 0.8 3.3 13.5 32.9 49.5 4.3 (0.9)

Note. M = mean; SD = standard deviation; EOLC = end-of-life care.

265

13.14 Participant responses as a percentage of the sample, means and standard deviations for each item in the Factors influencing the

provision of end-of-life care section of the survey instrument (n = 392)

Item SD D N A SA M (SD)

The medical staff support EOLC for dying patients in my unit 0.8 8.7 9.2 54.1 27.3 4.0 (0.9)

In my Unit, the team expresses its opinions, values and beliefs about providing care to dying patients 1.3 8.7 10.2 58.2 21.7 3.9 (0.9)

All members of the healthcare team in my Unit agree with and support EOLC when it is implemented 1.3 10.5 12.2 53.8 22.2 3.9 (0.9)

After a decision is made to withdraw life-sustaining treatment, less time is spent with the patient by the critical care medical team 7.4 25.0 12.0 42.1 13.5 3.3 (1.2)

After a decision is made to withdraw life-sustaining treatment, nurses are left to manage the care of the patient 6.4 27.6 9.9 40.3 15.8 3.3 (1.2)

In my Unit, EOLC is mostly allocated to junior nurses 36.2 48.5 10.7 2.8 1.8 1.9 (0.9)

In my Unit, nurses view the provision of EOLC as an important component of critical care nursing 1.5 5.1 12.8 44.1 36.5 4.1 (0.9)

The physical environment of my Unit is ideal for providing EOLC 16.8 43.4 20.4 16.6 2.8 2.5 (1.0)

Private rooms are available to care for the patient at the EOL 8.2 19.9 14.8 44.6 12.5 3.3 (1.2)

My Unit is adequately equipped (chairs/recliners/waiting room etc.) to support the comfort needs of the family during EOLC 14.5 33.2 14.8 28.6 8.9 2.8 (1.2)

My Unit is adequately staffed for providing the needs of dying patients and their families 5.4 24.5 19.1 42.6 8.4 3.2 (1.1)

When a patient dies in my Unit, nurses have sufficient time to spend with the family 3.8 21.4 14.0 50.5 10.2 3.4 (1.1)

There are policies/guidelines to assist in the delivery of EOLC in my Unit 6.4 28.6 22.2 34.2 8.7 3.1 (1.1)

Lack of staff results in no ‘down time’ between the death of a patient and the admission of the next patient 3.8 24.5 23.2 34.9 13.5 3.3 (1.1)

When a patient dies in my Unit, families have sufficient time to spend with the patient 0.8 5.1 12.0 61.5 20.7 4.0 (0.8)

In my Unit, families are involved in decisions about the dying patient 1.3 1.8 6.6 59.4 30.9 4.2 (0.7)

In my Unit, when a diagnosis with a likely poor outcome is made, families are informed of EOLC options 1.3 8.7 10.7 54.1 25.3 3.9 (0.9)

In my Unit, withdrawal of treatment is delayed whilst waiting for relatives of the patient to arrive 0.8 2.6 5.6 61.2 29.8 4.2 (0.7)

In my Unit, the patient’s preferences (or their likely preference according to family) for EOLC are considered 1.3 2.6 10.2 57.4 28.6 4.1 (0.8)

In my Unit, families are given adequate time to consider decisions to withhold/withdraw treatment for the patient (their relative) 0.5 2.6 11.5 60.5 25.0 4.1 (0.7)

In my Unit, family meetings with the health care team are held to resolve differing views about a patient’s treatment 1.0 3.6 7.9 56.9 30.6 4.1 (0.8)

I have received in-service education that assists me to support and communicate with families of dying patients 13.0 40.8 12.0 28.1 6.1 2.7 (1.2)

266

Item SD D N A SA M (SD)

I have received in-service education that assists me to provide care and manage patient symptoms at the EOL (n = 390) 12.6 42.1 13.6 25.6 6.2 2.7 (1.2)

My undergraduate nursing education included EOLC content (n = 390) 9.0 27.2 22.1 36.2 5.6 3.0 (1.1)

My postgraduate nursing education included EOLC content relevant to the critical care context (n = 390) 8.5 25.1 22.3 33.6 10.5 3.1 (1.2)

I draw on evidence from the literature to provide EOLC to patients and their families 2.3 17.1 20.7 48.2 11.7 3.5 (1.0)

I feel adequately prepared to care for patients at the EOL (n = 389) 0.5 6.9 13.6 49.1 29.8 4.0 (0.9)

I don’t know what to say to the family of a patient at the EOL (n = 391) 26.8 45.8 15.1 9.2 3.1 2.2 (1.0)

I have limited knowledge of symptom management for patient’s at the EOL (n = 391) 28.9 51.7 8.2 9.0 2.3 2.0 (1.0)

I am often exposed to death in the critical care environment 2.3 9.2 14.3 44.9 29.3 3.9 (1.0)

I have learnt from other nurses how to provide EOLC 2.3 6.6 10.7 63.5 16.8 3.9 (0.9)

Analgesia should be titrated to keep the patient comfortable even if this hastens death (n = 391) 0 0.5 2.6 36.1 60.9 4.6 (0.6)

Pressure area care should continue to be provided for a patient at the EOL (n = 391) 0 1.8 7.9 39.1 51.2 4.4 (0.7)

EOLC should include care of the patient’s family (n = 391) 0 0 0.5 24.6 74.9 4.7 (0.5)

Abrupt withdrawal of mechanical ventilation may contribute to dyspnoea (n = 387) 0.5 4.9 16.5 50.9 27.1 4.0 (0.8)

When withdrawing mechanical ventilation at the end-of-life, ventilator weaning (such as reductions in FiO2 and PEEP) should occur

prior to extubating the patient or placing a patient on a T-piece (n = 391) 5.1 22.0 29.2 31.7 12.0 3.2 (1.1)

Intravenous fluids should be provided to maintain hydration for the patient at the EOL (n = 390) 6.2 15.9 30.3 36.7 11.0 3.3 (1.1)

The provision of EOLC requires emotional detachment 23.0 51.5 11.5 10.2 3.8 2.2 (1.0)

It is OK to cry when a patient dies (n = 390) 1.3 1.5 10.0 52.3 34.9 4.2 (0.8)

When a patient dies in my Unit, counselling is available if I need it (n = 390) 4.9 12.8 14.6 56.2 11.5 3.6 (1.0)

In my Unit, nurses are expected to cope with death and dying (n = 391) 0.8 14.8 12.5 59.6 12.3 3.7 (0.9)

I feel supported when caring for a patient at the EOL (n = 390) 2.1 8.7 21.3 54.9 13.1 3.7 (0.9)

After caring for a patient at the EOL, my colleagues will ask me if I am OK (n = 391) 1.3 6.1 11.5 59.1 22.0 3.9 (0.8)

My supervisors/managers ensure staff caring for patients at the EOL are supported (n = 390) 3.6 16.9 23.3 45.6 10.5 3.4 (1.0)

My colleagues appreciate the stressors associated with caring for patients at the EOL (n = 391) 0.8 4.3 12.5 64.7 17.6 3.9 (0.7)

My supervisors/managers provide guidance that assists me to provide EOLC (n = 390) 5.9 21.8 28.7 37.9 5.6 3.2 (1.0)

267

Item SD D N A SA M (SD)

I feel a sense of personal failure when a patient dies (n = 391) 46.0 44.8 5.9 2.0 1.3 1.7 (0.8)

It is a privilege to care for a patient at the EOL and their family 0.8 1.0 7.1 33.9 57.1 4.5 (0.7)

I can influence the patient and family’s experience of EOLC (n = 391) 0.8 1.0 3.1 41.9 53.2 4.5 (0.7)

My own personal experiences of death have influenced the care I provide to patients at the EOL 1.0 9.4 17.3 37.2 34.9 4.0 (1.0)

EOLC is as important as curative care in the critical care environment (n = 389) 0.5 1.0 4.6 37.8 56.0 4.5 (0.7)

After a decision is made to withdraw treatment, I spend less time with the patient 37.0 43.1 8.2 10.7 1.0 2.0 (1.0)

Patients at the EOL require little nursing care (n = 388) 50.5 40.5 4.4 2.3 2.3 1.7 (0.9)

EOLC requires advanced nursing skills (n = 390) 1.8 12.6 18.5 40.3 26.9 3.8 (1.0)

In my unit, nurses that know the patient and family are allocated to care for the patient to maintain continuity of care 2.6 25.3 27.8 37.2 7.1 3.2 (1.0)

In my unit, a plan of care for the patient is clearly documented by the medical team after a decision to withhold or withdraw treatment

is made 2.6 21.2 16.6 46.2 13.5 3.5 (1.1)

In my unit, the critical care team and the patient’s primary medical or surgical team agree on a plan of care 3.3 11.5 25.8 49.7 9.7 3.5 (0.9)

Following a decision to withhold or withdraw treatment, patients remain in my Unit for EOLC 0.8 13.8 30.9 44.9 9.7 3.5 (0.9)

I am very much afraid to die 15.1 41.6 27.6 10.5 5.4 2.5 (1.0)

The thought of death seldom enters my mind 7.1 49.2 21.9 19.9 1.8 2.6 (0.9)

It doesn’t make me nervous when people talk about death 1.5 7.9 5.9 53.8 30.9 4.1 (0.9)

I dread to think about having an operation 17.3 41.1 18.9 18.9 3.8 2.5 (1.1)

I am not at all afraid to die 5.6 31.1 33.2 23.5 6.6 2.9 (1.0)

I am not particularly afraid of getting cancer 12.2 46.7 20.9 18.1 2.0 2.5 (1.0)

The thought of death never bothers me 7.1 49.7 23.7 16.8 2.6 2.6 (0.9)

I am often distressed by the way that time flies so very rapidly 5.9 31.4 33.2 23.7 5.9 2.9 (1.0)

I fear dying a painful death 1.5 18.4 18.9 43.4 17.9 3.6 (1.0)

The subject of life after death troubles me greatly 29.4 44.9 23.0 1.8 0.8 2.0 (0.8)

I am really scared of having a heart attack 13.0 46.7 26.5 12.5 1.3 2.4 (0.9)

I often think about how short life really is 5.1 21.2 21.9 42.1 9.7 3.3 (1.1)

268

Item SD D N A SA M (SD)

I shudder when I hear people talking about world war III 21.7 38.8 25.5 11.2 2.8 2.4 (1.0)

The sight of a dead body is horrifying to me 52.6 42.3 4.1 0.3 0.8 1.5 (0.7)

I feel that the future holds nothing for me to fear 7.1 33.7 37.5 17.1 4.6 2.8 (1.0)

Note. SD = strongly disagree; D = disagree; N = neutral; A = agree; SA = strongly agree; M = mean; SD = standard deviation; EOL = end of life; EOLC = end-of-life care.

269

13.15 Structure matrix for exploratory factor analysis with oblique rotation—Critical care nurses’ end-of-life care practices

Factor

1 2 3 4 5 6 7

Explain all interventions undertaken in the care of the patient .76 .22 .23 .27 .25 .38 –.27

Fully inform the family about all aspects of the plan of care .75 .22 .26 .36 .20 .49 –.21

Explain to the family what is happening to the patient .74 .18 .33 .32 .25 .35 –.14

Answer the family's questions about the patient's condition .68 .33 .19 .28 .26 .42 –.24

Introduce the nursing staff on the oncoming shift to the patient and family .65 .34 .36 .27 .22 .56 –.41

Liaise with other members of the health care team to meet identified needs of the patient and family .59 .28 .35 .39 .31 .49 –.07

Handover the family’s preferences for the patient’s end-of-life care to the oncoming nurse .59 .40 .29 .45 .16 .53 –.34

Provide information to the family about the dying process .57 .15 .32 .39 .14 .33 .02

Ask the intensive/critical care team to document the plan of care/treatment orders .48 .23 .18 .36 .20 .43 –.15

Attend family meetings with the intensive/critical care team .40 .33 .12 .22 .07 .32 –.05

Adjust (dim or turn off) the lighting around the patient’s bed space .23 .68 .28 .20 .20 .33 –.01

Silence alarms on all equipment .25 .64 .06 .15 .06 .25 –.24

Create a calm, peaceful environment .26 .61 .33 .25 .22 .33 .14

Place chairs around the bed for the family .49 .61 .32 .13 .28 .52 –.36

Cluster/coordinate care to minimise interruptions and provide privacy for the patient and family .32 .56 .47 .14 .25 .50 –.10

Draw curtains to provide privacy for the patient and family .40 .56 .28 .13 .27 .44 –.33

Turn off monitoring equipment .11 .53 .12 .15 –.01 .157 –.19

Remove the endotracheal tube if the patient is breathing spontaneously .21 .43 .07 .16 .12 .11 –.14

Permit family to visit outside of regular visiting hours .37 .40 .19 .19 .20 .39 –.29

Move the patient to a more private area of the unit .10 .39 .28 .16 .20 .209 .12

Administer medications to prevent/minimise dyspnoea and respiratory sections .20 .34 .12 .28 .28 .21 –.19

270

Factor

1 2 3 4 5 6 7

Provide the family with access to a private room to congregate .26 .31 .25 .15 .18 .25 .12

Provide opportunities for the family to create positive memories .42 .23 .71 .38 .36 .54 –.02

Encourage the family to reminisce about the patient .31 .14 .69 .18 .18 .33 –.14

Ask the family about their feelings .32 .21 .65 .42 .23 .45 –.15

Ask the family how they are coping .31 .28 .60 .36 .30 .46 –.17

Listen to the family reminisce about the patient .35 .23 .57 .25 .24 .45 –.17

Put up photographs of the patient, family members and/or pets .17 .33 .56 .09 .25 .35 .06

Play music .18 .27 .54 .17 .17 .34 .20

Ask the family their preferences for the patient's care .36 .26 .26 .78 .18 .35 –.09

Ask the family if they would like to be involved in the care of the patient .42 .22 .34 .75 .24 .38 –.04

Provide the family with options about the care of the patient .40 .20 .31 .67 .22 .36 .04

Ask the family about their knowledge of the patient's wishes for EOLC .47 .35 .24 .58 .10 .43 –.10

Document the family’s preference for the care of the patient .51 .17 .37 .54 .17 .46 –.09

Initiate (or implement) weaning from mechanical ventilation… .10 .25 .10 .34 .24 .19 –.05

Maintain the patient’s airway using oro/endotracheal suction .25 .17 .21 .18 .60 .17 –.02

Keep the patient sedated, if appropriate .32 .38 .14 .21 .55 .37 –.30

Administer fluids to the patient to maintain hydration .08 –.03 .13 .15 .55 .10 –.11

Administer analgesia via intravenous infusion .25 .42 .07 .16 .49 .35 –.32

Provide pressure area care .44 .26 .29 .01 .49 .24 –.13

Provide care to maintain patient hygiene .47 .38 .24 .00 .48 .40 –.26

Regularly obtain and document patient vital signs during the provision of EOLC .12 –.03 .25 .08 .46 .20 .03

Ask the family about spiritual needs .38 .24 .42 .30 .23 .86 –.12

Offer to arrange for a spiritual advisor or pastoral care representative to visit the patient and family .36 .24 .33 .26 .15 .84 –.14

Ask the family about cultural needs .40 .23 .42 .33 .28 .83 –.05

271

Factor

1 2 3 4 5 6 7

Support the family to participate in cultural practices related to dying .44 .28 .43 .33 .28 .83 –.06

Ask family members if there is anyone they would like to be there .49 .35 .39 .28 .21 .63 –.16

Stay in the room with the patient and family, if the patient and family wants .42 .30 .45 .26 .22 .60 –.17

Arrange for a counsellor or social worker to meet with the family .38 .28 .36 .17 .20 .44 .02

Encourage the family to touch the patient .41 .34 .39 .17 .32 .38 –.69

Encourage the family to talk to the patient .32 .29 .39 .12 .34 .31 –.60

Note. EOLC = end-of-life care.

272

13.16 Structure matrix for exploratory factor analysis with oblique rotation—Factors influencing the provision of end-of-life care

Factor

1 2 3 4 5 6 7 8

My supervisors/managers ensure staff caring for patients at the EOL are supported .83 –.01 –.15 .22 –.23 .20 .20 –.20

My supervisors/managers provide guidance that assists me to provide EOLC .77 –.08 –.19 .28 –.21 .24 .21 –.09

I feel supported when caring for a patient at the EOL .75 .08 –.20 .31 –.19 .30 .15 –.26

After caring for a patient at the EOL, my colleagues will ask me if I am OK .69 .23 –.19 .17 –.06 .24 .07 –.21

My colleagues appreciate the stressors associated with caring for patients at the EOL .65 .22 –.25 .18 –.16 .33 .02 –.30

When a patient dies in my Unit, counselling is available if I need it .55 –.14 –.03 .21 –.12 .20 .29 –.03

In my Unit, nurses view the provision of EOL care as an important component of critical care nursing .38 .23 –.19 .25 –.08 .28 .08 –.18

Lack of staff results in no ‘down time’ between the death of a patient and the admission of the next patient .34 –.13 –.18 .23 –.27 –.02 .00 –.08

EOLC is as important as curative care in the critical care environment –.01 .67 –.16 –.06 .03 –.06 .12 –.27

I can influence the patient and family’s experience of EOLC .07 .61 –.10 .01 –.05 –.04 .03 –.34

EOLC should include care of the patient’s family .01 .58 –.22 –.12 .07 .08 –.01 –.21

It is a privilege to care for a patient at the EOL and their family .14 .58 –.15 .02 –.04 –.10 .10 –.33

Analgesia should be titrated to keep the patient comfortable even if this hastens death .06 .40 –.15 –.03 .06 .00 .03 –.23

My own personal experiences of death have influenced the care I provide to patients at the EOL and their family –.05 .38 –.04 .03 .03 –.04 .10 –.14

It is OK to cry when a patient dies .08 .32 –.13 .01 –.14 –.05 .05 –.21

Pressure area care should continue to be provided for a patient at the EOL –.01 .32 –.08 –.02 .01 .07 .05 –.13

Abrupt withdrawal of mechanical ventilation may contribute to dyspnoea –.00 .28 –.22 .04 .11 –.03 .05 –.20

I am often exposed to death in the critical care environment .04 .24 –.23 .01 .17 .06 –.04 –.10

In my Unit, families are given adequate time to consider decisions to withhold /withdraw treatment for the patient .17 .06 –.74 .18 –.16 .20 .07 –.11

In my Unit, family meetings with the health care team are held to resolve differing views about a patient’s treatment .28 .03 –.73 .24 –.16 .28 .06 –.14

In my Unit, the patient’s preferences (or their likely preference according to family) for EOLC are considered .28 .07 –.61 .15 –.11 .41 .09 –.03

273

Factor

1 2 3 4 5 6 7 8

In my Unit, families are involved in decisions about the dying patient .206 .149 –.59 .239 –.155 .228 –.01 –.04

In my Unit, withdrawal of treatment is delayed whilst waiting for relatives of the patient to arrive .01 .227 –.509 .098 .045 .106 –.04 –.105

When a patient dies in my Unit, families have sufficient time to spend with the patient .37 .01 –.448 .407 –.27 .22 –.14 –.18

The physical environment of my Unit is ideal for providing EOLC .14 –.085 –.05 .72 –.118 .08 .088 –.04

My Unit is adequately equipped to support the comfort needs of the family during EOLC .13 –.00 –.189 .71 –.10 .16 .16 –.058

My Unit is adequately staffed for providing the needs of dying patients and their families .377 –.12 –.325 .68 –.185 .189 .049 –.17

Private rooms are available to care for the patient at the EOL .18 .07 –.097 .675 .026 .14 .08 –.018

When a patient dies in my Unit, nurses have sufficient time to spend with the family .46 –.079 –.409 .52 –.28 .20 –.11 –.396

In my Unit, nurses that know the patient and family are allocated to care for the patient to maintain continuity… .358 .00 –.16 .40 –.26 .257 .198 –.00

Following a decision to withhold or withdraw treatment, patients remain in my Unit for EOLC .21 .21 –.20 .318 –.126 .18 .04 –.13

After a decision is made to withdraw treatment, less time is spent with the patient by the critical care medical team –.22 .06 .12 –.189 .757 –.16 –.157 .04

After a decision is made to withdraw treatment, nurses are left to manage the care of the patient –.238 .026 .18 –.149 .727 –.18 –.088 .04

In my Unit, EOLC is mostly allocated to junior nurses –.22 –.13 .13 –.129 .35 –.18 .029 .22

After a decision is made to withdraw treatment, I spend less time with the patient .01 –.299 .20 .00 .33 .146 –.31 .27

In my Unit, nurses are expected to cope with death and dying .25 –.229 .03 .13 –.278 –.075 .00 .01

The medical staff support EOLC for dying patients in my Unit .29 .06 –.33 .16 –.175 .60 –.015 –.156

In my Unit, when a diagnosis with a likely poor outcome is made, families are informed of EOLC options .25 .025 –.425 .195 –.236 .59 .10 –.10

In my Unit, the team expresses its opinions, values and beliefs about providing care to dying patients .368 .035 –.16 .20 –.105 .57 .07 –.095

All members of the healthcare team in my Unit agree with and support EOLC when it is implemented .27 .089 –.248 .317 –.05 .538 –.119 –.096

In my Unit, a plan of care for the patient is clearly documented by the medical team… .27 –.047 –.31 .285 –.31 .50 .04 –.15

In my Unit, the critical care team and the patient’s primary medical or surgical team agree on a plan of care .278 .016 –.34 .238 –.21 .497 .05 –.107

There are policies/guidelines to assist in the delivery of EOLC in my Unit .307 –.01 –.10 .258 –.106 .347 .32 –.04

The provision of EOLC requires emotional detachment .00 .218 .01 –.06 –.19 –.25 –.085 –.159

I have learnt from other nurses how to provide EOLC .201 .180 –.198 .215 .106 .241 .026 .030

274

Factor

1 2 3 4 5 6 7 8

I have received in-service education that assists me to support and communicate with families of dying patients .324 –.090 .040 .216 –.055 .318 .627 –.302

I have received in-service education that assists me to provide care and manage patient symptoms at the EOL .289 –.117 .076 .219 –.068 .272 .597 –.324

My postgraduate nursing education included EOLC content relevant to the critical care context .246 .068 –.019 .148 –.110 .109 .461 –.115

I draw on evidence from the literature to provide EOLC to patients and their families .131 .216 –.282 .064 –.149 –.019 .457 –.263

Patients at the EOL require little nursing care –.049 .275 –.156 –.179 –.203 –.174 .336 –.132

EOLC requires advanced nursing skills .120 .301 –.069 .035 –.028 .027 .308 –.143

My undergraduate nursing education included EOLC content .170 –.008 .076 .174 –.007 .064 .306 –.026

When withdrawing mechanical ventilation at the EOL, ventilator weaning should occur… –.004 .071 –.025 –.001 .019 –.102 .262 –.055

Intravenous fluids should be provided to maintain hydration for the patient at the EOL .073 .044 –.065 .095 –.123 .080 .243 –.087

I feel adequately prepared to care for patients at the EOL .183 .288 –.163 .101 –.042 .112 .182 –.704

I don’t know what to say to the family of a patient at the EOL .141 .258 –.090 .087 –.109 .001 .142 –.662

I have limited knowledge of symptom management for patient’s at the EOL .023 .227 –.017 –.016 .050 .027 .125 –.508

I feel a sense of personal failure when a patient dies .181 .253 –.174 –.065 –.168 –.029 –.018 –.371

Note. EOL = end of life; EOLC = end-of-life care.