igniting innovation
Post on 18-Mar-2022
63 Views
Preview:
TRANSCRIPT
1
Igniting Innovation Improving Health, Creating Wealth
Nursing Technology Evidence Review Produced in collaboration with Nursing Technology Team NHS England
June 2015
2
Foreword
Health Information technology will have a profound impact on how nurses, midwives and
health visitors do their jobs. These changes will affect how and where they provide care,
how they interact with their patients and colleagues and how they measure their work and
its effectiveness. In the health care systems of high-income countries, many aspects of
nursing are increasingly supported by a wide array of technology such as electronic and
personal health records (EHRs and PHRs), biometric and telemedicine devices, and
consumer-focused wireless and wired systems. Technology is also integral to the
blueprint for the future of the the NHS; the NHS England Five Year Forward View is full of
references to technological healthcare delivery:
“Raise our game on technology”
“Exploit the information revolution”
“Accelerate useful innovation”
(NHS England 2014 4;31;32).
This brief review surveys the academic literature regarding a group of nursing-specific
technologies and applications. We have captured some of the messages arising from this
evidence and noted many gaps and unanswered questions. We are right at the beginning
of the learning curve for nursing technology in NHS and to climb it nurses must address
currently un-asked questions along with merely unanswered ones.
We hope this document will help nurses and midwives to achieve the greatest possible
improvements in the care they deliver to patients by harnessing these innovative systems
to the full.
Dr Lucy Sitton-Kent Dr Philip Miller
3
Contents
Glossary 5
1. Executive Summary: 8
2. Context
10
3. Methodology
Inclusion criteria
Exclusion criteria
Review process
11 11
12
12
4. Capabilities
13
4.1 Mobile access to digital care records
Effectiveness and accuracy
Financial impact
Impact on patients and staff
14 15 16 18
4.2 Digital images for nursing care
Effectiveness and accuracy
Combined with tele health
Imaging feet
Forensic digital imaging
Dietary intake
Legal implications
19
21
23
24
24
25
25
4.3 Remote face to face interaction
Effectiveness and accuracy
Financial impact
Patient and staff acceptability
27
28
30
33
4.4 Digital capture of clinical data at point –of-care and digitally enabled observations management Automated clinical data capture
35
37
4
Effectiveness and accuracy of automated data capture
Digital enabled management of clinical data
Effectiveness and accuracy of Early Warning Score (EWS) calculation and
response triggering
Prognosis/risk scoring, monitoring and decision-support
Effectiveness and accuracy of the generation of automated care
summaries/handovers
38
40
40
41
43
4.5 Safer care interventions
Drug administration
Blood Products
Effectiveness and accuracy of using patient identification
Theoretical work
45
47
50
52
53
4.6 Dashboards
Effectiveness and accuracy
Impact on patients and staff
55
57
60
4.7 E-Workforce deployment
Effectiveness, accuracy and skill mix
Financial impact
Self-Rostering: Recruitment, Retention, Wellbeing, Acceptability for staff
62
61
64
66
5. Limitations and learning
Methodology
Human factors
Technology
The Health technologist
69
69
71
72
73
6. Conclusion
75
7.Contacts 76
8. References 77
5
Glossary Term Abbreviation Synonyms Interpretation(s)
Automated Drug
Cabinet
ADC Automated drug cabinet: records stock
levels, patient administration and
reordering, controls access (e.g.
fingerprint), guides selection, provides
alerts and limited clinical decision
support
Adverse Drug Event ADE Event involving medication where harm
or risk of harm occurs
Bar-Code BC Machine readable pictorial
representation of patient ID or product
information
Bar Code Medicine
Administration
BCMA Administration procedure reliant on the
use of BC scanner and patient and
product specific bar-code labelling
Bluetooth Near-field communication technology
Clinical Decision
Support Systems
CDSS HIT application delivering, processing
or presenting clinical information to
assist disease management
Computerised
Physician / Provider
Order Entry
CPOE System for ordering of drugs,
investigations and other specific
procedures
Cross-matched to
transfused ratio
C:T ratio Number of units of cross-matched
blood units divided by the total number
of units actually transfused (1 is ideal;
>2.5 is considered wasteful)
Critical Care
Outreach Team
CCOT “Rapid
Response
Team”
Provision of (often nurse led) critical
care assessment and intervention
beyond confines of Intensive Care
department
Early Warning Score EWS Simple algorithm combining
conventional physiological observations
into an index to predict deterioration
and facilitate preemptive intervention
Electronic or Digital
Health Record
EHR / DHR Electronic
Medical
/Hospital /
Health
Record
Software which brings together key
clinical and administrative data using a
single interface (but may be available
on a variety of devices
Electronic
Prescribing and
Medication
Administration
EPMA Cf. BCMA
and CPOE
Linked electronic prescribing to
medicines administration record with or
without BC technology
6
E-rostering e-R Software to manage and generate
Nursing and other work schedules
Health Information
systems
HIS Data storage systems in health sector
Health Information
Technology
HIT Any information and communication
technology employed in the healthcare
sector
Informatics Study of the theory and practice of the
use of digital information
Identification ID
Intensive Therapy /
Care Unit
IT(C)U Level 3 care Mechanical ventilation
Length of Stay LOS
Non-Invasive Blood
Pressure
NIBP Via blood pressure cuff
Natural Language
Processing
NLP Computer analysis or processing of
unstructured human language e.g.
coded data derived from free-text
nursing records
Positive Patient
Identification
Safety convention requiring a patient to
identify themselves rather than simple
passive confirmation
Personal Digital
Assistant
PDA Term for hand-held digital device now
largely superseded by smart phones
Real-time Practically instantaneous or trivial delay
e.g. “automated data dash-board in
real-time”
Recognise and
Rescue Team
RRT Rapid
Response
Team
See CCOT
“Recognise and
rescue”
Principle of early detection of signs of
deterioration in patient’s clinical status
(or risk of) see EWS
Summary Care
Record
SCR Automatically generated summary of a
portion of patient EHR (generally from
Primary Care record and used in
Secondary Care)
Telecare Telecare refers to systems and
equipment designed to help disabled or
vulnerable people remain safe and
independent in their home. Alarms and
communication devices that alert a
family member, nurse or warden and/or
provide remote support to a person to
7
manage everyday tasks
Telehealth Telehealth is the delivery of health
services and information via
telecommunications technologies e.g.
Skype, smart phones, or Florence
(FLO) a personal self-monitoring and
communication tool
Telemedicine Remote consultation, examination and
diagnosis (and increasingly treatment)
via communication technology such as
video streaming
Telemedicine Video
Consultation
TVC Remote consultation between specialist
and patient using video link
Total Opportunity
Error
TOE The denominator of drug administration
errors i.e. all possible doses given
including “Stat” and PRN (as opposed
to pre-planned doses only) – used to
calculate ratio of errors
Universal Donor
Blood
UDB O negative blood which can be
transfused without rhesus and group
typing of recipient
VitalPAC Early
Warning Score
ViEWS Proprietary monitoring system
incorporating EWS principles
Visual Analogue
Scale
Conventionally 100 mm line allowing
marking of a self-reports of intensity
e.g. of pain or breathlessness
“Wrong Blood in the
Tube” errors
WBIT Sample taken from the wrong patient
and sent for laboratory processing (cf.
sample from correct patient mislabeled)
8
1. Executive Summary The NHS is embracing technology in the pursuit of efficient and effective care for patients. Nurses
have a key role to play in transforming care through such technology. The first feature to emerge
from this rapid review was the lack of high quality, accessible evidence to inform future
implementation of many nursing technologies. This absence of evidence does not mean these
technologies do not work but underlines how new much of this technology is and how much there
is to learn. This review can only present an incomplete picture of the impacts of many current
implementations. We conclude that it will remain challenging for many NHS organisations to make
timely informed decisions when choosing and implementing such technologies, particularly in
relation to
o the full financial and efficiency impacts
o costs, training and workload implications of implementation and ongoing
maintenance and development
o the ultimate effects on patients and staff
o the likely need to restructure work flows and other systems
Second, we identified a great need for discussion and agreement about the methodology used to
evaluate these health technologies. There is rarely a straightforward relationship between intended
causes and effects even when implementing the same technologies in different places. This is not
a reason to ignore research into implementation and evaluation of these technologies. Rather we
need to take a more comprehensive approach incorporating all available numerical and qualitative
data and use new theoretical models and frameworks to help guide the choice of approach.
This evidence review demonstrates that implementing and evaluating health technology is complex
and that these technologies interact with multiple systems and cannot be viewed in isolation.
Evaluation should use specific and sensitive measures to indicate which planned benefits were
realised and which were not. Many of the studies reviewed used very general quality improvement
measures over a relatively short follow-up period, these were not sensitive enough to measure the
impact of the particular technology, detect the effects of one technology being implemented with
other concurrent changes or distinguish if the technology was incompletely implemented, fully
integrated or sustained. In addition it is probable that there are studies that failed to achieve their
over-ambitious outcomes and therefore remained unpublished.
9
The third central feature of the research reviewed here is a lack of description and attention to
patient and staff involvement. Nurses and patients should be involved in the design and evaluation
of health technologies and we found few articles which focused on the experiences of users
themselves. The implementation phases of HIT are notoriously difficult and involvement of as
many stakeholders as possible at the earliest opportunity, can help to avoid or minimise problems.
Evaluations should be longitudinal, record problems as well as successes, assess interactions with
other technical and non-technical systems, be sensitive to the benefits required and be published
as widely and accessibly as possible to inform the many who will follow.
Despite these warnings the review discovered many instances of where new systems had
transformed care delivery, improved efficiency and safety and was well received by staff and
service users. The literature included:
Clear evidence of benefit from:
secure mobile access to digital care records
digital images when combined with additional clinical information
remote face to face interaction for both chronic illness and acute conditions such as stroke,
particularly in rural areas
increased accuracy and ease of automated digital capture of clinical data and continuous
measurement
increased sensitivity to some clinical events
improved availability and timeliness of information
reductions in blood and some drug administration errors
efficiency gains in resource use such as drug stock control and staff rostering
Mixed positive and negative evidence for
digitally enabled observations management
automated handovers
digital dashboards
E-roster deployment
10
2. Context In October 2012 the Prime Minister announced the establishment of a Nursing Technology Fund
(NTF) to support nurses, midwives and health visitors to make better use of digital technology in all
care settings to deliver safer and more effective and efficient care. The first round released £30m for
85 projects around 3 capabilities. The second round saw 62 organisations awarded funding totaling
almost £35M.
The capabilities for the 2nd round and the remit of this review are:-
• Mobile access to digital care records across the community (4.1)
• Digital images for nursing care (4.2)
• Remote face-to-face interaction (4.3)
• Digital capture of clinical data at point-of-care and digitally-enabled observations management
(4.4)
• Safer clinical interventions (4.5)
• Digital nursing dashboards (4.6)
• E- workforce deployment (4.7)
11
3. Methodology This rapid review is intended to address the need for a synthesis of the academic evidence base in
a selection of capability areas identified from the second round of NTF projects. To achieve this we
focused on conducting a search of published evidence and then extended this to the grey literature
to provide an overview of the contemporary position. Our objective was to review the available
evidence regarding the seven capabilities in terms of an understanding of the implementation
processes and the likely financial, patient and staff impacts.
Through this review, we aimed to synthesize literature published between 2008 and 2015 in a way
that was as rigorous as possible and useful to practitioners. Technology evolves rapidly (Jamal et
al., 2009) but key earlier material was included where pertinent. The pace of change in this area
means that our conclusions are essentially provisional but we feel justified in proposing suggestions
for evaluations and research priorities.
Literature searches were performed for each capability in the following databases: Cochrane library,
MEDLINE, CINAHL, Embase and Google Scholar. The reference sections of the selected articles
were scanned for additional papers. A hand search of key health websites was performed including
The King’s Fund and NICE. The evidence found was difficult to synthesise as a result of the lack of
settled terminology and methods with the choice of research question often reflecting the
opportunistic and naturalistic nature of many studies. While there is a need to integrate future
research and evaluations from the procurement or pre-implementation stage it is worth bearing in
mind that for many projects there simply is no baseline data as they represent the first
implementation of IT within the NHS. Such limitations are inherent but will become less important as
digital working becomes more pervasive and mature.
Inclusion criteria
The inclusion criteria initially focused on published systematic reviews in English on the Nursing
Technology used to deliver the seven capabilities in different settings. This was broadened due to a
lack of evidence and the review includes any studies where an impact had been discussed or
demonstrated such as “grey” literature (Lilford et al., 2009) and narrative commentaries. The
meaning of Nursing Technology was broadly defined to include different types of systems and tools
for data capture, information processing, decision support, management reporting and analytics,
12
telemedicine/telehealth applications, digital devices. We excluded capabilities used by patients
where there was no monitoring by a doctor or nurse and those designed for patient/ provider
education only. Technologies used by Midwives and Health Visitors were included, but most of the
literature found was concerned with nursing care in mental health, adult and paediatric contexts.
Exclusion criteria Studies written in languages other than English were excluded. Work published prior to 2008 was
excluded unless particularly relevant and not superseded or if regularly cited in later work. Some
studies were difficult to retrieve and were excluded if they could not be adequately reviewed within
the timeframe. Full-text articles were retrieved wherever possible but in a few cases abstracts were
used if sufficient detail or significant results were reported.
Review Process
The review and synthesis involved six steps:
1. Academic database and internet searches and review of taxonomies and key words
2. Review of summary records with reference to the review specification
3. An informal, iterative selection process and categorization under the seven NTF capabilities
4. Duplicates and articles reporting the same studies were removed
5. The methods used and the effects reported were evaluated for effectiveness and accuracy,
financial, staff and patient impact
6. Contextual factors and methodological challenges were identified to try and give a sense of
the reliability and applicability to the NTF capabilities specifically.
13
4. Capabilities The following sections will present syntheses of the evidence found for each of the seven
capabilities. The capabilities are not entirely discrete and there is some overlap with similar
systems providing multiple functionality.
14
Capability Details of review and search terms
Effectiveness and accuracy
Financial impact Impact for patients and staff Learning for implementation
4.1 Mobile access to digital care records
839 records
found; 364
abstracts
assessed; 21
papers included.
Search terms: Mobile working, nurs*, digital records, community mobile access, digital pens, smartphone, electronic patient care records, e-health records, personal digital assistants, tele*
The literature in this area did not focus on effectiveness in terms of quality of records.
One study reported an associated increase in amount of information recorded in notes.
Mobile access made nursing and midwifery notes more contemporaneous
Notes include access additional information and therefore can lead to improved care.
Little quantification of investment required for set up costs and maintenance.
Much anecdotal evidence about how deployment of technology increased productivity by reducing referrals, admissions, staff travel time, “no access” visits and data duplication.
Patient benefits were largely indirect e.g. nurses reported that they have increasing amount of face to face time.
Patients reporting that they liked to see what the midwives write. Such “open” records were confined to midwifery
Nurses were inventive with HIT using it to provide additional information for patients which was positively received.
Nurses and Midwives were generally positive about HIT including the ability to prepare by accessing information prior visits, reductions in duplication and increased time interacting with the patients.
Implementation required
good project management,
leadership, support, skills,
clinical engagement, good
connectivity, IT support and
secure systems.
If supported, staff will find
novel ways to use the
technology.
Important to map out how to move from paper to electronic records without loss of function
Applications reviewed potentially release significant resources, the challenge is how to measure this and recognize all implementation costs.
Considerable scope to free up nursing time for other activities valued by patients
15
4.1 Mobile access to digital care records The work reviewed under this capability relates to mobile access to an electronic health record
(EHR) in patients’ own homes and care homes (the latter are recognised as the patient’s own
homes and are not networked NHS sites). Mobile access refers to technologies such as digital
pens, smartphones, and laptops used by staff to view and edit records held centrally. Overall the
evidence in this category was of mixed quality ranging from individual Trusts reporting successful
projects on websites, through to case studies, several systematic reviews and the National Mobile
Health worker project (2013). The presentation within this capability reflects this, and is quite
descriptive as comparison between evidence sources cannot be made because of these different
structures, information supplied and assumptions made. Where no systematic reviews are available
there is an increased risk that equivocal or negative findings were not published which can create
bias. The evidence is grouped under three main themes; effectiveness and accuracy, financial
impact and the impact on staff and patients.
Electronic Health Records have been described as a digital repositories of retrospective, concurrent
and prospective (planned appointments etc.) patient data, accessible by multiple authorised users
and stored and exchanged securely to support continuing, efficient and quality integrated healthcare
and consist of a mixture of coded data and unstructured narrative text (Häyrinen et al., 2008). The
Department of Health (1998) distinguishes between EHRs and electronic patient records (EPRs):
EHRs being longitudinal records that follow patients’ care for a lifetime and EPRs describing records
relating to periodic care given by one provider, although this difference is not always reflected in the
literature (Black et al., 2011). The technologies that were noted in the literature were personal
digital assistants, digital pen systems, tough books, iPads, smartphones and laptops. Their purpose
includes the ability to access and update notes in real time from any location with an internet
connection and the option to obtain a second opinion from another health professional virtually
(Dewsbury 2014). Although many benefits were reported there was a lack of detail about the
financial investment required.
Effectiveness and accuracy
The literature here is largely international. Pare (2011) carried out a study that investigated the
effects of the introduction of mobile computing on the quality of home care nursing practice in
16
Québec. The completeness of the nursing notes in nine centres was compared and included 77
paper records (pre-implementation) and 73 electronic records (post-implementation). Overall, the
introduction of the software was associated with an improvement in the completeness of the nursing
notes. Patients felt that the use of mobile computing during home visits allowed nurses to manage
their health condition better and, hence, provide superior care services. The nurses reported a very
high level of satisfaction with the quality of clinical information collected.
Self-reported, improved levels of knowledge leading to better care from the use of EHRs is
described in the following studies. Braun et al., (2013) conducted a systematic review of Community
Health Workers, mainly in developing countries using a variety of Mobile Technology not just mobile
access to records. They found that in the 25 articles included, learning was the most significant
outcome for staff (n=9, 32%). Among the reviewed studies Lee et al., (2011) examined mobile
phone use for maternity, sexual and child health services in Indonesia. The study showed that
mobile phone use was positively associated with higher self-efficacy and this higher self-efficacy
was positively associated with health knowledge of maternal health practices in the areas of family
planning practices, prenatal, and child delivery processes. Another study of traditional birth
attendants explained that use of mobile phones enhanced their ability to independently meet the
clients’ health care needs through more timely access to relevant information (Chib et al., 2008).
Braun et al., (2013) also concluded that the outcome most frequently measured among studies 71%
(n=20) was improvement in the quality of care and this was most commonly measured in terms of
compliance to accepted standards. For example, a simulated experimental study used mobile
multimedia devices to support point-of-care clinical decisions by Community Health Workers in
Colombia and found that there were significantly fewer errors and better compliance with care
protocols over a range of clinical care situations, (Florez-Arango et al., 2011).
Financial impact A full detailed cost benefit analysis of the use of these technologies is not available in the current
literature. The most comprehensive picture is presented in the National Mobile Health worker
project (DOH 2013) but some detail is also reported on NHS and affiliated websites. Following
implementation of a tough book, the National Mobile health worker project found for the first phase
the majority of sites contacts increased and more time was spent with patients. Journeys and total
journey time rose, although to a lesser degree than activity, indicating improved efficiency.
Clinicians across the 11 pilot sites estimated that the devices allowed them to save 507 referrals,
(£42 per referral), equating to a saving of nearly 9 % across the pilot period and an estimated 49
17
admissions (£1735 per admission) were avoided saving approximately 21% across the same
period. Results varied significantly across the sites and services which was a reflection of the
differing local processes and approaches and there is insufficient detail to determine whether these
are gross or net savings. In Phase 2 there were 6 sites, however the number of completed data
returns for the final benefit period ranged between 14 -88%. This combined again with the
significant variations that were present across the sites led the authors to conclude that there could
not be any meaningful analysis of combined data across the project (DoH 2013). Some sites
demonstrated significant increases in productivity including large increases in contact activity for the
Specialist Nurses in Hartlepool (142%) and the Palliative Services in Birmingham (83-93%). There
were also significant increases in time spent with patients following deployment of the mobile
devices with Tower Hamlets demonstrating a 104% increase. They also found that productivity
could be maintained even in the presence of significant organisational change, staff absence or bad
weather, with the devices allowing business continuity throughout.
In terms of efficiency journeys were reduced, in Hartlepool the Specialist Nurses achieved an 11%
reduction and time staff spent travelling in Tower Hamlets showed a 33% drop. Data duplication
was also reduced, freeing up clinical time. This was especially significant where EPRs were
established and use optimised. In addition, “no access” visits decreased significantly as
demonstrated in Avon (38-47%), Northampton (37%) and Tower Hamlets (50%). Each “no access”
visit was calculated to cost £42, (University of Kent 2010). Although data was supplied about
projected savings there were no detailed financial costs included in the projects. There was no
outline of how much the technology cost to implement and maintain or how much of the projected
savings were realised.
Successes in terms of cost reduction have also been communicated through award processes and
published case studies. For example, midwives in Portsmouth have reported the success of digital
pens. Here the midwife carries specifically printed forms, on which they write as they would with an
ordinary pen. The digital pen takes snapshots of the information, converts this into an encrypted
format, which is sent into the midwife’s BlackBerry via Bluetooth. The files are then sent back to the
hospital as soon as a signal is detected reducing the amount of travel and data input time for
midwives. It has also streamlined processes within the hospital, since the information is
automatically fed into different systems. They estimated savings to be the equivalent of six full-time
midwives, or £220,000 per year based on a workforce of 130, http://www.ehi.co.uk/news/acute-
care/7607/warwick-to-use-ipads-for-patient-records.
18
Derbyshire Healthcare NHS Foundation Trust estimates that 85% of its data is now entered on the
same day due to the use of digital pens by its 800 staff allowing the nurses to spend more of their
day on direct patient care producing an estimated saving of about £800,000 a year (Duffin 2013).
Rotherham Community Health Services has introduced digital pens and reported a 35% increase in
productivity of its physiotherapists http://www.ehi.co.uk/news/ehi/5867/rotherham-physios-flex-
digital-pens. Kidd (2011) reported the benefits to nurse prescribers of being able to access records
remotely, in a study of 11 rural and urban sites. Mobile access led to increased efficiency, better
patient contacts and improved teamwork. However, for all these reports there is a lack of detail
about set up and maintenance costs.
When the investment cost of equipment and maintenance is included the financial impact is less
favourable. In the USA, Rantz et al., (2010) explored the patient benefit of the electronic medical
record (EMR). They investigated the unique and combined contributions of EMR at the bedside and
on-site clinical consultation by gerontology expert nurses and found both costs and quality of care
increased in nursing homes. 18 nursing homes in 3 USA states participated in a 4-group 24-month
comparison where the EMR and nurse consultation were tested. Costs, staffing, staff retention and
resident outcomes were measured. Total costs increased in both intervention groups that
implemented the technology. The authors suggest that this was due to the cost of the technology,
maintenance, support and on-gong staff training. Improvement trends were detected in resident
outcomes of Activities of Daily Living, range of motion, and high-risk pressure sores for both
intervention groups but not in comparison groups. Therefore in this study the implementation was a
quality initiative rather than a financial one.
Impact on Patients and Staff Detail about patient perspectives is limited although there is work around the quality of their care.
For staff the literature mainly focuses on what is required for implementation. Lau et al., (2010)
carried out a meta-level synthesis of 50 papers (published 1994-2008) and found that the important
factors influencing HIT success for staff were having in-house systems, users who were developers,
integrated decision support and benchmark practices, and addressing such contextual issues as
provider knowledge and perception, incentives, and legislation/policy. They found some evidence
for improved quality of care, but this varied across studies. The areas where HIT did not lead to
significant improvement included resource utilisation, healthcare cost, and health outcomes.
Significantly they concluded that in many of the papers it was evident that the studies were not
designed or had insufficient power/duration, to properly assess health outcomes. They recommend
19
caution when interpreting these findings, because there were wide variations in organisational
contexts and how the HIT were designed/implemented and perceived echoing the conclusions of
the Mobile Health Worker project.
Black et al’s (2011) systematic review into the impact of eHealth on quality and safety also found a
difficulty in establishing cause and effect. They thematically categorised eHealth technologies into
three main areas: (1) storing, managing, and transmission of data; (2) clinical decision support; and
(3) facilitating care from a distance. They found that there was relatively little empirical evidence to
substantiate many of the claims made in relation to these technologies and that best practice
guidelines in effective development and deployment strategies are lacking. What is interesting in
regard to this capability is that although a number of reviews claim to assess the impact of EHRs,
many actually investigated auxiliary systems such as Clinical Decision Support Systems (CDSS),
Computerised Physician Order Entry (CPOE), and ePrescribing. As a result, most of the impacts
assessed were more relevant to these other capabilities.
Zhang et al (2010) reported a study of 91 Canadian nurses finding that the nurses’ perception of
usefulness is the main factor in the adoption of mobile technology. Doran et al., (2012) in the United
State highlighted usability, organisational culture, evidence-based practice and adaptation of the
electronic clinical information system. MacLure et al., (2014) in a systematic review of medical and
non-medical practitioners’ views of the impact of eHealth on shared care, summarised the major
themes from 12 papers as organisational, social, technical and external. For the organisational
theme, the main influences were resource and time implications, culture of the workplace and
change management requirements. Social concerns focused on the impact on patient consultations,
extra workload, need for training for varying levels of IT literacy amongst staff, usability, patient
privacy and the practitioners’ role. The technical themes included the incompatibility of technological
systems and the need for shared definition and terminology, External factors highlighted the
interplay between national policy and strategy and regional autonomy. Lastly, Koivunen et al.,
(2015) in a study in Finland of 117 nursing professionals found that they perceived electronic
communication to speed up the management of patient data, increase staff co-operation and
competence and make more effective use of working time.
20
Capability Details of review and search terms
Effectiveness and accuracy Financial impact Impact for patients and staff Learning for implementation
4.2 Digital images for nursing care
1858 records found; 104 abstracts assessed; 37 papers included. Search terms: Digital images, digital photographs, nurs*, digital camera, smartphone Tele*
Principal use of digital images in the recording of wound healing.
Little robust evidence for the effectiveness of digital imaging alone.
Overall more accurate in combination with additional clinical information or software.
Some evidence of effectiveness when combined with teleconsultations to reduce demand for appointments.
No inclusion of investment costs in the literature for the equipment required
No inclusion of the cost/benefit if wounds are monitored effectively and advice sought when needed.
No inclusion of patient
feedback.
Limited staff perspectives in
literature.
One study mentioned staff
anxiety.
Another study reported increased staff confidence when expert viewed images remotely.
Legal implications for use of images and quality.
Images must be considered as part of the patient record.
Considerable scope for increasing skills in nurses working with specialists.
Systems need to be integrated and complete rather than require staff printing off pictures and putting them into a paper patient record.
Good image quality is very important.
21
4.2 Digital images for nursing care A digital image is one that is recorded via a digital camera or smartphone and transmitted
electronically. The most evidence emerged for still photography, in particular its use in the
diagnosis and monitoring of the healing of chronic and acute wounds. Much of the literature found
in this review was non-UK in origin. Most studies were too small to give reliable information about
sensitivity and accuracy of assessments or measurements obtained using digital imaging.
The digital image may often be printed and kept in paper notes but is sometimes stored
electronically in the EHR; this review considers how these images are incorporated into Telehealth
applications to ensure correct and timely treatment. Several other applications emerged from the
evidence review and have also been included such as the digital imaging of patients’ feet who have
diabetes. Although at present this is medically dominated it is foreseeable that this could be a useful
technology for nurses to employ routinely in the future. Other applications were found such as digital
images to record assault-related injuries and to monitor food intake. There is little evidence
regarding these applications currently. Generally, papers found in this area involve small numbers
of patients and focus on the effectiveness of the digital methods used. Again, cost-benefit analyses
are largely absent as is information on acceptability for patients or nurses.
Effectiveness and accuracy in wound assessment Digital photography has virtually replaced conventional film photography for clinical imaging. It is
commonly used to demonstrate if new procedures, dressings and ways of working, impact on
wound healing (Kaliyadan et al., 2008, MacBride et al., 2008, Varkey et al., 2008, Dufrene 2009,
Rennert et al., 2009, Bradshaw et al., 2011, Engel et al., 2011) and is recommended in NICE
guidance on pressure ulcer management (NICE 2014) .
Thompson et al., (2013) examined the validity and reliability of the revised Photographic Wound
Assessment Tool (revPWAT) on digital images taken of various types of chronic, healing wounds.
This Canadian multicenter trial involved 206 different photographs taken of 68 individuals with 95
chronic wounds of various aetiologies (diabetic foot, venous/arterial ulcers, pressure ulcers). An
initial wound assessment using the revPWAT was performed at the bedside and 3 digital
photographs taken. The revPWAT scores derived from photographs were assessed by the same
staff on different occasions and by different staff. They showed moderate to excellent intra-rater
correlation coefficients as well as test-retest and interrater reliability. They concluded that there was
excellent agreement between bedside assessments and assessments using photographs.
22
Florczak et al., (2012) carried out a study where nine USA wound care nurses used a camera-
enabled mobile device at the bedside as part of a wound assessment process. The greatest gains
for single evaluation items were reviewing wound progress and recognizing changes in wound
status. In this study, the wound nurses used the camera-enabled mobile device to take digital
photographs at the bedside, to enter wound assessments and risk assessments and to record
treatment recommendations. While this approach supported client-centered care and effective
communication, it also represented a competing pressure on nurses’ time. Some nurses struggled
with the introduction of a new assessment protocol incorporating digital photographs. Because their
system was not all electronic they also had to print it off and sign images to share with others, thus
increasing workload.
Baumgartner et al., (2009) in a study involving 48 patients and compared the validity of digital
photographs for the remote assessment of pressure ulcers as compared to existing practice of
bedside examination and diagnosis. The photographs were reviewed blindly by a dermatologist. The
sensitivity of the blinded assessment was 97% in terms of pressure ulcers being diagnosed at the
bedside and via photographs and areas judged to be unaffected at the bedside were also identified
correctly from the photographs. The results were only slightly less positive (92% sensitivity) when
limited to darker skin.
Terris et al., (2010) found a much less clear validation of digital imaging when they investigated the
assessment of stage 3 and 4 pressure ulcers in individuals with spinal cord injury in Cleveland
comparing beside with digital photograph assessments. Two wound-care nurses assessed 31
wounds among 15 participants during a 6 month period. One nurse assessed all wounds in person,
while the other used digital photographs. They achieved only moderate agreement about the
wounds which the authors felt may have been attributed to variation in assessor’s subjective
perception of qualitative wound characteristics.
Some authors have expressed caution in relying on digital images to diagnose and monitor wounds.
Buckley et al., (2009) conducted a descriptive comparative community study of 43 adult patients
with a total of 89 wounds. They found that Wounds, Ostomy and Continence nurses who provide
remote nurse-to-nurse consultations using digital images were at risk of recommending under or
over-treatment of patients' wounds. In dermatology, Janardhanan et al., (2008) found that some skin
conditions required a face-to-face consultation and they recommended a mixture of face-to-face
consultations and consultations via tele dermatology. For burns management, Hop et al., (2014)
found that burn size, but not depth, could be assessed reliably and validly by experts using
photographs of the burn wound. Finally Jesada et al., (2013) explored whether a digital photograph
23
obtained by a nurse in the acute care setting could be used to determine staging and wound
characteristics of a pressure ulcer when viewed remotely compared to bedside assessment. 100
digital photographs of pressure ulcers were obtained from 69 patients on general and critical care
medical-surgical nursing units. The study compared 13 wound characteristics and a total score on
the Bates-Jensen Wound Assessment Tool (BWAT) as well as the staging of pressure ulcers. Their
findings indicated that a digital photograph alone cannot reliably convey the characteristics of a
pressure ulcer. Jesada et al., (2013) therefore advocated a digital photo and clinical information,
systems designed to provide digital photography wound measurements and the development of
standardized education for examining digital photographs by wound experts, to increase the
accuracy of the assessment and documentation.
Other studies have used planimetry software (designed to measurement of surface area and
surface area-to-perimeter) to provide more accurate wound measurements from digital photographs
(Rogers et al., 2010, Wendelken et al., 2011). (Mayrovitz, and Soontupe (2009) in a small study of 6
wounds and 20 senior nursing students, used computerized planimetry of digital images. They
found area error rates of up to 7% with well-defined wound edges and 10% for wounds with less
defined borders. Miller et al., (2012) investigated the inter-rater and intra-rater reliability of a wound
imaging and measurement system called SilhouetteMobile in community nursing, in Victoria,
Australia. Seven community nurses including Wound Management Clinical Nurse Consultants and
Wound Resource Nurses assessed average wound surface area of 14 wound images as captured
using a wound imaging and measurement system. This again was a very small study but they
found that here was high inter-rater and intra-rater reliability, which was unaffected by image quality.
However, reliability was poor when tracing small wounds.
Combined with telehealth The following two small studies indicate how digital imaging could possibly be incorporated into
patient care to reduce hospital visits and inappropriate referrals. Chanussot-Deprez (2008) carried
out a small study of three patients with hard to heal ulcers in Mexico. A digital camera was used to
take images which were emailed to a wound care expert in Mexico City. The photographs allowed
the expert to diagnose and evaluate the wounds, two wounds healed and the decreased in size
although there was no information about time, they also they said it was cost-effective but no details
of cost were included.
Vowden and Vowden (2013) evaluated the effectiveness of a telehealth system, using digital pen-
and-paper technology and a modified smartphone to take digital pictures, to remotely monitor and
24
support the effectiveness of wound management in 39 nursing home residents. It was a cluster
randomised controlled pilot study, conducted in 16 selected nursing homes in Bradford for residents
with a wound of any aetiology or severity. Residents in the control homes, continued to receive
standard care directed by nursing home staff, while those in the evaluation homes received wound
care supported by input from remote experts. Analysis of individual patient management pathways
suggested that the combination of digital picture and clinical information provided, improved patient
outcomes. Such applications may offer cost savings by improving dressing product selection,
decreasing inappropriate onward referral and speeding healing. These studies are small and lack
detail about cost-benefit, but give a suggestion of possible benefits.
Imaging of feet While there is much medical literature looking at thermal imaging of the skin on the foot to assess
circulatory status (Nagase et al.,2011, Roback 2010, Roback et al., 2009), only one study was found
that looked at how the use of digital images would assist in patient care. Foltynski et al., (2010)
reported on a new way of managing patients with diabetes who needed foot care, with a nurse at
the patient’s home and a consultant in another location. Digital images of foot ulcers and blood
glucose and blood pressure measurements, were sent to the consultant for an opinion. The visiting
nurse reported increased confidence with the treatment of the foot ulcer and described the
consultations as a learning situation. Patients also expressed satisfaction with the process.
Forensic digital imaging Nurses may be involved in using digital imaging to document the effect of assault as it is seen as
valuable additional documentation of physical injuries after sexual assault. Ernst et al., (2012)
reported on the need for high image quality and naturalness in order that the images are reliable to
use. This issue of image quality is also picked up in Melville’s (2013) study which explored the
quality of photographs that are taken by professionals in child abuse cases. A total of 120 images
depicting 60 cutaneous lesions were selected for the study. Paired images of single lesions varied in
quality of focus, exposure, or framing. Seventy medical and nursing professionals evaluated the
images for quality (1-9 scale). Accuracy was defined as the agreement between the subjects and
the live examiners written documentation. Adequacy was the proportion of subjects that did not
indicate that the photograph was inadequate for interpretation. Mean accuracy among subjects was
64% and ranged from 35% to 84%. Accuracy was not predicted by subject profession, experience,
or self-rated computer skill. Image quality and adequacy were independently associated with
increased accuracy. They concluded that reliance on photographs alone is not sufficiently accurate
in the assessment of cutaneous trauma.
25
Dietary intake Navarro et al (2014) used digital imagery of food to estimate correlation with real-time visual
estimation of dietary intake reporting the intra- and inter-rater agreement for digitally captured plate
waste in hospitalised adults. 106 lunch plates of hospitalised adults were photographed
immediately prior to tray collection. Two raters used the modified Comstock method to assess food
waste from each of their digital images. Additionally, a randomly selected subset of photographs
was re-estimated 72 hours following the first estimation. The results showed that dietary intake
estimated from digital images has a high degree of both inter- and intra-rater reliability in
hospitalised patient populations. The findings were similar for regular diets but somewhat weaker for
soft diets.
Legal implications A theme to emerge from this literature review relevant to all professionals, was the importance of
recognizing the legal implications of taking and storing digital images. A detailed discussion on the
security of digital technology in the health sector is beyond the scope of this review.
There is clear guidance for medical practitioners, (GMC 2011) on the recording of photographs,
record keeping for nurses (Nursing and Midwifery Council 2015) and further advice by the Royal
College of Nursing (2012). Informed consent must be obtained for the photograph, with patients
understanding why the photo is being taken and how it will be used and stored. NHS organisations
should have their own policy on the taking and storing of photographs drawing on the Information
Security Management: NHS Code of Practice (DoH 2007). This states that digital or printed,
photographs must be treated as the same as paper or digital health records and as such staff are
required to abide by all relevant laws and national guidance including the Data Protection Act 1998
with regards to the capture, storage and use of patient images (NHS England 2014, Lakdawala et
al., 2013).
Harting et al., (2015) commented that the main choice was whether to have a digital camera or let
staff use smartphones. Many organisations have a Trust digital camera by which photographs are
taken. However Harting et al., (2015) says that they are usually not password protected and require
the nurse to download the pictures to a computer. Smartphones are highly portable, can rapidly
distribute a captured image and their password protection and/ or encryption abilities can better
secure data, making unwanted access less likely. Whatever is used, and despite password
protection, encryption and other technical safeguards, any storage, transport or dissemination
method can be compromised, potentially exposing identifiable data to misuse. All mobile technology
26
is subject to loss, damage or theft especially in a community settings entailing other risks such as
treatment delays and loss of record continuity.
King (2014) commented on the governance arrangements and training needs that were identified as
necessary prior to rolling out a community nursing project. The study evaluated a remote system of
wound management, incorporating the use of digital and mobile technology the use of a wound
template and the recording of remote images. Consent to photography, information governance
such as adequate encryption, ensuring that the correct images were uploaded to the correct record
and the practical issues of uploading and attaching images to records and reports were all
highlighted as important governance issues.
27
Capability Details of review
and search terms Effectiveness and
accuracy Financial impact Impact for patients and
staff Learning for
implementation 4.3
Remote face
to face
interaction
1398 records found;
230 abstracts
assessed; 28 papers
included.
Search terms:
video consultation,
video conference
teleconsultation,
teleconference,
nurs*, tele*, remote
interaction
Effective
technology
particularly in
the
management of
long term
conditions and
also the acute
treatment of
stroke where
specialist advice
is otherwise
unavailable.
Some evidence of
investment costings of
equipment but cost/benefit
analysis rare.
Overall results were mixed depending on service re-design e.g. not cost-effective if it involves both a doctors with the patient and a specialist.
Some studies but not all
reported reductions in
readmission rates and
length of stay.
One systematic review
reported telehealth to be
less costly than the non-
tele health alternatives.
Clear savings were
demonstrated where
medical cover was not
feasible e.g. in prisons
where transfer of patients
has additional costs or
reductions in admissions
from care homes or
transfers from rural
hospitals.
Mixed evidence of
patient and staff
satisfaction.
Positive benefits for
patients in ease of
access and
reduced expense.
Possible negative
impact of reduced
personal contact.
Technology always
alters relationships
and dynamics
between patients
and clinicians
thought rarely
reported or
analysed in detail.
Staff adopt different
roles when
implementing
telehealth.
Needs careful initial facilitation and integration with other human and technical systems.
Whole system HIT implementations probably provide the greatest benefits but there is little cost/benefit data.
Particularly tested and beneficial in rural areas.
Require HIT infrastructure and resources e.g. bandwidth and network coverage.
Considerable scope for nurses to develop further skills under remote supervision of specialists.
28
4.3 Remote face to face interaction Teleconsultation technology allows patients and clinicians to communicate with other clinicians
through a video conference link (Rice 2011). This section focusses on the consultation or
conference between health care staff and patients as a clinical tool, rather than between staff as an
educational or communication tool. Cruikshank (2010) says this improvement in access to expertise
could improve equality, access and clinical outcomes for patients but requires integration with other
systems. “Its effective usage requires access to video; Picture Archiving and Communications
(PACS) image sharing capabilities; remote diagnostics (e.g. ECG); and the patient record”
(Cruickshank 2010 p14). A range of evidence was found in this area, including several RCTs but
most of the research explores individuals’ implementation of the technology. The literature in this
capability is international in origin and there is a significant focus on applications in rural
communities where it is difficult either for health care providers to get specialist services to patients
or patients to services. The researchers in this areas acknowledge the huge challenges faced by
health services in delivering access to tertiary or secondary care to sparse populations. Video
consultation is seen as a central way to address this cause of inequality within the literature.
Themes of effectiveness, financial impact and acceptability to patients and staff emerged from this
literature. During such remote interactions, the nurse usually plays the role of being the person with
the patient and connecting to the specialist who is in another location.
Effectiveness and accuracy The literature on effectiveness of video consultation versus face to face consultation predominantly
focusses on medical diagnosis and treatment. Fatehi in Queensland, Australia, has reviewed 505
papers on the technical aspects of clinical videoconferencing but found that a high proportion of
telemedicine papers lack sufficient technical details which limits their repeatability and
generalisability (Fatehi et al., 2015a). Fatehi then went on to carry out a study which examined
videoconferencing and the decision making of endocrinologists concluding that it was as effective as
in-person consultation in decision making (Fatehi et al., 2015b). There is also literature that
describes effective management of patients, particularly those with chronic conditions (Chen et al.,
2013, Dhiliwal and Salins 2015). This is an area where nurses increasingly play a significant role
and where the results for telehealth are generally favourable.
Clemensen’s (2008) pilot study investigated whether video consultations in the home were a viable
alternative to hospital visits for patients with diabetic foot ulcers. They also measured the patients,
relatives, visiting nurses, and experts’ levels of satisfaction and increased confidence with this new
29
mode of assessment. Through field observations, semi-structured interviews, focus groups, and
qualitative analysis of tele-medical consultations in a pilot test they showed that it was possible for
experts at the hospital to conduct clinical examinations and decision making at a distance, in close
cooperation with the visiting nurse and the patient.
One chronic condition that has received significant interest is Chronic Obstructive Pulmonary
Disease (COPD). Sorknaes et al., (2011) and Saleh et al., (2014) evaluated the use of
teleconsultations with respiratory patients with COPD. Sorknaes et al., (2011) investigated the
effect of telemedicine video consultations (TVCs) on early readmissions of 100 non- randomly
allocated COPD patients in their homes after discharge from hospital. The patient received daily
TVC at home with a nurse based at the hospital for approximately one week and an additional
follow-up call. The patient could also call the nurse during the study follow-up period of 28 days. The
nurse and the patient were able to see and talk to each other and the nurse was able to measure
saturation and perform spirometry remotely. The results were transferred to the hospital by a secure
Internet connection. 12% of the TVC group compared to 22% of the control group were readmitted
because of exacerbation of COPD indicating that TVC is a significant protective factor for early
readmission. There are significant limitations to this study. The subjects were not randomly
assigned to treatment group, the duration of the study was short and the intervention was not
blinded. Analysis of the economic implications was limited to comparison to other, much older
studies.
In a later, randomised study of 266 patients with severe COPD discharged after acute exacerbation
from two hospitals, Sorknaes et al., (2013) found there were no significant difference in the total
mean number of hospital readmissions after 26 weeks between the intervention group and control
groups. There was also no significant differences between the two groups in mortality, time to
readmission, mean
number of total hospital readmissions, mean number of readmissions with COPD, mean number of
total hospital readmission days or mean number of readmission days with COPD measured at 4, 8,
12 and 26 weeks. Saleh et al., (2014) evaluated the effects of telemedicine video-consultation
(TVC) using similar programme but with follow-up for 12 months using a before and after design.
They compared the frequencies and durations of hospital re-admissions due to COPD
exacerbations within 6 and 12 month follow up after commencement of TVC to events at 6 and 12
months prior to TVC. The study showed the number of patients re-admitted and the number of re-
admissions due to COPD exacerbations were not reduced at 6 or 12 months post-TVC. The mean
30
length of re-admission at 12 months post-TVC was markedly reduced as compared to pre-TVC.
There was no cost-benefit analysis reported.
Remote teleconsultation has also been studied in the treatment of acute stroke. Audebert et al.,
(2009) carried out a follow-up study to assess the long-term impact of a stroke network with
telemedicine support. Over a 30 month period, 1938 stroke patients (and 1122 control patients) from
five community hospitals were followed-up. There were no significant reductions in composite
outcomes of reduced “death or institutional care” at 30 months but a significant reduction of “death
and dependency” at 30 months. Pedragosa et al., (2009) found an increase in the number of
patients evaluated by a specialised neurologist and Dharmasaroja et al., (2010) an increase in the
number and speed of thrombolytic treatments that occurred. Over an 11 month period,
Demaerschalk et al., (2010) studied 54 acute stroke patients and found no significant difference in
mortality, 90 day functional outcome, use of thrombolysis, rate of intracerebral haemorrhage after
treatment with thrombolysis or evaluation times, between telemedicine vs telephone consultations.
No consultations were aborted but technical problems occurred in 74% of telemedicine
consultations compared with none in telephone consultations. None of the technical issues
prevented correct treatment decisions but some did effect the time-to-treatment decision
(Demaerschalk et al., 2010).
Roots et al., (2011) included these studies in their systematic review investigating the use of
telemedicine in managing acute stroke and post-stroke rehabilitation. There were 20 observational
and two randomized, controlled acute studies involving 15872 patients and two observational and
four randomized controlled rehabilitation studies involving 112 patients that met the inclusion
criteria. From this review they concluded that telemedicine for stroke, in both acute and
rehabilitation settings, was demonstrated to be safe, effective, feasible and acceptable. It was also
shown to reduce geographical differences, increase diagnostic accuracy and increase uptake of
thrombolytic treatment. Further work is also in progress with Salisbury et al., (2014) exploring what
other conditions and patient groups internet videos and other remote consultation methods could be
effective.
Financial impact While there are systematic reviews of the evidence around telehealth, videoconferencing is usually
combined with other methods, particularly consultations by telephone and therefore it is difficult to
assess its impact independently. Wade et al., (2010) carried out a systematic review of economic
31
analyses of telehealth services using real time video communication. They were unable to perform
a meta-analysis of the effects of telehealth because of the heterogeneity of the outcome measures.
Instead they reviewed studies which report economic details with patient outcome data and a non-
telehealth comparator. 36 articles met the inclusion criteria. 61% of the studies found telehealth to
be less costly than the non-telehealth alternative, 31% found greater costs and 9% gave the same
or mixed results. The studies were classified according to where costs arose, with 23 reporting the
cost perspective of the health services, 12 were at societal level, and one included direct patient
perspectives. In three studies of telehealth in rural areas, the health organisation paid more for
telehealth, but due to savings in patient travel, the societal perspective demonstrated cost savings.
In regard to health outcomes, 33% of studies found improved health outcomes, 58% found
outcomes were not significantly different and 6% found that telehealth was less effective.
In the review Wade et al., (2010) highlighted that the organisational model of care was the most
important factor in determining the value of the service compared to the clinical discipline, the type
of technology, or the date of the study. They concluded that the delivery of health services by real
time video communication was cost-effective for home care and access to on-call hospital
specialists. Nonetheless they showed mixed results for rural service delivery, and found no
evidence that telehealth was cost-effective for local delivery of services between hospitals and
primary health care centres. This final funding was due to GPs being with the patient during the
consultation, (thereby doubling the medical costs) underlying the importance of the details of
implementation in delivering expected efficiency savings.
Detailed costings are available for TVC in a study by Barber et al., (2015) who highlighted that in
remote and rural areas, such as the Western Isles, specialist medical input i.e. stroke may not be
available or feasible. They reported a before and after study of a two year service intervention
where a weekly stroke multidisciplinary meeting in the Western Isles Stroke Unit was led by a
remote specialist stroke doctor via a videoconference link. They found that the median length of
stay fell by 10 days, primarily by empowering the team to discharge plan at an earlier stage. This
they calculated would lead to a potential recurring annual cost saving of €68,000 (one day in
hospital = €400). They reported that no other significant rehabilitation service changes were made
during the two-year project to account for these improvements. They acknowledge that the numbers
were small and that length of stay is a fairly crude outcome measure dependent on many other
factors. They costed the consultant coming to the hospital as €48,000, the cost of implementing the
service €7000, the videoconferencing unit €9967 and the license €177. This extension of equity and
32
quality in access to specialist care should be set against the need to demonstrate absolute
cost/benefit improvements but is difficult to quantify in cost-benefit comparisons. Finally in trauma
and general surgery contexts Latifi et al., (2009) carried out a retrospective analysis of 59
teleconsultations between five rural hospitals and a level One trauma center in the US. They
included set up costs (which were covered by a grant) and details on ongoing costs. For six
patients, the remote consultations were considered potentially lifesaving and 29% of patients
remained in the rural hospitals (eight trauma and nine general surgery patients). Treating patients in
the rural hospitals reduced transfers, saving an average of $19,698 per air transport or $2,055 per
ground transport, in addition to potential safety benefits. The telepresence of a trauma surgeon aids
in the initial evaluation, treatment, and care of patients, improving outcomes and reducing the costs
of trauma care.
Rice (2011) reported on the work at Airedale NHS Foundation Trust which had been developing a
range of services using high quality video links to provide consultant and other clinical professional
input to prisoners. Approximately 50% of all the patients remained in the remote location to receive
care. Previously all cases would have been transferred to the local secondary care for consultations,
indicating a likely cost saving to both the health care and prison services (not reported). Cruikshank
(2010) also commented on the Airedale work in over 18 UK Prisons and took a more financial focus
estimating that the average cost per escort episode (including mental health transfers) as £350 and
the average cost per bed watch episode was £3,731.
Pope et al., (2013) described the results of providing 24 hour access to immediate clinical support
using teleconsultation (the “telehub”) in both domestic and residential care settings in the Yorkshire
and Humber Region. This dedicated 24/7 service is staffed by experienced nurses who can receive
and make calls to patients over domestic broadband links, supported by a consultant general
physician. Consultations are viewed either on a domestic TV via set top box technology or on a
dedicated mobile video system. The consultations are documented using a shared EHR (TPP
SystmOne). COPD patients using the scheme had a 29.5% reduction in hospital admissions and a
36% reduction in length of stay for those admissions that did occur. Overall in the 14 nursing and
care homes supported using mobile teleconsultation equipment, there was a 50% reduction in
admissions compared with similar homes that had no access to teleconsultation and a 74% relative
reduction in Emergency Department attendances. They did not report details of the cost of
providing the hub nor additional training. In a later paper, Pope et al., (2014) explored whether this
change in practice could be sustained. Initial findings from 17 care homes supporting a population of
33
just over 1000 people, showed that there appeared to be a 53% reduction in ED attendance, 35%
reduction in admissions and a 59% reduction in total bed days for residents in the 12 months
following introduction of the service. Again the actual cost of these systems and the staff resource
requirements e.g. training, are absent as with most of the literature identified. The location of costs
and the relative impact of location of care on average costs must also be taken into consideration
when overall financial impacts are considered.
Patient and staff acceptability How patients and staff feel about using the technology and how it changes their interactions,
receives some attention in the literature and serves to indicate the complexity of introducing HIT.
From a patient perspective the evidence is mainly positive with schemes reporting high satisfaction
(Clemensen et al., 2008, Saleh et al., 2014, Sorkneas et al., 2011, Pope et al., 2013, 2014, Barber
et al., 2015). Johansson and Lindberg (2014) via a questionnaire distributed in rural areas of
northern Sweden, described the views of 26 residents of rural areas on accessibility to specialist
care and the use of video consultation as a tool to increase accessibility. They conclude that
although patients recognise savings in time, environmental damage and cost of travel as important,
there was uncertainty regarding video consultation as preferred solution. From a professional
perspective, Vuononvirta et al., (2009) carried out a study of the attitudes of multi-professional
teams to telehealth adoption in northern Finland health centres and revealed some complexity.
Observations and interviews were carried out with 30 professionals (physicians, nurses, psychiatric
nurses, physiotherapists). Overall attitudes were more positive than negative with diversity of
attitudes occurring in relation to time, situation, profession, health centre and telehealth application.
Ten different types of telehealth adopters were categorised: “enthusiastic user”, “positive user”,
“critical user”, “hesitant user”, “positive participant”, “hesitant participant”, “critical participant”,
“neutral participant”, “negative participant” and “positive non-participant”. This may be helpful for HIT
project staff and managers to consider when working with staff implementing such technologies.
There is some evidence that using this technology affects the nature of the clinician/patient
relationship e.g. Clemensen et al., (2008), Esterle and Mathieu-Fritz, (2013), Van Gurp et al.,
(2013). Wakefield et al., (2008) compared differences in nurse and patient communication profiles
between two telehealth modes: telephone and videophone. 28 subjects were enrolled in a
randomized controlled clinical trial evaluating a 90-day home-based intervention for heart failure.
They found that nurses on the telephone were more likely to use open-ended questions, responses
such as mmm and yeah, friendly jokes, lifestyle information, approval comments and checks for
34
understanding. Compliments were given and a sense of partnership was more evident but closed
questions were more common on the videophone. Nurses perceptions of the interactions were not
different between the telephone and videophone, nor did their perceptions change significantly over
the course of the intervention. The study did not collect patients’ perspectives.
Miller (2011) reviewed 10 interaction analysis studies of videoconference systems, describing and
categorising the communication and concluded that whilst some studies found telemedicine to be
more patient-centered than conventional medicine, others found it less so. There is currently limited
understanding of positive and negative effects of telemedicine on provider-patient relations.
35
Capability Details of
review and
search terms
Effectiveness and accuracy Financial impact Impact for patients and
staff
Learning for
implementation
4.4
Digital
capture of
clinical data
at point-of –
care and
digitally-
enabled
observations
management
Records found
1920; abstracts
assessed 145;
33 papers
included.
Search terms:
“Continuous
monitoring”,
“automatic
data capture”,
“automated
clinical
prediction”,
“clinical
decision
support
systems”,
automat*,
alert*, alarm*,
hand-held,
nurs*, observ*,
chart*, monit*
Automated data capture was more
accurate than manual recording
Continuous measurement more
comprehensive and less resource
intensive
Increased availability of data for
clinical audit and research but little
evidence of impact of this.
Greater sensitivity to some clinical
events e.g. apnoea
Evidence supporting integration of
nursing data into prognostic alerts
was mixed - most deployments are at
developmental stage
Some evidence that automation of
EWS calculation coupled with
decision support for nurses e.g.
when to escalate, is effective in
improving process outcomes and
timeliness
No conclusive evidence that
automated warnings reduce LOS or
mortality but may reduce cardiac
arrest calls via early detection of
deterioration.
Complex integration of clinical data
for prognosis and early warning
using predictive analytics in very
early stages of development - may
Likely
reductions in
nursing time but
not widely
quantified.
Reduction in
nursing
resource for
observations
may be
substituted for
other patient
care –
financially
neutral but
quality positive
Automated care
summary
generation
studies showed
reduction in
record
preparation and
handover times.
Little detail of
equipment and
implementation
costs e.g.
training.
Some studies
reported positive
feedback from staff
for automated data
capture and
automated care
summary generation
- primarily ease and
time saving.
No direct
involvement of
patients reported -
given most studies in
ITU and theatres this
is not unexpected.
Automated
measurement may
be able to reduce
“white coat”
hypertension other
systematic errors
and reduce
overtreatment.
Possible negative
patient experience if
automation leads to
overuse as
continuous
monitoring may be
uncomfortable
Significant
opportunities exist
for transformation
of “early warning”
systems
More
comprehensive
clinical monitoring
and production of
data for patient
trends, clinical
audit and
research.
Considerable
scope to free- up
nurse resource
for other care
activities
Analysis of time-
series data is
needed to reveal
impacts on
mortality, LOS
and ITU
escalation
“Big data”
analytics rarely
used or planned
Other factors are
important
36
lead to improvements in clinical
decision-making. Potential benefits from improved
accuracy, safety and completeness
of automated care summaries by
integrating with EHR were described.
Mixed results about whether it
reduces handover preparation time
or time to treatment delivery
Little information
on staff
interactions and
communication
or restrictive in lower
acuity settings. Risk of “alarm fatigue
confounders e.g.
trends towards lower
acuity of hospitalised
patients
Cost-benefit should
reflect reduced
workload and
increased quality
care and improved
patient experience.
Process data external
to the systems e.g.
clinical response times
should be integrated
for robust evaluation of
impact.
•Specificity of alarms
and alerts is of
concern to avoid
desensitisation
37
4.4 Digital capture of clinical data at point-of –care and digitally-enabled
observations management The physiological monitoring of patients is essential for the early detection and prevention of
deterioration in acute settings. NICE has published recommendations regarding the physiological
monitoring of patients in secondary care (NICE 2007 CG50) and mandates the use of “track and
trigger” or “recognise and rescue” models in all acute settings. Along the continuum from recognise
to rescue there are a number of points at which technology can enable staff to intervene more
rapidly and effectively. The following section presents two closely inter-related applications of digital
technology to acute healthcare: automated clinical data capture and storage, and management and
manipulation of such data to support clinical decision-making and handovers.
Clinical observations in acute care are predominantly recorded by nursing and increasingly, support
staff. The majority of interventions identified are limited to routine but vital and high-volume
observations such Non-invasive Blood Pressure (NIBP), respiratory rate, heart rate, oxygen
saturation and temperature. Literature regarding automatically captured parameters such as
ventilator settings or blood gas assays is also included if such data are then automatically stored as
part of a clinical record or integrated into a clinical decision support system. The automatic
integration of such observations with other parameters or the triggering of clinical responses
remotely, are dealt with in the section on management of digital clinical data as are the limited
number of sophisticated systems which integrate nursing data with other elements of the EHR such
as laboratory results to create alerts or hand-over documentation automatically. It is predicted that
such integration will become the norm in the coming decade and the limited impact of current
deployments reflect delays in radically changing working practices in healthcare needed to realise
the huge potential benefits (Wachter, 2015).
Automated Clinical Data Capture Timely and appropriate medical intervention for individual patients depends on the accurate
measurement and recording of physical and psychological parameters and the subsequent
availability of such data to inform clinical decision making. In recent years the collection of
physiological data at the bedside has increasingly relied on electronic devices to measure (but
rarely record) routine observations. Paper charts and clinical notes continue to be the norm in much
of the NHS regardless of acuity or level of care. Manual recording methods in nursing result in a
number of pitfalls such as missed or missing observations, transcription, legibility and rounding
errors and finally, errors in manually calculating useful clinical indexes such as Early Warning
38
Scores EWS (NICE CG50 2007, National Patient Safety Agency 2007). The literature in this area
largely focuses on the benefits of automated capture in terms of efficiency and accuracy of data
collection and patient outcomes, and only secondarily on cost-effectiveness or staff and patient
acceptability.
Effectiveness and Accuracy of Automated Data Capture There is some evidence that recording vital signs using HIT is faster and more accurate. A
simulation cross-over study of vital sign data entry carried out by Prytherch (2006) found that an
electronic system was 1.6 times as fast as manual recording with a reduction in errors. Bellomo
(2012) and Mohammed (2009) showed reductions in nursing time taken to record data when
electronic data capture was compared to manual charting (1.6 minutes and 13.9 seconds
respectively). Automated capture improved the completeness and accuracy of blood pressure
monitoring Hug (2011) in a study of NIBP monitoring in an intensive care environment. Automated
observations (filtered for extreme values indicating measurement error) improved predictions of
subsequent hypotensive episodes when compared to traditional episodic recording made by nursing
staff. Jones (2011) examined the use of the “Patientrack” system and found a 19% error rate on
paper was completely eliminated after the electronic system was introduced. In a small-scale cross-
over study of 6 nurses treating 5 patients in an intensive care setting, Wood and Finkelstein (2013)
found that a 30% error rate in traditional recording was eliminated after automation. Wood and
Finkelstein (2013) also reported evidence of efficiency gains as recording was instantaneous and
data were immediately available to multiple staff. Smith et al (2009) found the introduction of a bar-
code enabled, hand-held device used to upload vital signs automatically reduced errors of 10%
(paper) and 4.4% (manual entry into EHR) to less than 1% in an uncontrolled before and after study.
Taenzer et al (2014) carried out a comparison of automatic vs nurse recording of oxygen saturation
measurements. Nurses were found to overestimate SaO2 by 6% by comparison to the automated
record. This has important safety implications as it implies that desaturation events may be missed
with intermittent manual recording methods but could also reflect correct filtering out by nurses of
incorrect observations e.g. during repositioning of miss-placed probes. Signal et al. (2010)
compared automatic, continuous glucose monitoring on an intensive care unit to periodic manual
nurse blood sampling using point of care testing. This study predicted a small efficiency gain in
nursing time and improved glycaemic control but no full cost-benefit was undertaken. Both
Prytherch (2006) and Wood and Finkelstein (2013) reported positive staff evaluations of the
automated processes they studied.
39
Automation could have secondary benefits. A common and important source of bias in
measurements of blood pressure arises from patient anxiety during measurement taking creating
errors known as “white-coat” (UK) or “office” (US) hypertension. This can lead to over-detection and
subsequent over-treatment of hypertension, mainly in primary care. In a group of studies (Myers
2012 and 2014) and Nelson et al (2009), a fully automated system was used to measure NIBP over
10 minutes: the patient was instructed to lie down and relax and was left alone during the process.
The automated measurements were found to correlate more closely with continuous ambulatory
measurements and markers of end-organ hypertensive damage such a renal function, when
compared to traditional methods potentially reducing over treatment. Myers (2012) suggested that
the system may not yield significant benefits if staff fail to wait for the full assessment period or to
leave the patient alone for sufficient time emphasising the importance of human reactions to new
technology.
Automated systems can be set to obtain ad hoc, periodic or continuous measurements, thus
overcoming the need to rely on staff to carry out periodic assessments which can be missed. Once
monitoring equipment is placed e.g. NIBP cuffs or arterial lines, the system takes measurements
independently reducing workloads. Multiple or continuous measurements of rapidly fluctuating
physiological parameters are likely to correlate better with long-term physical status than single
random measurements. This principle, along with reductions in patient anxiety, underpins the
increased accuracy of continuous or automated system if implemented appropriately. Continuous
monitoring of vital signs for example, was shown in one study to decrease ITU length of stay (LOS)
(Bellomo 2012) and in a second, to reduce general ward and ITU LOS (Brown 2014). Evidence of
an effect on cardiac arrest rate was equivocal in both studies. There is a significant risk of bias in all
of the available evidence for effects of continuous monitoring on patient outcomes arising from the
uncontrolled and unblinded before and after study designs and demonstrating impacts on mortality
is likely to require very large studies.
There is limited evidence demonstrating benefits of employing automated regular or continuous
monitoring outside of hyper-acute settings (theatres and intensive care). Continuous or more
frequent intermittent monitoring, could have a significant impact on patient comfort and convenience
which may in turn lead to poorer compliance. One study, Brown et al., (2014), tested a non-contact
system to monitor heart rate and respiratory rate in acute admission patients in a general ward
setting. There were small reductions in the median length of stay in the intervention group on both
the study wards and for ITU stay for escalated patients suggesting more timely transfers. The study
40
was controlled but small, unblinded, and pseudo-randomised and no cost-benefit data were
presented.
Vergales (2014) investigated an automated stand-alone respiratory monitoring alarm in a neonatal
ITU setting. The algorithm to detect apnoea episodes demonstrated greater sensitivity than manual
nurse observations detecting three quarters of clinically significant episodes compared to one
quarter detected by standard nursing observations.
Automated alarms have been shown to improve the proportion of (non-automated) EWS triggers
receiving an appropriate response (Jones 2011). Timeliness of subsequent treatment was difficult to
measure because of poor documentation of attendance by rescue teams or medical staff.
Digitally enabled management of Clinical Data Once patient data is collected decisions need to occur. There are increasingly sophisticated
systems which allow for the integration and presentation of digital patient data once it is captured to
support clinical decision-making. The evidence can be delineated into three distinct applications:
Early Warning Score calculation and response triggering, algorithmic combination of
disparate electronic clinical records for screening,
prognosis or risk scoring, monitoring and decision support
the use of automatically populated templates or artificial intelligence and natural language
processing systems, to create real-time case or handover summaries to improve efficiency
or efficacy of clinical communication.
It is again worth emphasising that the deployment of such systems and the dramatic changes in
working practices they require are only just beginning even in high-income settings such as the US.
Generally current results are mixed with methodological problems in most of the studies and a
failure to fully utilize the resulting “big data” created by such systems.
Effectiveness and accuracy of Early Warning Score (EWS) calculation and response triggering Early detection of deterioration in a patient’s condition, the afferent arm of “recognise and rescue”, is
increasingly employed with the aim of reducing inpatient mortality and morbidity (Jones 2011).
There are a limited number of platforms that have been deployed and evaluated to facilitate the
automatic calculation of EWS and the automatic triggering of a clinical response e.g. Critical Care
outreach team in response to manually entered readings. Such systems typically combine bedside
41
records of vital signs and electronic nursing observations such a urine output and calculate a score
by weighting each element in line with published or locally created algorithms e.g. ViEWS, NEWS.
The performance of digital systems in terms of sensitivity, specificity and any improvement in patient
outcomes is potentially heavily dependent on the performance of the underlying equations and
thresholds (Romero-Brufau et al., 2014) with most published scores having high “false alarm” rates.
Research evidence into the derivation and calibration of the scoring systems is excluded from this
review unless specifically reported in digital implementation studies. More importantly there is very
little evidence of changes to workflows and staff behavior to maximise the impacts of or guide the
successful implementation of such systems.
Sawyer et al., (2011) demonstrated that automated alerts could reliably give early warning of
deterioration as a result of sepsis but subsequent work showed that such alerts did not reduce
hospital LOS or other outcomes when delivered automatically to senior nursing staff (Bailey et al.,
2013). A further randomised trial of electronic alerts delivered directly to “Rapid Response Team
nurses” by contrast, resulted in increased intensity of monitoring and did demonstrate a reduction in
hospital LOS by one day in a randomized trial (Kollef et al., 2014). There was no effect on rates of
transfer to ITU, mortality or need for long-term care at discharge beyond chance.
The remote telemetry and consultation systems employed in some ITU settings (with a consultant
advising on the care of large numbers of patients (e.g. Institute for Child Health, London) were not
considered for this review.
Prognosis/risk scoring, monitoring and decision support Systems using algorithms or analytics to combine vital signs, laboratory results, nursing
observations are rapidly being developed. Tepas et al., (2013) studied an automated system to
integrate patient parameters, laboratory data and nursing observations and other risk factors
interactively to calculate a Rothman Index of severity of illness. This correlated strongly with
morbidity and mortality in the peri-operative period. The automated score gave accurate predictions
of length of stay and mortality potentially allowing early intensive intervention and reductions in use
of unplanned intensive care. There was no clinical intervention, cost –benefit was not estimated and
the system required sophisticated integration of many clinical variables suggesting that widespread
adoption will take time. Woodford et al., (2012) observed the use of continuous monitoring of routine
vital signs recorded during pre-hospital trauma transfers of 120 patients and found that such
observations had similar sensitivity to conventional prognostic scores dependent on complex
42
investigations, suggesting potential benefits from earlier, cheaper and more accurate triage of such
patients.
A system called Artemis was implemented in a US Neonatal ITU (Blount et al., 2010 and McGregor
et al., 2011) to integrate clinical variables in real-time to assist with clinical decision–making and to
facilitate data-mining to develop new indicators of deterioration. Laboratory reports, nursing
observations and other elements of the EHR were combined to create prognostic calculations
updated iteratively for each patient. The three published papers relating to this system describe
detailed technical specifications, development and pilot implementation of the system in a NITU but
no clinical or outcome data are presented. Even where systems have such capability it is often the
case that even some early-adopting US hospitals are playing catch-up in developing the clinical or
statistical expertise needed to benefit fully from such an unprecedented wealth of information
(Wachter, 2015).
Better coordination of existing fragmented care may be easier both to achieve and demonstrate. A
study of an electronic system to facilitate immunosuppression therapy post-liver transplant was
conducted in a tertiary center in the US (Park 2010). The system automatically integrated historical
laboratory and prescribing data previously stored in multiple places. This allowed a liver specialist
nurse to improve the coordination of clinical teams to make immunosuppressant prescribing safer
and more efficient. There were significant reductions in toxicity and rejection events and estimated
cost savings were demonstrated in the 12 months after the implementation of the system. Caution
needs to be applied to these results, particularly in regards to the reproduction of similar results, as
it is a single system and implementation followed a safety review triggered by poor practice and
errors. Costs were imputed rather than actual but the large effect size detected suggests that the
system warrants further evaluation and other similar applications are possible. It is also unclear from
the article where this system lies on the continuum from HIT-mediated accessibility and usability of
data which is relatively simple but has high impact and sophisticated automated integration of data
to generate predictive indices and other decision support applications.
An observational before and after study examined the effect of the introduction of an electronic
physiological surveillance system (VitalPAC) which automatically stores physiological data,
calculates EWS scores and provides bedside decision support (Schmidt et al 2014). This rigorous
study provides evidence that full implementation was followed by a reduction in mortality at two
separate sites (7.75% to 6.42% and 7.57% to 6.15%) over a relatively long follow-up period. No
adjustment for existing reductions in national trends for in-hospital mortality were made and so it is
unclear how big an impact the system itself made. Jones (2013) describes the nurse-led
43
development of an automated EWS alerting system which displayed colour-coded warnings on
hand-held and desktop devices after manually entered vital signs data were added to the hospital
EHR. There was a reduction from 16 to 3 cardiac arrests after implementation of the system but
overall or case-mix adjusted hospital mortality was not reported.
Effectiveness and accuracy of the generation of automated Care summaries/handovers Care summaries, handover documentation and “ward rounds” all require the combination of recent
historical data and physiological observations to be efficient and effective. Such functions are
traditionally the preserve of nursing and junior medical staff and are time-consuming and subject to
errors and omissions (Randmaa et al., 2014). Systems have been developed for automatically
producing such summaries. Natural Language processing (NLP) systems are increasingly applied
to unstructured medical record data and have demonstrated very high levels of accuracy in
producing quantitative data sets for billing and research purposes (Murff, 2011, Fitzhenry 2013,
Nelson 2015). Such systems could help to generate shift care summaries with time savings and
accuracy increases as possible benefits.
Hunter (2012) developed a system with researchers from the University of Edinburgh to generate
automated shift to shift hand-over summaries for nursing staff working in neonatal intensive care.
Integration of quantitative and laboratory data along with NLP of free-text nursing care records
entered was used by the modular system “BT-Nurse” to create a shift “time-line” and summary.
Staff rated these automated handover records favourably. They noted potentially great benefits in
accuracy, safety and completeness particularly when compared to more junior staff giving
conventional handovers, however, there was no formal evaluation of these effects included in the
pilot study. This system is also dependent on the presence of sophisticated electronic data capture,
EHR and some modifications in staff record keeping practices and so may not be transferrable to
other organisation.
There are less sophisticated computer –generated summaries, which use an automatically
populated template, these have been shown to reduce hand-over duration by 44mins (Kochendorfer
et al., 2010). Accuracy has been shown to improve where a handover tool is linked the EHR and a
preparation time which could range from up to 20 minutes per day was reduced to up to 15 minutes
per day was reported for hand-over summaries by medical staff in a Neonatal ITU setting (Palma et
al 2011). A similar system showed no difference in the more clinically valid measure of time to
receive first clinical intervention after handover in a surgical setting but a reduction in median length
of stay by one day of borderline statistical significance may represent a real effect, perhaps by
44
reducing hand-over errors or increasing continuity of care (Ryan et al., 2011). Wohlauer et al.,
(2012) found an 11 minute reduction in preparation time by junior doctors for ward-round and hand-
over summaries, using automatically populated templates, along with a marginal reduction in the
numbers of patient “missed” during the ward round. Overall the evidence from such studies is mixed
and there is little evidence from a socio-technical perspective on the impact on working practices,
other clinical systems or objective accuracy or usefulness of such information in daily practice.
45
Capability Details of
review and
search terms
Effectiveness and
accuracy
Financial impact Impact for patients
and staff
Learning for implementation
4.5
Safer clinical
interventions
Records found
677; abstracts
assessed 182; 44
papers included.
Selected search
terms:
BCMA, ADC,
Adverse Drug
Events,
Medication
Administration
Errors
nurs*, barcode,
RFID, QR code,
blood transfusion,
positive patient
identification, total
opportunity error
Evidence of
effectiveness of
HIT in reducing
drug and blood
administration
errors is currently
limited but of high
quality showing
significant
impacts on safety
Bar codes can
reduce blood
administration
errors and some
drug
administration
errors.
Long term follow-
up studies
(including socio-
technical and
qualitative work)
are needed to
understand
workflows,
workarounds and
unintended
consequences
Broad financial data about
the cost of errors available
but savings are assumed
rather than measured and
not set against investment
required
Qualitative estimates of
severity of patient impact
from errors hampers reliable
cost-benefit estimation.
Some nurse resource saving
if single checker policy is
adopted with no loss of
safety.
One high quality study of a
single “whole chain” system
for blood transfusion in an
acute NHS Trust estimated a
net saving of £598k per year
Cost benefit highly context
specific e.g. high saving in
liver transplant aftercare
(US) but less impact in some
standard hospital settings
Limited cost/benefit reported
for UK studies
Faster and more
appropriate
delivery of blood
products carries
patient benefit
and may
improve
resource use by
clinicians
Reduction of
sampling and
labelling errors
reduce common
harm of further
venepuncture
and rarer harm
of
incompatibility
errors.
Blood laboratory
staff reported
reduced stress
at times of peak
demand
because of
automation of
many
processes.
Some studies with robust
observational outcome data
showed reductions in errors
When outcomes are restricted to
events causing harm, safety
benefits are often modest for drug
administration.
Uncontrolled before and after
studies carry risk of bias given
focus on errors during
implementation and study phases
may contribute to improvement.
Incident and error reporting data
is not robust enough to clearly
demonstrate the safety impact of
HIT.
“End to end” HIT likely to provide
the greatest safety benefits but
there is little cost/benefit data.
Socio-technological evidence
needed to understand
workarounds to better understand
how safety improvements are
actually achieved and can be
“locked in”
Systems generating automated
warnings may generate high
volumes of false or un-informative
alarms.
46
4.5. Safer care interventions The use of health information technology (HIT) systems to reduce errors in patient and product
identification, have mainly involved the employment of Bar Code Medicines Administration (BCMA)
and blood administration systems. These systems have the potential to demonstrate significant
reductions in “wrong patient” and “wrong product” errors by automating the processes of checking
normally undertaken by clinical staff, primarily nurses. When integrated with other HIT systems such
as Computerised Physician Order Entry (CPOE) and Automated Dispensing Cabinets (ADC),
BCMA systems may deliver greater improvements in safety than stand-alone systems. Highly
integrated systems are referred to as “end to end” (mainly drug delivery) or “whole chain” (Blood
delivery) and offer automation of the product management at every step. In addition, efficiency
improvements can be self-reinforcing for example where ADC systems could increase drug
availability because of automatic stock monitoring and therefore reduce omission errors or where
“whole-chain” HIT blood management systems could reduce delays in blood delivery which in turn
may reduce pre-emptive ordering by doctors thereby cutting wastage. Currently there is only limited
literature documenting research into BCMA and ADC standalone systems and the larger literature
on CPOE systems have a low UK installed base and are not included current review.
Patient or medicinal product identification and linking during the administration process can be
accomplished using bar-code systems or Radio Frequency Identification Chip (RFID) or similar
machine readable formats. Scanners may be connected to a mobile computer or computer enabled
drug cabinet or as a handheld device such as a PDA or smartphone with a suitable camera and
software. Generally the interventions reviewed here feature proprietary technology but increasingly
“off the shelf” products are used particularly with regards to hand-held devices such as PDAs and
smartphones. These are attractive because of relatively low unit costs. Such platforms support more
or less customisable “apps” or run proprietary software. In general the reliability of such devices has
drastically improved in recent years but may still exhibit problems during implementation which may
or may not be resolvable at site or necessitate “workarounds” to fit with existing workflows. Devices
may be subject to hardware or software limitations (interface usability, limited battery life) and all
such devices are subject to network problems (WLAN black-spots, dropped connections) which may
hamper effective implementation. These problems may feature in early-phase deployment but also
in the overall system life-cycle and where the exact details of supplier support and contractual
arrangements may become very important. There was little available evidence of such problems in
the literature although this is likely to indicate a degree of publication bias e.g. suppressed for
commercial or reputational reasons. Initial failure to integrate systems into established workflows,
47
poor design features and more rarely systemic failures can lead to “workarounds”, ad hoc staff
practices and behaviours designed to bypass an organisation’s systems and processes to
accomplish an activity (Wachter 2015). Ways to evaluate and reduce such practices have been
reported, for example where an override feature becomes overused (Pockras 2013). These effects
often dilute the predicted impacts on patient safety but are rarely reported. It is essential that the
NHS finds ways to learn lessons from early adoptions to speed up the innovation cycle and
maximize return on investment.
Nurses have a key role to play in the delivery of safe interventions in the areas of drug
administration and transfusion of blood products. Nurse administration of drugs and blood products
is the last step in complex chains of processes. Interception and reporting of errors at this point is
lower than at any other step having been estimated at around 2% of actual errors (Young et al.,
2010, Agrawal 2009, Leung et al., 2014). This means that HIT systems designed to reduce errors at
administration, wherever the error originates, have significant potential to prevent harm and waste.
This review examined the literature which has emerged in relation to drug administration and blood
products through HIT patient identification and product delivery systems. This review focusses on
technology to assist nurses in drug administration as they are by far the main administrators of
drugs and transfusions in the NHS.
Drug administration Leape et al., (1995), Kohn and Donaldson (2000) and Bates et al., (1998) describe evidence that
the drug prescription and administration stages of the medicines delivery process generate 39% and
38% of all drug errors respectively. In a review of 91 papers, Keers et al., (2013) found a lower rate
for administration errors of 19.6% falling to 8.0% when timing errors were excluded. A higher
reported incidence of errors for intravenous medication administration of 20.7% was reported again
after excluding timing errors. In terms of overall cost, the only data found was from the US where it
has been estimated that between 380 000 and 450 000 Adverse Drug Events (ADE including both
prescription and administration errors)) occur per year in the US costing $3.5 billion (a mean $8,433
per ADE (Aspden 2006 cited in Van der Linden et al., 2013).
Error detection and near-miss events are variously identified and measured. There is an emerging
consensus that direct and where possible covert, observation delivers the most robust metric
(Cottney 2014, Berdot et al., 2013, Keers et al., 2013). Direct or covert observation is justifiable
methodologically but is a costly research method and therefore underused. Drug safety could also
48
be addressed through Global Trigger Tools and other more efficient adverse event sampling
strategies but no such papers were identified during this review.
Drug administration errors are counted in a variety of ways and the choice of denominator is another
key to validity and generalisability. The most appropriate rate is “Total Opportunity [for] Error” (TOE)
which is defined as directly observed errors as a proportion of all prescribed and “as required” doses
(Keers 2013). Many studies exclude “wrong time” errors e.g. administration defined as +\- 30 or 60
minutes from prescription mandated time) given their lower likely harmful impact and the more
intractable nature of the problem which often arises from lack of nursing resources or other barriers.
Omission errors are also difficult to interpret given that documentation of rationale is often poor but
might often justify the nurses’ decisions e.g. based on a physiological observation being out of safe
range such as Heart Rate. A second important metric is Adverse Drug Events (ADE) although
these are variously defined, sometimes including prescription errors, adverse reactions or
reintroduction of drugs where previous allergy events occurred. ADEs have also been sub-
categorised by severity whether actual or predicted. Overall it appears that ADEs, like all drug
errors, are under-reported as incident reporting is known to be a poor source of data and sub-
clinical events may pass unnoticed.
Error rates may be higher in paediatric settings owing to small volumes of medications requiring
greater precision, the need for and complexity of weight-dose calculations, the use of weight
estimation methods in emergency settings and the variety of alternative preparations and route
choices (Chua et al., 2010). For example, even bar-code reading errors have been noted on
Neonatal ITU possibly because of the small radius of the children’s’ limbs resulting in poor fitting of
wrist-bands (Cochran et al., 2007). Morriss et al., (2009) reported an increase in errors of
administration after BCMA was introduced onto a neonatal ITU but a reduction in harmful events.
This may merely be a result of the fact that such systems dramatically increase the recording and
accuracy of administration times (and thus inflate the degree of error from this source) whilst
reducing actual patient harms making some studies difficult to interpret and generalisable
cost/benefit estimations more problematic.
DeYoung et al., (2009) and Helmons et al. (2009) both examined bar-code medication system
effects in ITU and ward settings on medication errors. They reported before and after rates of 19.7%
and 10.7% falling to 8.7% and 9.9% which were significant and non-significant respectively.
Rodriguez-Gonzalez et al., (2012) in a "disguised" observational study after BCMA implementation
in two GI units, found no "wrong patient" errors at all. Of the TOE of 20.7%, most errors (13.9%)
49
resulted from ignorance of nurses as to elements of administration e.g. before food or administration
via the wrong route. Hardmeier (2014) conducted a direct observation, post-implementation study of
BCMA and eMAR introduced into paediatric wards and found an error rate of 5% of all observed
doses with compliance varying by process and unit. The comparison of drug to the eMAR twice
(instead of once) was achieved 68% of the time. Compliance in checking two forms of Patient ID
varied between 31% and 70% by unit. However in both cases there was no available comparison
with an existing baseline as with many transitions from paper to computerised system. Koppel et
al., (2008) undertook an observational mixed-methods study at 5 hospitals using BCMA systems
and developed a typology of clinical staff workarounds and identified 15 types of workaround
(including collecting BC wristbands at the drug trolley and carrying pre-scanned medications to the
bedside) arising from 31 separate causes including damaged / absent wrist bands, hardware
malfunctions, non-BC medications, emergencies and failed connectivity.
Coleman and colleagues conducted a well-designed evaluation using a combined Computerised
Physician Order Entry (CPOE) and Administration recording system of four different interventions to
reduce missed or overdue drug doses (Coleman et al., 2013). Data from the system were analysed
over time to demonstrate the independent effect of each implementation against existing trends over
a 4-year period. The overall effect of the interventions was to reduce antibiotic and non-antibiotic
omission rates from 10.3% to 4.4% and 16.4% to 8.2% respectively. The time-series analysis
indicated that the introduction of a prescription pause function for prescribers, a clinical dashboard
reporting missed doses and root cause analysis meetings were associated with significant “step-
change” reductions in omitted doses of antibiotics but real-time visual reporting to staff of delayed
doses, was not. The sophisticated analysis allowed the inclusion of existing safety trends to be
accounted for and to evaluate sequential interventions directly. Serrano et al., (2012) reported on
the effects of the introduction of BCMA systems onto an onco-haematology unit and found, during a
short follow-up period, that significant errors were detected by the system including attempts to re-
administer the same prescription and wrong patient identification. Holmstock et al. (2014) used a
time-series design to analyse dose omissions and drug administration efficiency during the
implementation of an Electronic Prescribing and Medication Administration system on one medical
and one surgical ward. Only a short report of this work could be found and although omission rates
are low compared to other papers found no evidence is reported to demonstrate that this was
directly related to the introduction of the system.
The impact of BCMAs on time spent on drug administration is not conclusive either in terms of the
reducing time taken or whether systems themselves are bypassed or otherwise subverted to save
50
time. Dibbed et al., (2011) reported an increase in time spent on patient care vs drug treatment
from introduction of a BCMA system. Hassink et al., (2012) found no difference and Franklin et al.,
(2007) an increase in time that nurses spent on medication administration tasks after
implementation. Poon et al., (2010) reported that 20% of drug scanning steps were missed and
Rack et al. (2012) reported nurse focus group evidence that a majority of nurses said they failed to
use the scanner for patient or drug identification at least once on the previous shift. It is known that
interruptions to medication tasks result in greater error rates (Raban and Westbrook 2014) but no
evidence was found to suggest medication-related HIT could have a direct effect on this problem.
Blood Products The administration of blood products is a key element of care in many settings, primarily GI
medicine, trauma, haematology and to a lesser extent, oncology and obstetrics. Errors may occur in
numerous steps in the process from collection of products from pharmacy to final administration.
The review found no published evidence for interventions to reduce errors during the donation and
production phases of the blood supply chain both carried out by specialist nurses. Evidence is
restricted to studies of patient sampling and administration where available.
Pre-analytic errors (mislabeling or incorrect patient sample) are a leading cause of health laboratory
error but are almost certainly underreported as an unquantifiable number of errors remain
undetected (Snyder et al., 2012). Transfusion to the wrong recipient is referred to as a 'never event'
emphasising its gravity. Nurses play a key role in the administration of blood, in both the correct
collection and labelling of “group and save” and cross-matching samples from patients in hospital
and the subsequent administration of products to patients. “Wrong blood in tube” errors (WBIT) vary
between 1 in every 1500 to 3000 samples (Cottrell 2013a, 2013b) and are not as rare as transfusion
errors allowing some statistical inferences to be made. HIT interventions have been deployed to
reduce error during both of these steps. Other important benefits of HIT are likely to arise from
increased efficiency and reduced wastage of a finite resource. The Cross-matched to Transfused
Ratio (CTR) provides a good proxy for safety and efficiency. This is the proportion of units
transfused that have received patient-donor specific antigen testing. Lower ratios indicate more
efficient use of blood with higher ratios associated with increased laboratory costs and waste. In
emergency situations restricted testing is employed (e.g. “type specific”) or “universal donor” blood
(UDB) is supplied. The risk of haemolytic transfusion reaction has been shown to be low in
emergency settings using type-specific blood although fully cross-matched blood is still preferable
(Goodell et al. 2010). However their devastating patient impact and high cost means they warrant
preventative interventions with the aim of their complete elimination.
51
There is evidence from a single site study which suggests that HIT such as Bar Code matching of
recipient, sample and product can reduce WBIT errors by approximately half from 1 in every 12,322
to 1 in every 26,690 samples and wrong products transfused from 1 in 27,523 to 1 in 67,935
(Murphy 2012b). An automated alarm was triggered if a BC mismatch was detected at both the
bedside and remotely within the transfusion laboratory. Of the 343 such alarms triggered 37 were
“true” mismatches and all of these potentially devastating events were avoided.
Bar Code systems have also been shown to increase other markers of safety such as closer
adherence to standard procedures amongst nurses (Murphy 2012a & 2012b; Miller et al., 2013). It
was shown that, in common with other applications of safety HIT, benefits may arise from the
enforcement of existing rules or the prevention of short-cuts and multi-tasking rather than merely
from automation of checking steps. These important socio-technical mechanisms which can
contribute to improved safety are often not reported in full but could provide explanations as to why
some systems lead to workarounds compromising safety whilst others do not. This lack of detailed
examination of mechanisms hampers much of the HIT literature and reduces its usefulness to the
NHS as a whole.
Reducing the time to delivery of blood products increases the likelihood of full cross-matching being
possible without reducing time-to-vein in emergencies. This in turn reduces wasteful pre-emptive
ordering, overuse of universal donor blood and increases physician acceptability of automated
systems (Murphy 2012b). Elsewhere, Murphy (2012c) presents estimated net savings of £598k per
year with yearly costs of £357k for system, support and training set against yearly gross savings of
£955k as a result of improved productivity, greater than national trend reductions in blood use (12%
vs 9% national from SHOT data) and reduced wastage (estimates are reported in (Murphy 2012c) in
US dollars and converted here using December 2012 mean exchange rate of 1.61 (source:
http://www.x-rates.com last accessed 17/06/2015).
Murphy (2012b) found a combination of patient barcodes, hand held devices and electronically
controlled fridges led to a faster supply of urgent blood from 18 minutes to 45 seconds. Miller et al.
(2013) conducted an uncontrolled before and after study of implementation of 2D BC in single
haematology/oncology clinic in Australia and noted positive verbal identification increased from 57%
to 94%, cross-referencing wrist band to patient ID increased from 36% to 94% and cross-
referencing of “compatibility tag” to wrist band increased from 48% to 99%. Nurses were able to
scan the BC at the bedside and print labels on demand with portable printer further increasing
efficiency. Labelling errors decreased from 103 per year to 8. There was no detailed staff, time or
workflow analysis presented.
52
Collectively, the relatively short follow-up phases for these studies may not adequately capture
longer-term shifts in acceptance by staff or the development of workarounds or indeed improved
efficiencies as familiarity increases and systems “bed in”. Murphy (2012b) did report that the BC
technology supported a reduction from 2 to 1 nurse required for bedside checking whilst actually
improving safety rates. As with other HIT efficiency gains, this can be considered a financial saving
or as a nursing quality improvement as time savings can be reinvested in improving patient
experience.
Sellen et al. (2015) conducted a systematic review but found only 13 papers (reporting 6 studies) in
this area. They included grey and qualitative literature as long as there was evidence of a "research
method". No studies were found that analysed serious transfusion events as these are so rare but
Cross-matched to Transfused Ratio (CTR) is used as a proxy measure of safety and efficiency.
Reductions in time to blood issue of between 96% and 98% were found with reductions in CTR of
between 23% and 65%. Other relevant results were blood waste due to hardware failure, staff time
saving (of around 100 hours per week), increased traceability of units and lowered laboratory staff
stress during emergencies or high-demand periods. Cottrell (2013) conducted a systematic review
of blood transfusion technologies and found evidence for the effectiveness of various interventions
to reduce WBIT errors. All studies showed improvement but did not demonstrate longer-term
sustainability and it is not clear which intervention or combination of interventions is superior.
Effectiveness and accuracy of other safety systems Evaluations of CPOE systems themselves are not included in this review although advantages may
include reductions in prescription-related errors in terms of legibility, timing (as a result of alerts and
instant updates), availability (compared to a single paper drug card) and completeness (automatic
inclusion of essential elements such as route or rate instructions) which could all increase safety.
We have reviewed articles where outcomes directly relevant to administration errors were included.
No papers were identified formally evaluating interpretation of prescriptions by nursing staff at the
point of administration or enforcing pre- administration checks for allergy, pain, heart rate etc.
although these would have been relevant and systems increasingly include such safety functions.
Fitzhenry (2011) studied a BCMA enabled warfarin dosing system which included automated dose
error alerting. Alerts were scored from 0-10 in terms of the likely clinical patient impact by staff.
Only 99 of 2404 alerts were deemed “clinically meaningful” using a subjective numerical rating
scale. Further evidence regarding CPOE emerges from Manias et al’s., (2012) review of evidence
of the impact on drug errors in ITU from a variety of safety interventions. This showed that CPOE
did not always reduce prescribing errors but did lead to reductions in harmful Adverse Drug Events
53
(ADEs). The same review found evidence that smart pumps had no effect on infusion errors and a
25% workaround rate was observed during the study period in one paper in an adult ITU setting.
Two further reviews of BC technology and drug administration errors reached different conclusions
with Young et al., (2010) suggesting that BCMA had mixed effects on error rates and Leung et al.,
(2014) finding up to 80% reductions in administration errors where systems were integrated with
CPOE and pharmacy processes. The complexity of such technologies and the hospital workflows
within which they are embedded combined with the expense of rigorous evaluations of individual
implementations and lack of baseline data, means there is insufficient published evidence to
substantiate projected benefits.
In a before and after study based on two Intensive Care units (Chapuis et al., 2010) the rate of all
administration errors fell further in a unit equipped with ADCs compared to a control unit with
traditional drug storage but there was no difference in errors leading to patient harm suggesting the
multifactorial nature of drug administration errors. Cottney and Innes (2015) showed that other
factors affecting safety, including interruptions, the number of “stat” (due immediately) or “pro re
nata” (as required, at the discretion of nursing or medical staff or request of the patient) doses, bed
occupancy and burden of drug administration workload all increase error rates significantly and
none of these are directly amenable to a technological fix as these do not reduce workload for
nurses to any great extent.
Theoretical and qualitative work
Toussi et al., (2014) carried out a “failure mode and effects analysis” study of medication safety. A
panel examined medication processes and ascribed scores for likelihood and severity of different
errors. Potential errors were still possible despite automation but such systems were more
amenable to the inclusion of decision support and automated alert features reducing the risk some
error types significantly. Finally, Lawler’s et al. (2011) systematic review of the socio-technical
aspects of HIT implementation showed that BCMA does improve medication administration
accuracy when implementation accommodates actual workflows and staff needs and priorities but
no cost/benefit data are presented. Lawler commented on communication difficulties between
teams and suggests workarounds with unintended consequences tend to emerge because of poor
design or fit with existing care processes.
Song et al., (2015), used a survey technique to examine the views of nurses already using a BC
system and found that feedback and communication regarding errors were positively associated
54
with perceived usefulness and ease of use, whereas the age of staff was a negatively associated
with intention to use the systems. Smeuler et al., (2014) conducted a qualitative study of 20 nurses
regarding their role in drug administration safety finding that level of nursing knowledge, experience
and workload were all significant factors impacting on nurses' ability to perform safe drug delivery.
All of which may reduce the efficacy of HIT interventions unless automated alerts and reminders
implemented via ADCs can be shown to mitigate these effects directly. So far this has not been
demonstrated in the academic literature.
55
Capability Details of review
and search terms
Effectiveness and
accuracy
Financial impact Impact for
patients and staff
Learning for implementation
4.6
Digital nursing
dashboards
744 records found;
93 abstracts
screened;
45 papers included.
Search terms
Nursing dashboards
Dashboards
Clinical dashboards
Few papers detail
the importance of
dashboards
focusing on the
quality of the
measures not the
quantity.
The impacts on
clinical nursing
are not clear in
the literature
although there are
improvements
reported such as
increased
communication
and information
sharing.
Lack of
quantifiable data
from EHRs
Unclear how
accurate or ‘real-
time’ dashboard
data can be.
Limited financial
information in
the literature.
One example of
how much these
systems cost
and savings in
terms of a GP
urgent care
dashboard
preventing
hospital
admissions –
possible causal
link.
No mention
of patient
perspectives
or utility
Staff
feedback
positive -
dashboard
had helped
the nurses
achieve
outcomes of
care.
Some
feedback
that decision
support tools
were
required as
visual
presentation
of the data
was
insufficient to
achieve
changes
Involvement of the staff in the
design is important – need for
evidence on increasing impact as
staff may not necessarily use them.
Dashboards need to be accessible
- user guide is useful.
Lack of detail of how data is
analysed longitudinally (as opposed
to day to day reactivity). Snapshots
of real-time data can be very
misleading.
Research is needed into the
motivational effects as evidence for
direct feedback on performance is
limited - only multi-modal
approaches to quality improvement
have any evidence base.
56
4.6. Dashboards “Dashboard” is a word increasingly used in a variety of ways within healthcare to describe frequently
updated graphical displays of various indicators such as bed occupancy. They can be used in the
other capabilities described in this review such as monitoring patient acuity, workforce deployment
and ratios or patient flow performance in Emergency Departments (Spektor et al., 2008, McHugh et
al., 2011, McCaughey et al., 2015). They are increasingly used to inform commissioning in specialist
services and to compare the performance of organisations and individuals (NHS England 2015,
Redwood et al., 2013, Maben et al., 2012). In this section we examine the literature on digital
nursing dashboards and their impact on performance metrics. The key characteristics of quality and
clinical dashboards, which separate them from other systems such as computerized decision
support systems (CDSS) and electronic medical record (EMR) are that they
provide summary data on performance measured against targets (often related to quality of
care or productivity)
use data visualization techniques (such as graphs) to provide intuitive feedback to leaders or
individual clinicians (Dowding et al., 2015).
Dashboards visually display nursing metrics so staff can easily see and understand the standard of
care that is being provided in their unit. Many sources suffer from a lack of detail about the costs of
these dashboards, whether the measures are real-time or a visual display of daily/weekly/monthly
measures and whether they impact on clinical behavior. There is a predominance of examples of
areas developing and introducing a dashboard with little evidence for longer-term impacts.
Nursing metrics are ways of measuring the quality of nursing care. They include the use of
indicators to measure nurse-delivered interventions, outcomes and patient experiences (Griffiths et
al., 2008). The term “metrics” originated in business where it is used to describe and display
measurements of results or production. Nursing metrics are designed to quantify important nursing
care and procedures in a specific area (Foulkes 2011). The policy driver for these is evident in High
Quality Care for All: NHS Next Stage Review (Department of Health DH 2008a) and Framing the
Nursing and Midwifery Contribution: Driving up the Quality of Care (DH 2008b). There is a
significant amount of literature documenting the importance of metrics both in this country and
abroad (Aydin et al., 2008, Brown et al., 2008), Chandraharan 2010, Flannigan et al., 2010, Royal
College of Nursing (RCN) 2011, Doran et al., 2011,Morrow et al., 2012, Felkey and Fox 2014) and
57
work looking at establishing them in certain areas nationally (NHS England 2015). But it is unclear
what effect they have (Dowding et al., 2015).
Nursing metrics are focused on outcomes for which nurses could realistically be held accountable
such as infection rates, number of falls, and number and grade of pressure ulcers (Foulkes 2011),
however metrics need to be visible and accessible if they are to assist in promoting quality
improvement. Dashboards are a visual representation of these metrics such that staff can easily see
the standard of care provided and the impact of specific quality improvement initiatives (Brown et
al., 2010). The way in which information is presented to individuals may also affect the decisions
they make (Workman 2008). The majority of clinical dashboards tend to use the same universally
recognized colour coding, based on the traffic light system e.g. green for performance within targets,
areas in amber where performance is falling slightly behind targets, and areas in red where there
has been a major deviation from the agreed targets, (Flannigan et al., 2010). The idea is to avoid a
common failing: that nurses collect data but do not always ‘own it or understand it’ (Foley and Cox
2013). However the value of such dashboards is dependent upon the quality of the data entered
(Collins and Draycott 2015). The underuse of dashboards in the United Kingdom was noted by the
NHS Next Stage Review (DoH 2008). It advocated the development of dashboards in all clinical
areas and that every provider of NHS services should systematically measure, analyze, and
improve quality, developing local frameworks to use national relevant indicators.
Looking at the literature in this area there is a definite international mix with two themes emerging:
the effectiveness of the dashboards and the perspective of staff introducing and working with
dashboards. The literature is not always specific to nursing as it is often ward teams that are
included rather than the individual professions.
Effectiveness and accuracy A key message which emerged from a Royal College of Nursing summit on clinical dashboards
(RCN 2011) was that the choice of indicators used to measure quality of care was vital and should
be based on what is important to the healthcare service user/patient and nursing team. Otherwise
there was a risk that dashboards could be used to measure what is convenient, rather than what is
necessary. There was also the risk of too much data being displayed.
Simms et al., (2013) in a study of consultant-led maternity units in the South West found that there
were 352 different quality indicators (QIs), covering 37 different indicator categories, with up to 39
different definitions for one particular QI. Fraizer and Williams (2012) in a paper detailing the
introduction of a dashboard system in a US health provider that used monthly data stressed that
58
“de-cluttering” was important and that effective dashboards focused on not the quantity of
measures, but on the quality of the key measures being reported.
The literature contains many examples of dashboards being designed and implemented.
McLaughlin et al., (2014) reported that in 2007, the University of California Los Angeles (UCLA)
Department of Neurosurgery successfully created a quality dashboard to help manage process
measures and outcomes and ultimately to enhance clinical performance and patient care. Guha et
al., (2013) reported on the design of acute gynecology dashboard focusing on activity, workforce
and clinical indicators, updated monthly. They documented that since the creation of the dashboard
it has gained popularity and is now used in a variety of activities including gynecology business
planning, sourcing new equipment, initiation of acute gynecology theatre lists and enabling
consistent feedback on clinical incidents and training of multidisciplinary staff members.
At an organisational level, Clark et al., (2013) described the internal Medicine Program at The
Prince Charles Hospital Queensland, which developed a clinical dashboard that displays locally
relevant information alongside relevant hospital and state wide metrics to inform daily clinical
decision making. The data reported on the clinical dashboard is driven from data sourced from the
electronic patient journey board in real time as well as other Queensland Health data sources,
providing clinicians including nurses, access to a wealth of local unit data presented in a simple
graphical format. This can have the effect of aiding the identification and confirmation of patient flow
problems, assisting identifying the root causes and enabling the evaluation of patient flow solutions.
These are useful studies about the design of dashboards, but for both studies there is a lack of
detail about the effectiveness and the costs associated with implementation.
Two major pushes for dashboard deployment in the UK were the Clinical Dashboard pilot
programme,
http://webarchive.nationalarchives.gov.uk/20130502102046/http://www.connectingforhealth.nhs.uk/s
ystemsandservices/clindash/overview and the QIPP National Urgent Care Clinical Dashboard in
https://www.networks.nhs.uk/nhs-networks/qipp-urgent-care-gp-
dashboard/documents/4718_dashboards_4pp_a4_WEB_AW_FINAL.pdf. The pilot programme
involved the development and implementation of 24 dashboards, encompassing over 500 metrics,
utilising multiple clinical systems across in 12 projects over 10 Strategic Health Authorities. The
QUIPP dashboard, when fully implemented, will display information to help 2000 GP practices pro-
actively manage and co-ordinate their patient’s healthcare, especially for the most vulnerable
patients and those with long-term conditions by providing Real-Time information from the local acute
trust(s) on ED attendances, admissions and discharges combined with near real-time information
59
from out of hours (OOH) and the walk in center to each GP practice, enabling a whole system view
of urgent care. The learning from both of the programmes is difficult to access, with information on
websites, see above.
The QIPP programme built on the success in NHS Bolton Urgent Care Clinical Dashboard (UCCD),
which was developed as part of the NHS Connecting for Health led Clinical Dashboards pilot
programme (Talbot 2012). The UCCD was designed by local GPs and combined information that
they already received from the local Acute Trust on A& E attendances, admissions and discharges,
with information from out-of-hours and the walk-in centre providers. This information was then
provided to 56 GP practices through the dashboard which gave a more complete picture of patient
care. The dashboard, was used in different ways within practices to give GPs, nurses and active
case managers a clearer picture of urgent care activity, helping them pro-actively manage and co-
ordinate patient care. This, alongside a range of other local service improvement initiatives, led to
2009-2010 ED admissions falling by over 3% compared to a regional increase of 9%. Unscheduled
hospital admissions fell by over 4%, with one practice showing reductions of 17 %. It was calculated
that over the year 2009-2010 the reductions equated to cost savings of over £600,000. The costs of
implementation were £6,000 or less for several sites that only needed to purchase software. For
practices which invested in additional staffing and/or software, the average total spend was £43,000
(Hedgecock and Lewis 2012).
Dashboards do not always present the clarity which is expected. Shaw et al., (2013) examined the
introduction of an electronic health record (EHR) query tool which extracted EHR data from a
pediatric intensive care unit for 295 patients to provide a real-time visual display showing
compliance with patient safety activities. The display included data on presence of consent for
treatment, restraint orders and deep venous thrombosis (DVT) prophylaxis. Baseline compliance
and duration of non-compliance with these activities were established prior to, and one month
following, the activation of the dashboard. Patients included in the analysis were equally divided pre
and post intervention and were similar in age, race, gender and mortality. The median time from
admission to attainment of caregiver consent was 327 minutes in the pre-intervention time period
and 219 minutes in the post-intervention time period, clinically but not statistically significant. There
were no significant differences in compliance rates for DVT prophylaxis or restraint orders between
time periods.
Dowding et al. (2015) performed a literature search from 1996–2012. 122 full text papers were
retrieved of which 11 reported an empirical evaluation of dashboards, reporting either their impact
on desired outcomes (patient/professional behavioral) or clinician perceptions of the utility of such
60
dashboards in clinical practice. Overall the results indicated considerable heterogeneity in
implementation setting, dashboard users and indicators used. In terms of impact there was
evidence that in contexts where dashboards were easily accessible to clinicians (such as in the form
of a screen saver) their use was associated with improved care processes and patient outcomes
(Starmer et al., 2008, Zaydfudim et al., 2009,). Dowding et al., (2015) concluded that it remained
unclear what dashboard characteristics (e.g. type of graphical display, method of presentation) are
related to improved outcomes and how clinicians incorporated use of dashboards into their
everyday practice. Similar to much HIT research Dowding et al., (2015) reported that the majority of
studies had some element of potential bias, with 5 studies being of low quality meaning that any
significant results should be treated with caution. Keen et al., (2014) are due to report in 2017 on a
study examining ward use of dashboards occur at four NHS Trusts and whether they achieve the
openness, transparency and candor envisaged by Francis (2013) and supported by Keogh (2013).
Also work by Cookson et al., (2013) through consultation with a wide range of NHS stakeholders is
developing equity dashboards for presenting findings in an accessible and useful form to local and
national decision makers to identify emerging changes in equity performance and understand how
their actions are influencing inequalities is due to deliver in 2016
Impact on patients and staff In terms of the acceptance by staff, Daley et al., (2013) explored the experiences and attitudes of
mental health professionals working in three acute elderly care areas, to a new clinical dashboard
system which used a variety of different graphics. Metrics were tracked from baseline to 6 months.
Staff completed a questionnaire 3 months after the initial implementation, the results of which
suggested there was improved access to information, increased communication and information-
sharing, increased staff awareness, and data quality associated with the introduction of the
dashboards. Jeffs et al., (2014) in a self-reported qualitative study undertaken in Toronto explored
the perceptions and experiences of 51 front-line nurses and five managers from six units, of the
implementation of a unit-level dashboard. They described the Care Utilising Evidence (CUE)
dashboard as a visual tool that displayed data on the impact of the nursing care provided to
patients. Staff used the tool to keep track of processes of care and patient outcomes and
experiences at a unit level. They were then able to use the performance data to identify quality care
improvements specific to their unit and it also acted as a reminder. Importantly the nurses reported
that before the CUE dashboard they were often unaware of the outcomes of care and that it gave
them encouragement for continuing to provide care.
61
Safdari et al., (2014) and Frazier and Williams (2012) described how involving staff within their ED
department/hospital can be helpful in establishing key performance indicators and designing the
dashboard. Russell et al., (2014) described the collaborative approach taken by NHS Lanarkshire,
which involved nursing staff, programme leads and the eHealth team in the development of a
general ward nursing dashboard as a means of ensuring safe, effective person-centered care. Staff
in 70 acute adult inpatient wards, in three general hospitals, had been assisted to understand and
use it. To ensure consistency of understanding, a user’s guide was developed both within the
dashboard (as an information button) and as a leaflet. The guide helped staff to refresh their
knowledge and develop a deeper understanding of the information available to them because “Data
alone cannot ensure care is of high quality” (Russell et al., 2014 pp 43).
In their small study Effken et al., (2011) explored the decision making of 20 nurse managers in three
Arizona hospitals. They described the nurse manager's environment in terms of the constraints it
imposes on the nurse manager's ability to achieve targeted outcomes through organisational goals
and priorities, functions, processes, as well as work objects and resources (e.g., people, equipment,
technology, and data). They found that nurse managers receive a great deal of data, much in
electronic format. Although dashboards were perceived as helpful because they integrated some
data elements, what was lacking was the decision support tools to help nurse managers with
planning or answering “what if” questions. No nursing example was found of such a decision
support tool, but Dolan et al., (2013) explored the feasibility of such a supportive tool in interactive
decision dashboards in the medical treatment of knee osteoarthritis. They created a computerized,
interactive clinical decision dashboard and performed a pilot test of its clinical feasibility and
acceptability with 25 doctors using a multi-method analysis. The mean time spent interacting with
the dashboard was 4.6 minutes and the interactive clinical decision dashboard scored well for
mechanical and cognitive ease of use, clarification of values, reduction in decisional uncertainty,
provision of decision-related information and decision-aiding effectiveness. In a later paper Dolan
and Veazie (2015) reported that such an interactive decision dashboard Increased confidence in
medical decision making.
62
Capability Details of review and search terms
Effectiveness and accuracy
Financial impact Impact for patients and staff
Learning for implementation
4.7 E-Workforce Deployment
Records found 580; abstracts assessed 28; papers included 12. Search terms: e-roster, nurs*, roster, rota, workforce planning, off-duty, autom*, digital
Lack of baseline data on rota quality and efficiency to assess the impact of implementation
Limited evidence detailing implementation from a practitioner-manager perspective.
Some evidence of improvement in rule adherence and reduction in changes after rota release
No evidence that e-roster is used to increase self-rostering and therefore improve employee wellbeing
Literature mainly anecdotal or narrative - no detailed cost benefit data found
No studies have demonstrated financial savings net of implementation, training and licenses
Significant savings in senior nurse time likely but robust evidence is lacking
Financial impact of savings from improved analytics and management oversight and control has not been reported
There is little evidence that e-roster systems as currently used facilitate greater control for staff
One study reported that the system was acceptable to nurses.
No studies measuring relative “fairness” or other acceptability indicators for e-roster versus existing systems or between units or sites
No studies reported the impact on senior nurses in terms of ease or time saving
No reports of systems that allow greater accessibility to rostered staff
There is a need for research to compare implementation and outcomes between sites now that sufficient systems have been deployed.
The degree to which HIT is integrated and adapted to local practices may strongly influence the realisation of benefits such as safer nursing cover and appropriate skill-mix.
Fidelity to planning during implementation via training and senior monitoring may significantly increase the benefits achieved e.g. by using automation and rule-setting functionality.
Benefits likely in reducing senior nurse time and improving skill-mix and safety - little published evidence.
Analytics to manage costs are possible with integration with ESR and payroll systems – published evidence lacking
Attendance management systems are needed to ensure payroll reflects actual activity and reduce mis-payment
Monitoring compliance may be more important if e-roster is extended to other, less directly supervised staff groups.
63
4.7 E-Workforce Deployment Adequate skill-mix provision is a key determinant in the realisation of safety and care quality
benefits proposed for e-rostering (NHS Employers 2007). Within nursing this concern is increasingly
driven by increases in the introduction of nurse practitioner roles often entailing a great range of
skills sets needed across different wards and departments and at different times. To enable the
management of complex skill-mixes via e-roster systems, considerable up-front data input is
required as well ongoing record “maintenance” to ensure the accuracy of information regarding staff
capabilities, training and accreditation. The evidence discussed below is grouped under four broad
themes: Effectiveness and accuracy of rotas including management of skill-mix; financial impacts;
self-rostering and staff impact acceptability.
Effectiveness, accuracy and skill-mix
The importance of accurate nursing workforce management can hardly be overstated, especially in
the light of recent safety reports e.g. Francis 2013. E-rostering systems have been proposed as
having the potent to deliver improvements in safety such as adequate skill-mix provision and
matching of staff supply to demand. In addition e-roster systems may impact on the degree to which
planned rosters become the actual worked hours and deployment often termed rota “robustness”.
Many commercially available e-roster systems incorporate tools for managing skill-mix which could
support increased productivity by ensuring that required skills are available consistently. No
literature was identified measuring the impact of e-roster implementation on skill mix availability or
indeed the degree to which the functionality of deployed systems was utilised or integrated with
Electronic Staff Records. In addition, accurate information on patient flows, acuity and case-mix are
also required to target workforce planning using an e-roster system. There was no literature
regarding the adaptation of e-roster processes to incorporate information on variability in patient
demand e.g. modelling patient flow and incorporating into e-roster constraint or rule setting. There
was also no literature regarding the use of e-roster analytics in longer-term variations such as
seasonal planning or predicting the need for part-time or annualised hour’s staff.
Where automatically generated rosters are adapted post hoc to reflect unexpected sickness, peaks
in workflow and other threats, this result in reductions in the efficiencies gained from advanced
planning. Very few studies measured the robustness of electronically generated rosters in real-world
64
settings either in terms of their ability to meet soft and hard constraints (e.g. European Working
Time Directive) or adherence to planned shift allocations or staff requests. Wong et al.,(2014)
examined the use of an electronic roster system and found that a two-staged process including an
initial roster meeting demands and requests followed by an iterative adjustment process, delivered
effective rosters when assessed in terms of completeness of cover, skill-mix, fairness, staff-
wellbeing and continuity of staffing across consecutive shifts. There was no comparison to previous
systems and robustness to sickness or workload variation was not assessed.
Drake conducted a study (2014b, 2014c) of e-roster system implementations in NHS hospitals but
found no data describing rosters created prior to implementation to enable a before and after
comparison. Instead cross-sectional comparisons were made between sites after implementation of
various systems. The study found enormous variation in the degree to which automatic functions
were utilised by senior nurses in practice, from 6% to 80% of all shifts allocated. Rule violations and
post hoc changes were more likely where there was a greater reliance on manual roster creation
suggesting that automation performed well in meeting staffing needs where it was used. Increasing
self-rostering was not associated with an increased need for retrospective changes suggesting that
automation using e-rostering could realise the benefits of greater staff autonomy without loss of
roster robustness. Despite the adoption of electronic systems, problems such as unfilled shifts
persisted and contributed to safety concerns. Unfilled shift rates varied across hospitals with rates of
up to 15% of all planned shifts left vacant in some units. E-roster system analytics may make
identifying such problems easier and may permit more effective work force planning but this has not
been tested in any of the literature identified.
The degree to which e-roster implementation is localised at hospital or even ward level, may
strongly influence the realisation of safe and adequate of worked rosters. Although e-roster
implementation for nursing has reached 90% coverage, unscheduled inspections on wards reveal
staff short-falls compared to workforce plans (Hockley and Boyle 2014).
Financial impacts E-roster systems are suggested as likely to positively impact on both high agency spend and payroll
errors (NHS Employers). Ronneberg (2010) reported a reduction in the requirement for agency staff
after the introduction of an e-roster system. The degree to which this is generalisable to other
settings is unclear. The time devoted to the management of rostering by senior nursing staff may
also be reduced by the implementation of electronic systems. Wong et al., (2014) reported that
65
roster generation processes using an electronic system took 10-15 minutes compared with manual
processes taking “2 hours” or more. This was merely to fulfil hard constraints e.g. no empty shifts
implying that further time would have been required to finesse the roster before implementation and
this time was not estimated. Other literature found concurs that a conventional manual approach
may take hours of senior staff time to generate even preliminary versions of a roster and much more
for larger more complex settings (NHS Employers 2007). In addition the size of the ward or unit and
possibly the complexity of shift patterns and workload may also be related to the size of any
potential efficiency benefits derived from automation. Again context is a central concern in predicting
likely cost/benefit improvements. No work was identified in the public domain quantifying this aspect
of e-roster implementation. Hertfordshire Partnership NHS Foundation Trust reported savings of
£0.75m over a three year period after the introduction of an e-roster
(http://www.nhsemployers.org/case-studies-and-resources/2012/11/hertfordshire-partnership-nhs-
foundation-trust---e-rostering-to-reduce-agency-spend last accessed 23/04/2015). Little supporting
data are presented in this case-study. It is not clear where such savings emerged e.g. whether from
reduced agency spending or reductions in other errors arising from to manual processes. It is not
reported whether savings are net of implementation costs e.g. software licenses and training, or
whether they includes indirect costs such as ESR maintenance or residual manual work by nursing
managers in altering rosters post-hoc.
Direct links to payroll systems may also reduce overpayment compared to manual systems, for
example where shifts are cancelled at short notice. Further efficiencies may arise simply from
increased legibility and accuracy of e-rosters compared to paper but this was not reported in the
literature in keeping with the overall lack of evidence from prior to automation regarding roster
practices. Improvements in the accessibility of rosters for nursing staff may increase efficiency and
communication ensuring that planned shifts are actually worked.
The literature on e-roster systems does demonstrate that automation allows roster creation in a
matter of minutes compared to many hours for manual systems (Wong 2014, Glass and Knight
2010, NHS Employers 2007). Overall NHS Employers estimate that typical improvements from the
introduction of e-rostering reduce senior nurse time from three days to about three hours. No
literature was found quantifying such benefits and the degree to which automated rosters need
refinement has not been widely researched in live systems.
66
The National Audit Office (2010) identified large levels of payment errors arising from the failure to
update payroll systems in line with staff turnover, missed shifts or unearned supplemental payments
for unsocial hours during sickness periods. Hockley and Boyle (2014) extrapolate from conservative
local figures to estimate that nationally the NHS may lose up to £30m per year in overpayments due
to rostering errors often arising from failures to link separate roster and payroll systems. No
literature was identified which reported changes in the levels of payment errors as a result of
implementation of linked e-roster systems. The greatest threat to purported payroll savings from e-
roster deployment remains the reliance on manual reporting of shift attendance by supervisory staff.
E-roster deployment has not generally been accompanied by widespread, automation of attendance
information e.g. via biometric “clock-in” systems (Hockley and Boyle 2014). The same report uses
data from Basildon and Thurrock University Hospitals NHS Trust, to illustrate the potential level of
savings to the NHS at over £40m (this paper reports funding by Kronos Ltd, supplier of e-roster
systems to the Trust concerned). It is possible that this estimate is at the upper limit of possible
savings and arises from a single Trust that has fully implemented automated attendance recording
systems rather than a stand-alone e-roster product.
Most of the evidence for the cost-benefit impacts of e-roster systems is anecdotal or assumed to
arise simply from the rich functionality available in many commercially available systems and the
scale of the existing waste. Crucially the limited degree to which previous practice was understood
or analysed, limits the reliability of these assumptions of savings to the NHS. No high-quality, time-
series auditing evidence adjusted for confounding factors has been identified within the literature but
this may be available locally to guide implementation and evaluation of financial impacts and
cost/benefits.
Self-Rostering: Recruitment, Retention, Wellbeing, Acceptability for staff
Many of the benefits of e-rostering are thought to arise from the degree to which such systems can
facilitate greater staff autonomy in managing their working patterns (NHS Employers). There is
strong evidence that increasing employee “work time control” or degree of self-rostering delivers
benefits to employees as a result of improved work-life balance, work-family conflicts and to
employers in the form of job-related outcomes such as improved performance, motivation and
reduced actual or intended turnover (Nijp et al 2012). The same review did not find consistent
67
evidence of any impacts on employee wellbeing or health however and there appears to be little
sign that these measurements are routinely monitored.
Rönnberg and Larsson (2010) used data derived from shift requests in the form of self-written roster
lines submitted by ward nurses, to generate rosters automatically which were then compared to
convention rosters written by ward managers. No cost-benefit data was presented but electronically
generated rosters met compliance constraints to a greater degree than the conventional process. It
was reported that there was less discrepancy between staff requests and what was allocated using
the automated process which is likely to have had a positive impact on staff acceptability.
Drake (2014b, 2014c) qualitatively compared hospital rostering policies (14 NHS hospitals) reporting
that the requirement for “fairness” in rostering processes was the commonest consideration to be
mentioned. Though assertions of increased fairness are widely made in support of e-roster system
implementation, Drake claims that fairness is poorly defined and may be no more nuanced than an
adherence to a few rules (2014a). Such impersonal implementation may not be accepted as “fair”
by nurses in practice. The original automated version of a roster may not survive manual revisions
which impact on objectively important attributes of the roster e.g. adequate skill-mix (2014b). But
perceived fairness may actually depend on manual revisions in the roster process rather than
impersonal automation to reflect individual personal circumstances etc. Unless systems enable staff
to make many personal requests or write their own rota (within limits) easily and quickly, it is unlikely
that even maximally automated e-roster systems will increase satisfaction levels markedly without
subsequent manual revisions. Such radical implementation of e-roster functions may be less time
consuming than manual revisions in a paper system or poorly implemented e-roster. Failure to take
advantage of such functioning could reduce the impacts on “empowerment” and improved well-
being (Drake 2014a).
Overall there is little evidence that e-roster systems are currently used in ways that facilitate greater
staff control nor that they lead to increased acceptability of rosters for staff they cover and therefore
improvements in contingent human resource metrics thought to underpin likely benefits.
No other literature was identified measuring the extent to which self-rostering was facilitated by e-
roster implementation in practice. No studies were identified which measured acceptability of e-
roster implementation in terms of perceived work-life balance, well-being or in actual or proxy
measures of staff turnover.
68
The benefits of e-roster rollout across the NHS fall broadly into those thought to arise directly
(recording, reporting, accessibility and auditing functions) and those that are inferred as secondary
benefits (improved planning and skill-mix and reduced agency spend) but are more or less
contingent on effective and faithful implementation. As with other HIT applications, implementation
may be as important as the original choice of systems and demonstrating planned benefits in terms
of high-level outcomes and performance is difficult. Overall there is as yet, no evidence that the
purported benefits of e-rostering systems have been realised within the NHS and demonstrating
benefits such may require larger comparisons across organisations and/or monitoring of local
practices.
69
5. Limitations and learning The quality of the evidence in the area of nursing HIT is variable and there is a lack of scientific
evaluation for the implementation of existing and developing health technologies in nursing. Four
themes emerge that need consideration: methodological issues, human factors, technological
implications and the health care context.
Methodology A variety of methodological issues arise in the context of HIT and these problems are essentially
those of assessing complex health interventions in general: how to establish tight cause-effect
relations; how to assess the impact of elements of an intervention where confounding effects cannot
be controlled for; how research should be reported so as to defend it against a “purist” research
agenda which demands demonstrable, generally quantitative, impacts on high-level outcomes such
as mortality. Despite widespread deployment of HIT to improve safety, drive quality improvement
and control costs, high-quality evidence of efficacy remains sparse. In many cases poor
methodological approaches such as small sample sizes, short follow-up periods, failure to consider
confounding effects of existing trends and lack of attention to the socio- technical aspects of
implementation and embedding, all reduce the utility of what is published for commissioners and
senior managers. A final overarching problem is that HIT for nursing is in its infancy. There is little
comparable base-line data and no allowance for the cultural lag inherent in transforming workflows
and existing practices needed to maximize the potential benefits of digitisation.
All 21st century digital technology has an increasingly short life-cycle and this is nowhere more
important than in healthcare. As the projects described here mature, ongoing evaluation of
interventions throughout the whole life-cycle of a HIT deployment is called for. This is essential to
avoid missing the development of unintended consequences such as workarounds, redundancy,
misfits with other human and digital systems and asynchronous obsolescence. Local
implementation takes time to demonstrate benefit and requires considerable project management
which must include cultural and behavioural shifts. The evaluation phases governed by the
academic research cycle are typically too short to study maturing systems and so the need for
“more research” is better expressed as “robust ongoing evaluation” reflecting a lack of any definable
destination or endpoint. Academic peer-review is essential to drive methodological refinement but
valid techniques must be made more widely available to NHS organisations themselves so they can
to assess their own progress internally. Sharing of learning must increase dramatically with
70
consideration given to supporting staff in publishing, not necessarily academically, positive and
negative findings, of what is effectively “action research”. Most importantly of all, “failures” must be
recast as learning opportunities and must be shared free of the risk of reputational damage to
individuals and organisations from commercial and political interests.
There was a lack of available information on many important factors such as the costs of equipment,
software, training, support and maintenance of systems. This precludes the establishment of
realistic cost-benefit analyses. Similarly, a reduction of all economic evaluations to financial bottom
lines risks obscuring the sometimes dramatic impacts of releasing nursing time to increase
qualitative aspects of care such as communication and demonstrating compassion which are so
important to patients and families. Robust qualitative and observational data can often indicate
when a quality improvement has occurred even when a financial one has not. Few of the current
evaluations of the implementation phases of HIT projects reviewed here include a robust qualitative
analysis of staff and patient acceptability, behavioral intentions, acculturation or adaptive responses
to technology especially where systems appear to conflict with staff professional pressures and
patient need. Barriers to uptake and usability must be understood and overcome and research
should incorporate socio-technical theoretical models to ensure comprehensive assessment of
impacts on staff and patients as well as accounting for process improvements.
There remains a lack of consensus around the evaluation of highly complex service interventions
(Lilford et al., 2009). Catwell and Sheikh (2009) say that we need to have the means to demonstrate
the “true” benefits of HIT systems. Quantitative outcomes measured with quasi-experimental before
and after methods with suitable controls where possible are the only option in most cases of HIT
investment. Careful selection of controls or individual participants, stepped-wedge implementations
and the inclusion of confounding secular trends in modelling using Interrupted Time Series analysis
and other quasi-experimental techniques are all robust techniques with a good fit to situations where
“the plane must be rebuilt in flight” as the saying goes. These techniques are not all highly complex
and interpretation of longitudinal analysis is perfectly possible for senior clinical nurses with suitable
training and support. Time-series principles have been used in managerial contexts e.g. Statistical
Process Control Charts to highlight unexpected performance deteriorations have been used by non-
academics in industry for decades but remain underused in the NHS. The collection, manipulation
and analysis of the increasing amount of available data in a form useful for senior staff should be
viewed as a core competence for nursing and managerial staff responsible for deploying and
refining HIT.
71
Many templates exist. Talmon et al., (2009) detail a list of principles for properly describing the
evaluation of HIT projects, to allow the reader to place the evaluation the right context and judge the
validity and generalisability of findings. Lilford et al. (2009) discuss ways to address complexity
using pre-implementation and pilot testing incorporating principles from the Medical Research
Council (2008) framework for evaluation of complex interventions. They advocate “methodological
pluralism”, undertaking combined quantitative (“what happened?”) and qualitative work (“why did it
happen?”) when evaluating IT systems, with each informing the other. Greenhalgh et al., (2010)
suggest that HIT programmes should be studied at three levels: macro (national policy, wider social
norms and expectations); meso (organisational processes and routines); micro (particular
experiences of patients and professionals). Such research can result in effective tactics to calibrate
systems and maximise desired impacts and minimize disruption.
Lau et al., (2010) notes that formal studies HIT often lack sufficient power or duration to assess
health outcomes in statistical terms. These limitations can be mitigated by using longitudinal as
suggested above. In many cases, the technology is only part of a complex set of interventions that
included changes in clinical workflow, provider behaviours and scope of practice but these elements
are too often ignored (Wachter 2015). But it is here in the messy detail, that real impacts and
improvements will occur.
Human factors In much of the HIT literature examined in this review the tensions between fidelity and local
adaptation, the emergence of workarounds, underutilised functionality and failure to eradicate
duplication and redundancy is evident but seldom addressed directly. Nurses in particular, given
the unstructured nature of their work and holistic responsibility for patient well-being rather than
narrow treatment, often develop “workarounds” to mitigate the impact of inflexible systems on their
clinical priorities and the priorities of their patients. This hierarchy should be reversed. Workarounds
occur overwhelmingly due to failures in systems and implementation of HIT in context-insensitive
ways (Marini 2009). Rack et al’s (2012) focus group study in the US reported that over half of
nurses interviewed said that they had bypassed the BCMA system in their previous shift. Other
behaviours can include barcodes fixed to beds instead of patients or stored on a drug trolley or in a
remote preparation room (Hardmeier et al., 2014). Lawler e al., 2011, suggests the inclusion of
nurses in the implementation or even procurement phase of HIT planning could help mitigate these
down-stream effects by ensuring that human resource impacts are addressed. There also needs to
be much more exploration of the impact of HIT for the patient. HIT designed to reduce
72
documentation and improve communication should result in more patient contact whilst ensuring
that interprofessional working is enhanced and not displaced. If a nurse can access a social history
on a PC in the office will she/he still spend as much time welcoming a patient to a ward? If a doctor
can issue nurse orders from an office and therefore not inform a nurse that relatives appear upset
and confused about a treatment plan will she do so?
Carlisle (2012) has documented that the biggest barrier to successful HIT implementation is lack of
proper project management and leadership. These are vital to ensuring nurses embrace technology.
Too often she says implementation occurs in organisations that lack the skills to deploy technology
well. There is a lack of clinical engagement, responsive IT support and early process-mapping,
these conceal, significant problems if neglected. This can result in unnecessary resistance to HIT by
staff and is often misinterpreted. For example, nurses were found to be embarrassed because they
felt patients became aware of their lack of keyboard skills and proficiency in using new technology
(Carlisle 2012). Greenhalgh et al., (2008) advocates the widespread use of socio-technical change
models applicable to the implementation of HIT. This could mean that rather than view an EHR as
an end in itself with success measured in terms of reduction in costs of paper records or speed of
implementation. Success could be measured as more frequent and timely access to records or
improvements in legibility and reduced transcription errors. Instead of talking about “implementing
the EHR”, implementers should think of “integration” of the EHR as a part of the infrastructure for a
wider strategic change with a longer timeframe but a greater impact.
Lau et al., (2010) make three recommendations. First, when reproducing a successful project
attention must be paid to specific features and key factors that are critical to “making the system
workable” locally by early and continued involvement of frontline staff and assessments of the
impact on their day-to-day work. Second, contextual issues should be addressed, for example,
where information is required in one place and at one (a clinic) but resides in multiple systems.
Third, return-on-investment can only be judged by ‘measuring the clinical impact […] of
implementation’. Human factors are currently neglected or undervalued in HIT evaluations but are
the key to delivering quality and patient experience improvements.
Technology Although much everyday technology such as smart phones appear flawless we rarely use a fraction
of the available functionality and then only for trivial individual tasks. This is not a good benchmark
by which to judge healthcare HIT. Usability and system resilience have profound effects on
73
acceptance because they have clinical and not just usability impacts. Poor connectivity can result in
prolonged log in times or clinical applications “timing out”, resulting in the loss of work (Kidd 2011).
This is not merely frustrating but often dangerous. As a result of the imperative to deliver care,
hardware and software failures result in staff developing workarounds to side-step a label printer
failure in an emergency room, fingerprint recognition errors on an ADC in a busy psychiatric unit or
network “black-spots” in community midwifes practice area.
Maturation of systems tend towards the resolution of these problems (e.g. Franklin et al., 2007).
This requires recognising (and acknowledging) where and how failures have an impact, identifying
legitimate workarounds and developing safer alternatives (Barber 2010). Other considerations may
include new opportunities for cross-infections such as multiple use keyboards in outpatient clinics,
access to limited computers in pharmacy, restrictions on simultaneous record entry in multiple
trauma teams and the introduction of less efficient workflows. Some HIT implementations fail
because the host organization had insufficient understanding of what it was actually doing before
the HIT implementation and what they really needed a system do.
Wade et al., (2010) hold that telehealth using video has particular requirements and hence
associated costs such as video screens, connectivity of sufficient bandwidth and reliability, physical
space at each location, and health providers’ time to deliver services. Again these must be
considered in the planning stage and incorporated into any evaluation or research as they can
easily derail implementation. Lastly a central problem for evaluating the effectiveness of HIT which
are commercial systems are that their algorithms are not “open-label” and therefore evidence from
academic studies of generic approaches cannot be generalised to their commercial applications.
The Health Technologist
Health Information technology is not simply about replacing paper forms with their electronic
equivalents and when using evidence from HIT in situ it can be difficult to generalise between sites,
settings and especially health systems. In particular, there is a preponderance of evidence from
outside the UK; this should preclude simplistic extrapolation of reported benefits to the NHS. Cost
benefit analyses and even the likelihood of successful implementation are not clear even where high
quality evaluations have been completed if the context is sufficiently different. For example, positive
evaluation of BCMA systems may be partially reliant on “Unit dose" dispensing common in US
tertiary hospitals, where pharmacy prepares and issues individualized drug supplies to facilitate
accurate billing. Such individualised dispensing is rare in UK where drugs are supplied to ward and
74
departments in generic multi-dose packaging. If systems are dependent on one of more such steps
the overall intervention may not deliver the expected safety or cost benefits. The evidence from drug
administration error research suggests that too crude an understanding of the aetiology of drug
errors will reduce understanding of what works, what does not and at what cost.
In the US, the head start over the rest of the world in the use of HIT driven by recent financial
incentives to digitization (e.g. HITECH) has gradually led to the emergence of hybrid clinical-IT
specialists; staff embedded within IT departments but with clinical backgrounds (Wachter 2015).
The American Health Information Management Association (www.ahima.org/) offers accreditation
as a “certified health technology specialist” and in the UK the Council for Health Informatics
Professions offers yearly registrations, a code of conduct and accreditation of training and education
(www.ukchip.org/).
75
6. Conclusion Throughout the research for this review a central theme has surfaced frequently. Existing evidence
and evaluations of HIT along with many implementations themselves have failed to live up to
expectations for similar reasons. Mass implementation of HIT requires not just new computers and
monitors but new ways of working, new skills and new understandings.
Introducing Nursing technologies is a complex process requiring appropriate, responsive and
flexible technology, good infrastructure, effective project management and a clear evaluation system
which may rarely be of a purely academic nature. This rapid review has described the existing
academic and non-academic knowledgebase regarding the likely effectiveness, financial impact and
impacts on staff and patients of seven capabilities of current HIT. Where available, it has
synthesized evidence for their impact on the nurses they are designed to support, the organisations
which implement them and the patients whose outcomes they are designed to improve. The
evidence varies in both scope and quality but rarely is it comprehensive enough to provide
assurance that implementation will be straightforward. This reflects the need for the NHS to learn
rapidly from its own experimentation of these advances. Academic research alone will not provide
the necessary evidence required by the NHS to achieve quality improvement goals. Robust multi-
methodological evaluation is possible and the results from large exemplar projects should receive
wide attention. But small scale, rapid cycle or continuous monitoring of impacts should be central to
any local implementation plan and sufficient skills and external support should be available to
clinicians and IT specialists within each organization to enable them to undertake planning and
evaluation where it is needed.
76
7. Contacts
Dr Lucy Sitton- Kent Research Fellow for Innovation and Translation East Midlands Academic Health Science Network And Nottingham University Business School Level C, Institute of Mental Health University of Nottingham, Innovation Park, Triumph Road, Nottingham, NG7 2TU t: +44 (0) 115 7484262 e: lucy.sitton-kent@nottingham.ac.uk w: http://www.emahsn.org.uk/ Dr Philip Miller Knowledge and Evaluation Researcher East Midlands Academic Health Science Network Level C, Institute of Mental Health University of Nottingham, Innovation Park, Triumph Road, Nottingham, NG7 2TU t: +44 (0) 115 7432457 e: Philip.miller@nottingham.ac.uk w: http://www.emahsn.org.uk/ Professor Rachel Munton East Midlands Academic Health Science Network Level C, Institute of Mental Health University of Nottingham, Innovation Park, Triumph Road, Nottingham, NG7 2TU t: +44 (0) 115 8231300 e: rachel.munton@nottingham.ac.uk w: http://www.emahsn.org.uk/
77
8. References Agrawal, A., (2009) Medication errors: prevention using information technology systems. British Journal of Clinical Pharmacology, 67(6), 681-686. Audebert, H. J., Schultes, K., Tietz, V., Heuschmann, P. U., Bogdahn, U., Haberl, R. L., and Schenkel, J. (2009). Long-term effects of specialized stroke care with telemedicine support in community hospitals on behalf of the Telemedical Project for Integrative Stroke Care (TEMPiS). Stroke, 40(3), 902-908. Aydin, C.E., Burnes Bolton, L., Donaldson, N., Storer Brown, D., Mukerji, A., (2008) Beyond Nursing
Quality Measurement: The Nation’s First Regional Nursing Virtual Dashboard,
http://www.ncbi.nlm.nih.gov/books/NBK43614/
Bailey, T.C., Chen, Y., Mao, Y., Lu, C., Hackmann, G., Micek, S.T., Heard, K.M., Faulkner, K.M., and Kollef, M.H., (2013) A trial of a real-time Alert for clinical deterioration in Patients hospitalized on general medical wards. Journal of Hospital Medicine, 8(5), 236-242. Barber, N, (2010) Electronic prescribing – safer, faster, better? Journal of Health Services Research & Policy. 15: 64-67 Barber, M., Frieslick, J., Maclean, A., Williams, J., and Reoch, A., (2015). The Western Isles Stroke Telerehabilitation (Specialist Medical Consultation) Service–implementation and evaluation. European Research in Telemedicine/La Recherche Européenne en Télémédecine, 4(1), 19-24. Bates, D. W., Leape, L. L., Cullen, D. J., Laird, N., Petersen, L. A., Teich, J. M., ... and Seger, D. L. (1998). Effect of computerized physician order entry and a team intervention on prevention of serious medication errors. Jama, 280(15), 1311-1316. Baumgarten, M., Margolis, D. J., Selekof, J. L., Moye, N., Jones, P. S., and Shardell, M. (2009). Validity of pressure ulcer diagnosis using digital photography. Wound repair and regeneration, 17(2), 287-290. Bellomo, R., Bailey, A.M., Beale, R., Clancy., G, Danesh V, Hvarfner A, Jimenez E, Konrad D, Lecardo, M., Pattee, K., Ritchie, J., Sherman, K., Tangkau P, (2012) A controlled trial of electronic automated advisory vital signs monitoring in general hospital wards. Critical Care Medicine, 40(8), 2349-2361. Berdot, S., Gillaizeau, F., Caruba, T., Prognon, P., Durieux, P., and Sabatier, B. (2013) Drug administration errors in hospital inpatients: a systematic review. PLoS One, 8(6), e68856. Black, A.D., Car, J., Pagliari, C., Anandan, C., Cresswell, K., Bokun,T., McKinstry, B., Procter, R.,Majeed, A.,and Sheikh, A. (2011) The Impact of eHealth on the Quality and Safety of Health Care: A Systematic Overview. PLoS Med 8(1): e1000387. doi:10.1371/journal.pmed.1000387 Blount, M., Ebling, M.R., Eklund, J.M., James, A.G., McGregor, C., Percival, N., Smith, K.P. and Sow, D. (2010) Real-time analysis for intensive care: development and deployment of the artemis analytic system. IEEE Engineering in Medicine and Biology Magazine, 29(2), 110-8. Bradshaw, L. M., Gergar, M. E., and Holko, G. A. (2011). Collaboration in wound photography competency development: a unique approach. Advances in skin and wound care, 24(2), 85-92.
78
Braun, R., Catalani, C., Wimbush, J., Israelski, D. (2013). Community Health Workers and Mobile Technology: A Systematic Review of the Literature. PLoS ONE 8(6): e65772. doi:10.1371/journal.pone.0065772: Brown, D.S., Aydin, C.E., Donaldson N. (2008), Quartile dashboards: translating large data sets into performance improvement priorities. Journal for Healthcare Quality,30 (6):18-30.
Brown, D. S., Donaldson, N., Burnes Bolton, L., and Aydin, C. E. (2010). Nursing‐Sensitive Benchmarks for Hospitals to Gauge High‐Reliability Performance. Journal for Healthcare Quality, 32(6), 9-17. Brown, H., Terrence, J., Vasquez, P., Bates, D.W. and Zimlichman., E. (2014) Continuous monitoring in an inpatient medical-surgical unit: A controlled clinical trial. American Journal of Medicine, 127(3), 226-232. Buckley, K.M., Adelson, L.K., and Agazio, J.G. (2009), Reducing the risks of wound consultation: adding digital images to verbal reports, Journal of Wound, Ostomy, and Continence Nursing. 36(2):163-70. Carlisle, D. (2012). Remote access to closer care, Nursing Standard. 26(29): 20-22. Catwell, L., and Sheikh, A. (2009), Evaluating eHealth Interventions: The Need for Continuous Systemic Evaluation, 6 (8) PLoS Medicine:1- 6. Chandraharan, E. (2010). Clinical dashboards: do they actually work in practice? Three-year experience with the Maternity Dashboard. Clinical Risk, 16(5), 176-182. Chanussot‐Deprez, C., and Contreras‐Ruiz, J. (2008). Telemedicine in wound care. International Wound Journal, 5(5), 651-654. Chapuis, C., Roustit, M., Bal, G., Schwebel, C., Pansu, P., David-Tchouda, S., Foroni, L., Calop, J., Timsit, J.F., Allenet, B., Bosson, J.L. and Bedouch, P. (2010) Automated drug dispensing system reduces medication errors in an intensive care setting. Critical Care Medicine, 38(12), 2275-81. Chen, L., Chuang, L. M., Chang, C. H., Wang, C. S., Wang, I. C., Chung, Y., ... and Lai, F. (2013). Evaluating self-management behaviors of diabetic patients in a telehealthcare program: longitudinal study over 18 months. Journal of medical Internet research, 15(12): e266 Chib A, Lwin MO, Ang J, Lin H, Santoso F (2008) Midwives and mobiles: using ICTs to improve healthcare in Aceh Besar, Indonesia. Asian Journal of Communication,18:4, 348–364. Chua, S., Chua, H. and Omar, A. (2010) Drug administration errors in paediatric wards: a direct observation approach. European Journal of Pediatrics, 169(5), 603-611. Clark, K. W., Whiting, E, Rowland, J, Thompson, L E., Missenden, I, and Schellein, G (2013). Breaking the mould without breaking the system: the development and pilot of a clinical dashboard at The Prince Charles Hospital. Australian Health Review 37, 304–308. Clemensen, J., Larsen, S. B., Kirkevold, M., and Ejskjaer., N. (2008). Treatment of diabetic foot ulcers in the home: video consultations as an alternative to outpatient hospital care. International journal of telemedicine and applications, vol. 2008, Article ID 132890.
79
Cochran, G.L., Jones, K.J., Brockman, J., Skinner, A. and Hicks, R.W. (2007) Errors prevented by and associated with bar-code medication administration systems. Joint Commission journal on quality and patient safety / Joint Commission Resources, 33(5), 293-301, 245. Coleman, J.J., Hodson, J., Brooks, H.L. and Rosser, D. (2013) Missed medication doses in hospitalised patients: A descriptive account of quality improvement measures and time series analysis. International Journal for Quality in Health Care, 25(5), 564-572. Collins, K. J., and Draycott, T. (2015). Measuring quality of maternity care. Best Practice and Research Clinical Obstetrics and Gynaecology. Published Online: March 31, 2015, Publication stage: In Press Corrected Proof Cookson. R., Ferguson, B., Fleetcroft, R., Goddard, M., Goldblatt, P., Laudicella, M., Raine, R. (2013) HSand DR - 11/2004/39: Developing indicators of change in NHS equity performance, http://www.nets.nihr.ac.uk/projects/hsdr/11200439 Cottney, A., and Innes, J. (2015) Medication-administration errors in an urban mental health hospital: A direct observation study. International Journal of Mental Health Nursing, 24(1), 65-74. Cottney, A. (2014) Improving the safety and efficiency of nurse medication rounds through the introduction of an automated dispensing cabinet. BMJ Quality Improvement Reports, 3(1). u204237-w1843. Cottrell, S., and Davidson, V. (2013a) National audit of bedside transfusion practice. Nursing standard (Royal College of Nursing (Great Britain): 27(43), 41-48. Cottrell, S., Watson, D., Eyre, T.A., Brunskill, S.J., Dorée, C. and Murphy, M.F. (2013b) Interventions to Reduce Wrong Blood in Tube Errors in Transfusion: A Systematic Review. Transfusion Medicine Reviews, 27(4), 197-205. Cruickshank, (2010) “Healthcare without Walls: A Framework for Delivering Telehealth at Scale, from 2020 Health, November 2010 url: http://www.2020health.org/dms/2020health/downloads/reports/2020telehealth LOW.pdf Daley,K., Richardson,,J., James,I., Chambers, A., and Corbett, D. (2013) Clinical dashboard: use in older adult mental health wards, The Psychiatrist , 37, 85-88 Demaerschalk, B. M., Bobrow, B. J., Raman, R., Kiernan, T. E. J., Aguilar, M. I., Ingall, T. J., ... and Meyer, B. C. (2010). Stroke Team Remote Evaluation Using a Digital Observation Camera in Arizona The Initial Mayo Clinic Experience Trial. Stroke, 41(6), 1251-1258. Department of Health (1998) Information for Health. A National Strategy for the Modern NHS 1998-2005. DH Publications, London. Department of Health (2007) Information Security Management: NHS Code of practice, DH Publications, London Department of Health (2008a) High Quality Care for All: NHS Next Stage Review, DH Publications, London
80
Department of Health (2008b) Framing the Nursing and Midwifery Contribution: Driving up the Quality of Care, DH Publications, London Department of Health (2013), National Mobile Health Worker Project: Final Report, DH Publications, London Dewsbury, G. (2014),The use of patient management systems in the community, British Journal of Community Nursing 19:11, 548-550 DeYoung, J., Venderkooi, M. and Barletta, J. (2009) Effect of bar-code-assisted medication administration on medication error rates in an adult medical intensive care unit. . American Journal of Health-System Pharmacy, 66, 1110-1115. Dharmasaroja, P. A., Muengtaweepongsa, S., and Kommarkg, U. (2010). Implementation of telemedicine and stroke network in thrombolytic administration: comparison between walk-in and referred patients. Neurocritical care, 13(1), 62-66. Dhiliwal, S. R., and Salins, N. (2015). Smartphone applications in palliative homecare. Indian journal of palliative care, 21(1), 88-91 Dolan, J.G., and Veazie, P.J. (2015), Balance Sheets Versus Decision Dashboards to Support
Patient Treatment Choices: A Comparative Analysis. The Patient-Patient-Centered Outcomes Research, 1-7. Dolan, J.G., Veazie, P.J., Russ, A.J., (2013), BMC Medical Informatics and Decision Making 13 (51), Development and initial evaluation of a treatment decision dashboard. BMC medical informatics and decision making, 13(1), 51. Doran, D., Mildon, B., and Clarke, S. (2011) Toward A National Report Card in Nursing: A Knowledge Synthesis. Prepared for the Planning Committee for a Think Tank entitled “Toward a National Report Card for Nursing” Nursing Health Services Research Unit, University of Toronto.
Doran, D., Bloomberg, L. S., Reid-Haughian, C., and Cafazzo, J. (2012). A Formative and Summative Evaluation of an Electronic Health Record in Community Nursing. In NI 2012: Proceedings of the 11th International Congress on Nursing Informatics (Vol. 2012). American Medical Informatics Association. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3799176/ Dowding, D., Randell, R., Gardner, P., Fitzpatrick, G., Dykes, P., Favela, J.,and Currie, L. (2015). Dashboards for improving patient care: Review of the literature. International Journal of medical informatics, 84(2), pp 87-100. Drake, R. (2014a) Five dilemmas associated with e-rostering. Nursing Times, 110(20), 14-16. Drake, R.G. (2014b) The nurse rostering problem: from operational research to organisational reality? Journal of Advanced Nursing, 70(4), 800-810. Drake, R.G. (2014c), The ‘Robust’ roster: exploring the nurse rostering process. Journal of Advanced Nursing, 70(9), 2095-2106.
81
Duffin, C. (2013) Digital pen means nurses can spend more time with patients, Mental Health Practice Nov 17 (3): 9-9. Dufrene, C. (2009) Photography as an Adjunct in Pressure Ulcer Documentation, Critical Care Nursing Quarterly. Realities and Myths in Critical Care Practice. 32(2):77-80. Dwibedi, N., Sansgiry, S.S., Frost, C.P., Dasgupta, A., Jacob, S.M., Tipton, J.A. and Shippy, A.A. (2011) Effect of bar-code-assisted medication administration on nurses’ activities in an intensive care unit: A time–motion study. American Journal of Health-System Pharmacy, 68(11), 1026-1031. Effken, J. A., Brewer, B. B., Logue, M. D., Gephart, S. M., and Verran, J. A. (2011). Using Cognitive Analysis to fit decision support tools to nurse managers’ work flow, International Journal of Medical Informatics, 80, (10), 698 – 707. Engel, H., Huang, J. J., Tsao, C. K., Lin, C. Y., Chou, P. Y., Brey, E. M., and Cheng, M. H. (2011). Remote real‐time monitoring of free flaps via smartphone photography and 3G wireless internet: A prospective study evidencing diagnostic accuracy. Microsurgery, 31(8), 589-595. Ernst, E. J., Speck, P. M., & Fitzpatrick, J. J. (2012). The element of naturalness when evaluating image quality of digital photo documentation after sexual assault. Advanced Emergency Nursing Journal, 34(3), 250-258. Esterle, L., and Mathieu-Fritz, A. (2013). Teleconsultation in geriatrics: Impact on professional practice. International Journal of medical informatics, 82(8), 684-695. Fatehi, F., Armfield, N. R., Dimitrijevic, M., and Gray, L. C. (2015a). Technical aspects of clinical videoconferencing: a large scale review of the literature. Journal of telemedicine and telecare, 21(3), 160-166. Fatehi F., Gray, L.C., Russell, A.W., Paul, S. (2015b), Validity and reliability of video teleconsultation for the management of diabetes: A randomized controlled trial. Diabetes Technology & Therapeutics, (A120-A120). Felkey, B. G., and Fox, B. I. (2014). Health System Dashboard: It’s All Coming Together. Hospital pharmacy, 49(5), 485-486. FitzHenry, F., Murff, H.J., Matheny, M.E., Gentry, N., Fielstein, E.M., Brown, S.H., Reeves, R.M., Aronsky, D., Elkin, P.L., Messina, V.P. and Speroff, T. (2013) Exploring the Frontier of Electronic Health Record Surveillance: The Case of Post-Operative Complications. Medical care, 51(6), 509-516. Flannigan, C., Toland, U., and Hogan, M. (2010). Dashboards in neonatology. Clinical Audit, 2, 79-81. Florczak, B., Scheurich, A., Croghan, J., Sheridan Jr, P., Kurtz, D., McGill, W., and McClain, B. (2012). An observational study to assess an electronic point-of-care wound documentation and reporting system regarding user satisfaction and potential for improved care. Ostomy/wound management, 58(3), 46-51.
82
Florez-Arango JF, Iyengar MS, Dunn K, Zhang J (2011) Performance factors of mobile rich media job aids for community health workers. Journal of the American Medical Informatics Association,18 (2): 131–7. Foley, B. and Cox, A. (2013). Work organisation and innovation - Case study: Nottingham University Hospitals NHS Trust, UK. Dublin: European Foundation for the Improvement of Living and Working Conditions. http://digitalcommons.ilr.cornell.edu/intl/248/?utm_source=digitalcommons.ilr.cornell.edu%2Fintl%2F248andutm_medium=PDFandutm_campaign=PDFCoverPages Foltynski, P., Wojcicki, J.M., Ladyzynski, P., Migalska-Musial, K., Rosinski, G., Krzymien J., Waldemar Karnafel, W.(2010), Monitoring of Diabetic Foot Syndrome Treatment: Some New Perspectives, 35, (2):176–182, Foulkes, M. (2011) Nursing metrics: measuring quality in patient care. Nursing Standard, 25, 42, 40-45. Francis, R. (2013) Report of the Mid Staffordshire NHS Foundation Trust public inquiry. HC 898. London. Franklin, B.D., O'Grady, K., Donyai, P., Jacklin, A. and Barber, N. (2007) The impact of a closed-loop electronic prescribing and administration system on prescribing errors, administration errors and staff time: A before-and-after study. Quality and Safety in Health Care, 16(4), 279-284. Frazier, J. A., and Williams, B. (2012). Successful implementation and evolution of unit-based nursing dashboards. Nurse Leader, 10(4), 44-46. General Medical Council (GMC 2011) New guidance on Making and using visual and audio recordings of patient, http://www.gmc-uk.org/guidance/9457.asp Glass, C.A. and Knight, R.A. (2010),The nurse rostering problem: A critical appraisal of the problem structure. European Journal of Operational Research, 202(2), 379-389. Goodell, P.P., Uhl, L., Mohammed, M. and Powers, A.A. (2010) Risk of Hemolytic Transfusion Reactions Following Emergency-Release RBC Transfusion. American Journal of Clinical Pathology, 134(2), 202-206. Greenhalgh, T. Stramer, K. Bratan, T, Byrne, E. Mohammad, Y, Russell, J, (2008),Introduction of shared electronic records: multi-site case study using diffusion of innovation theory.BMJ (Clinical research ed.) 337 (2008): a1786 Greenhalgh,T., Stramer, K., Bratan, T., Byrne, E., Russell, J., and Potts, H. (2010), Adoption and non-adoption of a shared electronic summary record in England: a mixed-method case study, BMJ (Clinical research ed.) 340 (2010), c3111. Griffiths P, Jones S, Maben J, Murrells T. (2008) State of the art metrics for nursing: A rapid appraisal. National Nursing Research Unit, King’s College London Guha, S. Hoo,W. P. and Bottomley, C.(2013) Introducing an acute gynaecology dashboard as a new clinical governance tool, Clinical Governance: An International Journal, 18:3 , 228-237
83
Hardmeier, A., Tsourounis, C., Moore, M., Abbott, W.E. and Guglielmo, B.J. (2014) Pediatric Medication Administration Errors and Workflow Following Implementation of a Bar Code Medication Administration System. Journal for Healthcare Quality, 36(4), 54-63. Harting, M. T; DeWees, J. M.; Vela, K. M; Khirallah, R. T. (2015), Medical photography: current technology, evolving issues and legal perspectives. International Journal of Clinical Practice. 69(4):401-409, Hassink, J.J.M., Jansen, M.M.P.M. and Helmons, P.J. (2012) Effects of bar code-assisted medication administration (BCMA) on frequency, type and severity of medication administration errors: a review of the literature. European Journal of Hospital Pharmacy: Science and Practice, 19(5), 489-494. Häyrinen, K, Saranto K, Nykanen P., (2008) Definition, structure, content, use and impacts of EHRs: a review of the research literature. International Journal of Medical Informatics; 77: 291-304. Health and Social Care Information Center HSCIC (2013) Urgent Care Clinical Dashboards ClearWaterWatcher, http://systems.hscic.gov.uk/qipp/library/cwwatcherfs.pdf Hedgecock, L and Lewis, J (2012) Dashboard data, The Commissioning Review http://www.thecommissioningreview.com/article/dashboard-data Helmons, P., Wargel, L. and Daniels, C. (2009) Effect of bar-code-assisted medication administration on medication errors and accuracy in multiple patient care areas. American Journal of Health-System Pharmacy, 66, 1202-1210. Hockley, T. and Boyle, S. (2014) NHS staffing: not just a number. London School of Economics, London. Holmstock, L., Verellen, E., Franklin, B.D. and McLeod, M. (2014) Developing a tool for evaluating the impact of an electronic prescribing and administration system on medication safety in an English NHS hospital trust. International Journal of Pharmacy Practice, 22, 14-15. Hop, M. J., Moues, C. M., Bogomolova, K., Nieuwenhuis, M. K., Oen, I. M. M. H., Middelkoop, E.and Baar, M. V. (2014). Photographic assessment of burn size and depth: reliability and validity. Journal of wound care, 23(3), 144-152. http://informahealthcare.com/doi/abs/10.1586/erd.10.35 Hug, C.W., Clifford, G.D. and Reisner, A.T. (2011) Clinician blood pressure documentation of stable intensive care patients: an intelligent archiving agent has a higher association with future hypotension. Critical Care Medicine, 39(5), 1006-14. Hunter, J., Freer, Y., Gatt, A., Reiter, E., Sripada, S. and Sykes, C. (2012) Automatic generation of natural language nursing shift summaries in neonatal intensive care: BT-Nurse. Artificial Intelligence in Medicine, 56(3), 157-72. Jamal, A., McKenzie, K., and Clark, M. (2009),The impact of health information technology on the quality of medical and health care: a systematic review. Health Information Management Journal, 38 (3), 26-37.
84
Janardhanan, L., Leow, Y. H., Chio, M. T., Kim, Y., and Soh, C. B. (2008). Experience with the implementation of a web-based teledermatology system in a nursing home in Singapore. Journal of telemedicine and telecare, 14(8), 404-409. Jeffs, L., Lo, J., Beswick, S., Chuun, A., Lai, Y., Campbell, H., and Ferris, E. (2014). Enablers and Barriers to Implementing Unit-Specific Nursing Performance Dashboards. Journal of nursing care quality, 29(3), 200-203. Jesada, E., Warren, J., Goodman, D., Iliuta, R., Thurkauf, G., McLaughlin, M., Johnson, J., and Strassner, L. (2013), 'Staging and Defining Characteristics of Pressure Ulcers Using Photographs by Staff Nurses in Acute Care Settings', Journal of Wound, Ostomy and Continence Nursing, 40,(2),150-156. Jesada, E, Warren, J, Goodman, D, Iliuta, R, Thurkauf, G, McLaughlin, M, Johnson, J, and Strassner, L (2013), 'Staging and Defining Characteristics of Pressure Ulcers Using Photographs by Staff Nurses in Acute Care Settings', Journal of Wound, Ostomy and Continence Nursing, vol. 40, no. 2, pp. 150-156. Johansson, A. M., Lindberg, I., and Söderberg, S. (2014). Patients’ experiences with specialist care via video consultation in primary healthcare in rural areas. International journal of telemedicine and applications, 2014, 9. 143824. Jones, B.G. (2013) Developing a vital sign alert system. American Journal of Nursing, 113(8), 36-44. Jones, S., Mullally, M., Ingleby, S., Buist, M., Bailey, M. and Eddleston, J.M. (2011) Bedside electronic capture of clinical observations and automated clinical alerts to improve compliance with an Early Warning Score protocol. Critical care and resuscitation: Journal of the Australasian Academy of Critical Care Medicine, 13(2), 83-88. Kaliyadan, F., Manoj, J., Venkitakrishnan, S., and Dharmaratnam, A. D. (2008). Basic digital photography in dermatology. Indian Journal of Dermatology, Venereology, and Leprology, 74(5), 532. Keen, J., Long, A., McGinnis, E., Randell, R., Willis, S., Ginn. C., and Whittle, J. (2014) HSand DR - 13/07/68: Information Systems: Monitoring and Managing from Ward to Board http://www.nets.nihr.ac.uk/projects/hsdr/130768 Keers, R.N., Williams, S.D., Cooke, J. and Ashcroft, D.M. (2013) Prevalence and Nature of Medication Administration Errors in Health Care Settings: A Systematic Review of Direct Observational Evidence. Annals of Pharmacotherapy, 47(2), 237-256. Keogh, B., (2013), Review into the quality of treatment and care provided by 14 hospital Trusts in England: overview report. NHS England, London. Kidd R. (2011), Benefits of mobile working for community nurse prescribers. Nursing Standard, 25(42):56-60.
85
King, B. (2014) Influencing dressing choice and supporting wound management using remote ‘tele-wound care’ British Journal of Community Nursing. Suppl: S24-31. Kochendorfer, K.M., Morris, L.E., Kruse, R.L., Ge, B. and Mehr, D.R. (2010) Attending and resident physician perceptions of an EMR-generated rounding report for adult inpatient services. Family Medicine, 42(5), 343-349. Kohn L, C.J., Donaldson M (2000) To Err Is Human: Building a Safer Health System, National Academy Press, Washington D.C. Koivunen, M., Niemi, A., and Hupli, M. (2015). The use of electronic devices for communication with colleagues and other healthcare professionals–nursing professionals’ perspectives. Journal of Advanced Nursing. 71 (3) 620–631, Kollef, M.H., Chen, Y., Heard, K., LaRossa, G.N., Lu, C., Martin, N.R., Martin, N., Micek, S.T. and Bailey, T. (2014) A randomized trial of real-time automated clinical deterioration alerts sent to a rapid response team. Journal of Hospital Medicine (Online), 9(7), 424-9. Koppel, R., Wetterneck, T., Telles, J.L. and Karsh, B.T. (2008) Workarounds to Barcode Medication Administration Systems: Their Occurrences, Causes, and Threats to Patient Safety. Journal of the American Medical Informatics Association, 15(4), 408-423. Lakdawala, N; Bercovitch, L; Grant-Kels, J (2013), A picture is worth a thousand words: Ethical dilemmas presented by storing digital photographs in electronic health records, Journal of the American Academy of Dermatology. 69(3):473-475 Latifi, R (2009), Initial experiences and outcomes of telepresence in the management of trauma and emergency surgical patients. The American Journal of Surgery,198 (6):905 -10 Lau, F., Kuziemsky, C., Price, M., and Gardner, J. (2010). A review on systematic reviews of health information system studies. Journal of the American Medical Informatics Association, 17(6), 637-645. Lawler, E.K., Hedge, A., and Pavlovic-Veselinovic, S. (2011) Cognitive ergonomics, socio-technical systems, and the impact of healthcare information technologies. International Journal of Industrial Ergonomics, 41(4), 336-344. Leape, L.L., Bates, D.W., Cullen, D.J. and et al. (1995) Systems analysis of adverse drug events. JAMA, 274(1), 35-43. Lee S, Chib A, Kim JN (2011) Midwives’ cell phone use and health knowledge in rural communities. Journal of Health Commun 16(9): 1006–23. Leung, A.A., Denham, C.R., Gandhi, T.K., Bane, A., Churchill, W.W., Bates, D.W. and Poon, E.G. (2014) A Safe Practice Standard for Barcode Technology. Journal of Patient Safety, Publish Ahead of Print. Lilford RJ, Foster J, Pringle M (2009) Evaluating eHealth: How to Make Evaluation More Methodologically Robust. PLoS Med 6(11): e1000186.
86
Maben, J., Morrow, E., Ball, J., Robert, G., and Griffiths, P. High Quality Care, (2012), Metrics for Nursing. National Nursing Research Unit, King’s College London, MacBride, S; Wells, M; Hornsby, C; Sharp, L; Finnila, K; Downie, L (2008) A Case Study to Evaluate a New Soft Silicone Dressing, Mepilex Lite, for Patients With Radiation Skin Reactions, Cancer Nursing. 31(1):E8-E14, MacLure K, Stewart D, Strath A. (2014) A systematic review of medical and non-medical practitioners’ views of the impact of ehealth on shared care, European Journal of Hospital Pharmacy 21: 54–62. Manias, E., Williams, A. and Liew, D. (2012) Interventions to reduce medication errors in adult intensive care: a systematic review. British journal of clinical pharmacology, 74(3), 411-423. Marini, S, Hasman, A, (2009) Impact of BCMA on medication errors and patient safety: a summary. Studies in health technology and informatics, 146: 439-444 Mayrovitz, H. N., and Lisa B. SoontupeL. B (2009). "Wound areas by computerized planimetry of digital images: accuracy and reliability." Advances in skin and wound care 22.5: 222-229. McCaughey, D., Erwin, C. O., and DelliFraine, J. L. (2015). Improving Capacity Management in the Emergency Department: A Review of the Literature, 2000-2012. Journal of Healthcare Management, 60(1), 93-115. McGregor, C., Catley, C., James, A. and Padbury, J. (2011) Next generation neonatal health informatics with Artemis. Studies in health technology and informatics, 169, 115-119. McHugh, M., VanDyke, K., McClelland, M., and Moss, D. (2011). Improving patient flow and reducing emergency department crowding: A guide for hospitals. http://www.ahrq.gov/research/findings/final-reports/ptflow/ptflowguide.pdf McLaughlin N, Afsar-Manesh N, Ragland V, Buxey F, Martin NA. (2014) Tracking and sustaining improvement initiatives: leveraging quality dashboards to lead change in a neurosurgical department, Neurosurgery;74 (3):235-43; Medical Research Council (2008) Developing and evaluating complex interventions: new guidance. Available:http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000186 Melville, J; Lukefahr, J L., Cornell, J; Kellogg, N, D. Lancaster, J L. (2013), The Effect of Image Quality on the Assessment of Child Abuse Photographs, Pediatric Emergency Care. 29(5):607-611 Miller, C., Karimi, L., Donohue, L., and Kapp, S. (2012). Interrater and intrarater reliability of silhouette wound imaging device. Advances in skin and wound care, 25(11), 513-518. Miller, E. A. (2011). The continuing need to investigate the nature and content of teleconsultation communication using interaction analysis techniques. Journal of Telemedicine and Telecare, 17(2), 55-64.
87
Miller, K., Akers, C., Magrin, G., Whitehead, S. and Davis, A.K. (2013) Piloting the use of 2D barcode and patient safety-software in an Australian tertiary hospital setting. Vox Sanguinis, 105(2), 159-166. Mohammed, M., Hayton, R., Clements, G., Smith, G. and Prytherch, D. (2009) Improving accuracy and efficiency of early warning scores in acute care. British Journal of Nursing, 18(1), 18-24. Morriss Jr, F.H., Abramowitz, P.W., Nelson, S.P., Milavetz, G., Michael, S.L., Gordon, S.N., Pendergast, J.F. and Cook, E.F. (2009) Effectiveness of a Barcode Medication Administration System in Reducing Preventable Adverse Drug Events in a Neonatal Intensive Care Unit: A Prospective Cohort Study. The Journal of Pediatrics, 154(3), 363-368.e1. Morrow, E., Robert, G., Maben, J. and Griffiths, P. (2012) Implementing largescale quality improvement – lessons from The Productive Ward: Releasing time to care TM. International Journal of Health Care Quality Assurance, 25(4) 237-253. Murff, H.J., FitzHenry, F., Matheny, M.E. and et al.,. (2011) Automated identification of postoperative complications within an electronic medical record using natural language processing. JAMA, 306(8), 848-855. Murphy, M.F. (2012a) End-to-end electronic transfusion management in hospital practice. Vox Sanguinis, 103, 38. Murphy (2012b) Transfusion errors: How they occur and novel systems to improve the safety of blood administration. Transfusion Alternatives in Transfusion Medicine, 12(2):1-9 Murphy, M, Fraser, E, Miles, D, Noel, S, Staves, J, Cripps, B, Kay, J, (2012c) How do we monitor hospital transfusion practice using an end-to-end electronic transfusion management system? Transfusion, 52(12): 2505-2512 Murphy, M, Stanworth, S, Yazer, M, (2011) Transfusion practice and safety: Current status and possibilities for improvement. Vox Sanguinis, 100(1): 46-59 Myers, M.G. and Godwin, M. (2012) Automated office blood pressure. Canadian Journal of Cardiology, 28(3), 341-6. Myers, M.G. (2014) Eliminating the human factor in office blood pressure measurement. Journal of Clinical Hypertension, 16(2), 83-6. Nagase, T., Sanada, H., Takehara, K., Oe, M., Iizaka, S., Ohashi, Y., and Nakagami, G. (2011). Variations of plantar thermographic patterns in normal controls and non-ulcer diabetic patients: novel classification using angiosome concept. Journal of plastic, reconstructive and aesthetic surgery, 64(7), 860-866. National Audit Office / Department of Health (2006) Good practice in managing the use of temporary nursing staff. London (online resources at http://www.nao.org.uk/wp-content/uploads/2006/07/05061176_Good_Practice.pdf last accessed 16/06/2015) National Patient Safety Agency (2007) Recognising and responding appropriately to early signs of deterioration in hospitalised patients. Report. Published at http://www.npsa.nhs.uk/corporate/news/deterioration-in-hospitalised-patients/?locale=en
88
Navarro, D A; Singer, P; Leibovitz, E; Krause, I; Boaz, M, (2014), Inter- and intra-rater reliability of digitally captured images of plate waste, Nutrition and Dietetics. 71(4):284-288. Nelson, M.R., Quinn, S., Bowers-Ingram, L., Nelson, J.M. and Winzenberg, T.M. (2009) Cluster-randomized controlled trial of oscillometric vs. manual sphygmomanometer for blood pressure management in primary care (CRAB). American Journal of Hypertension, 22(6), 598-603. Nelson, S, Lu, C, Leng, J, Cannon, G, He, T, Zeng, Q, Halwani, A, Sauer, B (2015) The use of natural language processing of infusion notes to identify outpatient infusions, Pharmacoepidemiology and Drug Safety. 24(1): 86-92
NHS Connecting for Health (2012), Pioneering Dashboards are leading innovation, http://webarchive.nationalarchives.gov.uk/20130502102046/http://www.connectingforhealth.nhs.uk/engagement/clinical/publications/cc17/dashboards NHS Employers (2007) Electronic rostering: helping to improve workforce productivity A guide to implementing electronic rostering in your workplace. London (online report avauilable at http://www.nhsemployers.org/~/media/Employers/Publications/E-Rostering%20guidance.ashx last accessed 16/06/2015) NHS England (2014) The NHS Five Year Forward View, NHS England, London NHS England 2015, Specialised services quality dashboards http://www.england.nhs.uk/commissioning/spec-services/npc-crg/spec-dashboards/ NICE (2014) NICE Pressure ulcers: prevention and management of pressure ulcers [CG179] Published date: April 2014 http://www.nice.org.uk/guidance/cg179 NICE (2007), Acutely ill patients in hospital: Recognition of and response to acute illness in adults in hospital [CG50] published date: July 2007 http://www.nice.org.uk/guidance/CG50 Nijp, H.H., Beckers, D.G., Geurts, S.A., Tucker, P. and Kompier, M.A. (2012) Systematic review on the association between employee worktime control and work―non-work balance, health and well-being, and job-related outcomes. Scandinavian journal of work, environment and health, 38(4), 299-313. Nursing and Midwifery Council NMC (2015), The Code, NMC, London Palma, J.P., Sharek, P.J. and Longhurst, C.A. (2011) Impact of electronic medical record integration of a handoff tool on sign-out in a newborn intensive care unit. Journal of Perinatology, 31(5), 311-317. Paré, G., Sicotte, C., Moreault, M. P., Poba-Nzaou, P., Nahas, G.,and Templier, M. (2011). Mobile computing and the quality of home care nursing practice. Journal of Telemedicine and Telecare, 17(6), 313-317. Park, E.S., Peccoud, M.R., Wicks, K.A., Halldorson, J.B., Carithers, R.L., Jr., Reyes, J.D. and Perkins, J.D. (2010) Use of an automated clinical management system improves outpatient immunosuppressive care following liver transplantation. Journal of the American Medical Informatics Association, 17(4), 396-402.
89
Pedragosa, À., Alvarez-Sabin, J., Molina, C. A., Sanclemente, C., Martín, M. C., Alonso, F., and Ribo, M. (2009). Impact of a telemedicine system on acute stroke care in a community hospital. Journal of Telemedicine and Telecare, 15(5), 260-263. Pockras, P (2013) Reconcling Pyxis overrides after the implementation of EPIC. Online Journal of Nursing Informatics, 17(3) (online at http://ojni.org/issues/?p=2868 last accessed 16/06/2015. Poon, E.G., Keohane, C.A., Yoon, C.S., Ditmore, M., Bane, A., Levtzion-Korach, O., Moniz, T., Rothschild, J.M., Kachalia, A.B., Hayes, J., Churchill, W.W., Lipsitz, S., Whittemore, A.D., Bates, D.W. and Gandhi, T.K. (2010) Effect of Bar-Code Technology on the Safety of Medication Administration. New England Journal of Medicine, 362(18), 1698-1707. Pope, R. M., Malin, R., Buchan, M., Pearson, C., and Wagner, A. (2014). Evidence of sustained benefit: Use of teleconsultation to support people living in residential care settings. International Journal of Integrated Care, 14(8). Inter Digital Health Suppl; URN:NBN:NL:UI:10-1-116540. Pope, R., Muchan, M., Malin, R., Binks, R., and Wagner, A. (2013). The results of 24 hr teleconsultation with people at home and in residential care settings. International Journal of Integrated Care, 13(7). T&T Conf Suppl; URN: URN:NBN:NL:UI:10-1-115664 Prytherch, D.R., Smith Gb Fau - Schmidt, P., Schmidt P Fau - Featherstone, P.I., Featherstone Pi Fau - Stewart, K., Stewart K Fau - Knight, D., Knight D Fau - Higgins, B. and Higgins, B. (2006) Calculating early warning scores--a classroom comparison of pen and paper and hand-held computer methods. Resuscitation, 70(2), 173-8. Raban, M.Z. and Westbrook, J.I. (2014) Are interventions to reduce interruptions and errors during medication administration effective?: a systematic review. BMJ Quality and Safety, 23(5), 414-421. Rack, L.L., Dudjak, L.A. and Wolf, G.A. (2012) Study of Nurse Workarounds in a Hospital Using Bar Code Medication Administration System. Journal of Nursing Care Quality, 27(3), 232-239. Randmaa, M. Mårtensson, G., Swenne, C. L., & Engström, M. (2014) SBAR improves communication and safety climate and decreases incident reports due to communication errors in an anaesthetic clinic: a prospective intervention study. BMJ Open 4.1: e004268. Rantz, M., Hicks,J.,Petroski,L Madsen, G .F., Alexander, R. W., Galambos, G., Conn, C. Scott-Cawiezell, V. Zwygart-Stauffacher,J. and Greenwald, M (2010), Cost, staffing and quality impact of bedside electronic medical record (EMR) in nursing homes. Journal of the American Medical Directors Association, 11, (7) 485-493. RCN (2011) Nursing dashboards – measuring quality. Royal College of Nursing: London RCN (2012) RCN guidance Nursing staff using personal mobile phones for work purposes, RCN, London Redwood S Ngwenya NB, Hodson J Ferner RE .Coleman JJ (2013), Effects of a computerized feedback intervention on safety performance by junior doctors: results from a randomized mixed method study, BMC Medical Informatics and Decision Making 13:63,
90
Rennert, R., Golinko, M., Kaplan, D., Flattau, A., and Brem, H. (2009). Standardization of wound photography using the wound electronic medical record. Advances in skin and wound care, 22(1), 32-38. Rice, P. (2011), Teleconsultation for Healthcare Services: A workbook for implementing new service models, Health Innovation and Education Cluster (HIEC) Yorkshire and the Humber, http://yhhiec.org.uk/wp-content/uploads/2011/10/11092020_tele_consultation_workbk.pdf. Roback, K., (2010) An overview of temperature monitoring devices for early detection of diabetic foot disorders, Expert Review of Medical Devices, 7, (5):711-718. Roback, K., Johansson, M. and Starkhammar, A. (2009). Feasibility of a thermographic method for early detection of foot disorders in diabetes. Diabetes technology and therapeutics, 11(10), 663-667. Rodriguez-Gonzalez, C.G., Herranz-Alonso, A., Martin-Barbero, M.L., Duran-Garcia, E., Durango-Limarquez, M.I., Hernandez-Sampelayo, P. and Sanjurjo-Saez, M. (2012) Prevalence of medication administration errors in two medical units with automated prescription and dispensing. Journal of the American Medical Informatics Association, 19(1), 72-8. Rogers, L.C., Bevilacqua, N.J., Armstrong, D.G and Andros, G., (2010), Digital planimetry results in more accurate wound measurements: a comparison to standard ruler measurements. Journal of Diabetes Science and Technology, 4(4), 799-802. Romero-Brufau, S., Huddleston, J.M., Naessens, J.M., Johnson, M.G., Hickman, J., Morlan, B.W., Jensen, J.B., Caples, S.M., Elmer, J.L., Schmidt, J.A., Morgenthaler, T.I. and Santrach, P.J. (2014) Widely used track and trigger scores: Are they ready for automation in practice? Resuscitation, 85(4), 549-552. Rönnberg, E. and Larsson, T. (2010) Automating the self-scheduling process of nurses in Swedish healthcare: a pilot study. Health Care Management Science, 13(1), 35-53. Roots A, Bhalia, A and Bims J.(2011) “Telemedicine for Stroke: A Systematic Review”, Journal of Neuroscience Nursing, 7:2, 481-489. Russell, M., Hogg, M., Leach, S., Penman, M., and Friel, S. (2014). Developing a general ward nursing dashboard. Nursing Standard, 29(15), 43-49. Ryan, S., O’Riordan, J.M., Tierney, S., Conlon, K.C. and Ridgway, P.F. (2011) Impact of a new electronic handover system in surgery. International Journal of Surgery, 9(3), 217-220. Safdari, R., Ghazisaeedi, M., Mirzaee, M., Farzi, J., Goodini, A. (2014), Development of balanced key performance indicators for emergency departments strategic dashboards following analytic hierarchical process, Health Care Management, 33, (4): 328-34. Salisbury, C., Campbell, J., McKinstry, B., Ziebland, S., Atherton, H., Gibson, A.(2014), The potential of alternatives to face to face consultation in general practice, and the impact on different patient groups publication data August 2017, HSand DR - 13/59/08, http://www.nets.nihr.ac.uk/projects/hsdr/135908
91
Saleh, S., Larsen, J. P., Bergsåker-Aspøy, J., and Grundt, H. (2014). Re-admissions to hospital and patient satisfaction among patients with chronic obstructive pulmonary disease after telemedicine video consultation - a retrospective pilot study. Multidisciplinary Respiratory Medicine, 9(1), 6. Sawyer, A.M., Deal, E.N., Labelle, A.J., Witt, C., Thiel, S.W., Heard, K., Reichley, R.M., Micek, S.T. and Kollef, M.H. (2011) Implementation of a real-time computerized sepsis alert in nonintensive care unit patients. Critical Care Medicine, 39(3), 469-73. Schmidt, P.E., Meredith, P., Prytherch, D.R., Watson, D., Watson, V., Killen, R.M., Greengross, P., Mohammed, M.A. and Smith, G.B. (2014) Impact of introducing an electronic physiological surveillance system on hospital mortality. BMJ Quality and Safety, 24(1), 10-20. Sellen, K.M., Jovanovic, A., Perrier, L. and Chignell, M. (2015) Systematic review of electronic remote blood issue. Vox Sanguinis, 108 (4) published online: 31 Mar 2015 Issue Shaw. A., Jacobs, B., Stockwell D., Futterman C. and Spaeder, M., (2013) Effects of a Unit-Wide dashboard on compliance with patient safety measures, Critical Care Medicine. 41(12 SUPPL. 1):A160-A161. Signal, M., Pretty, C.G., Chase, J.G., Le Compte, A. and Shaw, G.M. (2010) Continuous glucose monitors and the burden of tight glycemic control in critical care: can they cure the time cost? Journal of Diabetes Science and Technology, 4(3), 625-35. Simms RA , Ping H, Yelland A Beringer AJ, Fox R and Draycott TJ (2013) Development of maternity dashboards across a UK health region; current practice, continuing problems, European Journal of Obstetrics, Gynecology, and Reproductive Biology 170(1):119-24, Smeulers, M., Onderwater, A.T., Zwieten, M.C.B. and Vermeulen, H. (2014) Nurses' experiences and perspectives on medication safety practices: an explorative qualitative study. Journal of Nursing Management, 22(3), 276-285. Smith, L.B., Banner, L., Lozano, D., Olney, C.M. and Friedman, B. (2009) Connected care: reducing errors through automated vital signs data upload. CIN: Computers, Informatics, Nursing, 27(5), 318-23. Snyder, S.R., Favoretto, A.M., Derzon, J.H., Christenson, R.H., Kahn, S.E., Shaw, C.S., Baetz, R.A., Mass, D., Fantz, C.R., Raab, S.S., Tanasijevic, M.J. and Liebow, E.B. (2012) Effectiveness of barcoding for reducing patient specimen and laboratory testing identification errors: A Laboratory Medicine Best Practices systematic review and meta-analysis. Clinical Biochemistry, 45(13–14), 988-998. Song, L.P.R.N., Park, B.P. and Oh, K.M.I.P.R.N. (2015) Analysis of the Technology Acceptance Model in Examining Hospital Nurses' Behavioral Intentions Toward the Use of Bar Code Medication Administration. CIN: Computers, Informatics, Nursing, 33(4), 157-165. Sorknæs AD, Madsen H, Hallas J, Jest P and Hansen-Nord M. (2011) Nurse tele-consultations with discharged COPD patients reduce early readmissions – an interventional study. The Clinical Respiratory Journal, 5: 26–34.Denmark
92
Sorknaes, A. D., Bech, M., Madsen, H., Titlestad, I. L., Hounsgaard, L., Hansen-Nord, M.,and Østergaard, B. (2013). The effect of real-time teleconsultations between hospital-based nurses and patients with severe COPD discharged after an exacerbation. Journal of telemedicine and telecare, 19(8), 466-474. Spektor, M., Ramsay, C., Sommer, B., Likourezos, A., and Davidson, S. J. (2008). 144: Electronic Dashboard and a Multidisciplinary Hospital-Wide Team Decrease Patient Throughput Intervals and Reduce Number of Admitted Patients Held in the Emergency Department. Annals of Emergency Medicine, 52(4), S87. Starmer, J., and Giuse, D. (2008). A real-time ventilator management dashboard: toward hardwiring compliance with evidence-based guidelines. In AMIA Annual Symposium Proceedings (Vol. 2008, p. 702). American Medical Informatics Association. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2656080/ Taenzer, A.H., Pyke, J., Herrick, M.D., Dodds, T.M. and McGrath, S.P. (2014) A comparison of oxygen saturation data in inpatients with low oxygen saturation using automated continuous monitoring and intermittent manual data charting. Anesthesia and Analgesia, 118(2), 326-31. Talbot , A (2012),Supporting urgent care using clinical dashboards: transforming data into knowledge, http://www.healthcareconferencesuk.co.uk/news/clinical-dashboards-an-update-on-the-national-urgent-care-dashboard Talmon, J., Ammenwerth, E., Brender, J., de Keizer, N., Nykänen, P., and Rigby, M. (2009). STARE-HI—Statement on reporting of evaluation studies in Health Informatics. International Journal of Medical Informatics, 78(1), 1-9. Tepas, J.J., 3rd, Rimar, J.M., Hsiao, A.L. and Nussbaum, M.S. (2013) Automated analysis of electronic medical record data reflects the pathophysiology of operative complications. Surgery, 154(4), 918-24. Terris, D. D., Woo, C., Jarczok, M. N., and Ho, C. H. (2010). Comparison of in-person and digital photograph assessment of stage III and IV pressure ulcers among veterans with spinal cord injuries. Journal of Rehabilitation, Research and Development, 48(3), 215-224. Thompson, N; Gordey, L; Bowles, H; Parslow, N; Houghton, P (2013) Reliability and Validity of the Revised Photographic Wound Assessment Tool on Digital Images Taken of Various Types of Chronic Wounds. Advances in Skin and Wound Care. 26(8):360-373, Toussi, M., Berthelot, H. and Simon, G. (2014) Study of medication errors which may occur in the process of electronic prescription and dispensing using failure mode and effects analysis (FMEA). Pharmacoepidemiology and Drug Safety, 23, 236-237. University of Kent (2010)‘Unit Costs of Health and Social Care 2010’ published by the Personal Social Services Research Unit, University of Kent Van der Linden, C.M.J., Jansen, P.A.F., Grouls, R.J.E., van Marum, R.J., Verberne, M.A.J.W., Aussems, L.M.A., Egberts, T.C.G. and Korsten, E.H.M. (2013) Systems that prevent unwanted represcription of drugs withdrawn because of adverse drug events: A systematic review. Therapeutic Advances in Drug Safety, 4(2), 73-90.
93
Van Gurp, J., van Selm, M., van Leeuwen, E., and Hasselaar, J. (2013). Transmural palliative care by means of teleconsultation: a window of opportunities and new restrictions. BMC medical ethics, 14(1), 12. Varkey, P; Tan, N; Girotto, R; Tang, W; Liu, Y; Chen, H, (2008)A Picture Speaks a Thousand Words: The Use of Digital Photography and the Internet as a Cost-Effective Tool in Monitoring Free Flaps. Annals of Plastic Surgery. 60(1):45-48, Vergales, B.D., Paget-Brown, A.O., Lee, H., Guin, L.E., Smoot, T.J., Rusin, C.G., Clark, M.T., Delos, J.B., Fairchild, K.D., Lake, D.E., Moorman, R. and Kattwinkel, J. (2014) Accurate automated apnea analysis in preterm infants. American Journal of Perinatology, 31(2), 157-62. Vicente, M. S., Armengol, M. V., Rodríguez, M. A., Crespo, A. M., and Buil, L. O. (2012). Effect of bar-code technology on the safety of cytostatic drugs administration. European Journal of Hospital Pharmacy: Science and Practice, 19 (2), 142. Vowden, K and Vowden, P (2013), A pilot study on the potential of remote support to enhance wound care for nursing-home patients, Journal of Wound Care, 22, (9), 481-488 Vuononvirta, T., Timonen, M., Keinänen-Kiukaanniemi, S., Timonen, O., Ylitalo, K., Kanste, O., and Taanila, A. (2009). The attitudes of multiprofessional teams to telehealth adoption in northern Finland health centers. Journal of Telemedicine and Telecare, 15(6), 290-296. Wachter, R (2015) The digital doctor: Hope, hype and harm at the dawn of medicine’s computer age. MaGraw-Hill Education, New York. Wade, V. A., Karnon, J., Elshaug, A. G., and Hiller, J. E. (2010). A systematic review of economic analyses of telehealth services using real time video communication. BMC health services research, 10(1), 233. Wakefield, B. J., Bylund, C. L., Holman, J. E., Ray, A., Scherubel, M., Kienzle, M. G., and Rosenthal, G. E. (2008). Nurse and patient communication profiles in a home-based telehealth intervention for heart failure management. Patient education and counseling, 71(2), 285-292. Weiner, J.P (2012) Doctor patient communication in the e-healh era, Israel Journal of Health Policy Research, 1:33 Wendelken, M.E., Berg, W.T., Lichtenstein, P., Markowitz, L., Comfort, C., Alvarez, O.M., (2011), Wounds measured from digital photographs using photo digital planimetry software: validation and rater reliability.Wounds. 2011 Sep; 23(9):267-75. Wohlauer, M.V., Rove, K.O., Pshak, T.J., Raeburn, C.D., Moore, E.E., Chenoweth, C., Srivastava, A., Pell, J., Meacham, R.B. and Nehler, M.R. (2012) The Computerized Rounding Report: Implementation of a Model System to Support Transitions of Care1. Journal of Surgical Research, 172(1), 11-17. Wong, T.C., Xu, M. and Chin, K.S. (2014) A two-stage heuristic approach for nurse scheduling problem: A case study in an emergency department. Computers and Operations Research, 51, 99-110.
94
Wood, J. and Finkelstein, J. (2013) Comparison of automated and manual vital sign collection at hospital wards. Studies in Health Technology and Informatics, 190, 48-50. Woodford, M.R., Mackenzie, C.F., DuBose, J., Hu, P., Kufera, J., Hu, E.Z., Dutton, R.P. and Scalea, T.M. (2012) Continuously recorded oxygen saturation and heart rate during prehospital transport outperform initial measurement in prediction of mortality after trauma. The Journal of Trauma and Acute Care Surgery, 72(4), 1006-11. Workman, M. (2008). An experimental assessment of semantic apprehension of graphical linguistics. Computers in Human Behavior, 24(6), 2578-2596. Young, J., Slebodnik, M. and Sands, L. (2010) Bar Code Technology and Medication Administration Error. Journal of Patient Safety, 6(2), 115-120. Zaydfudim, V., Dossett, L. A., Starmer, J. M., Arbogast, P. G., Feurer, I. D., Ray, W. A., May, A,K and Pinson, C. W. (2009). Implementation of a real-time compliance dashboard to help reduce SICU ventilator-associated pneumonia with the ventilator bundle. Archives of Surgery, 144(7), 656-662. Zhang H; Cocosila M; Archer N. (2010),Factors of adoption of mobile information technology by homecare nurses: a technology acceptance model 2 approach. CIN: Computers, Informatics, Nursing. 28(1):49-56.
top related