ED 260 661

AUTHORTITLE

INSTITUTION

DOCUMENT RESUME

HE 018 656

Strickland, Ora L., Ed.; Damrosch, Shirley P., Ed.Research in Nursing Practice; Education, andAdministration: Collaborative, Methodological, andEthical Implications. Proceedings of the ResearchConference of the Southern Council on CollegiateEducation for Nursing (3rd, Baltimore, Maryland,December-2-3, 1983).Maryland Univ., Baltimore. School of Nursing.;Southern Council on Collegiate Education for Nursing,Atlanta, GA.Dec 83161p.Southern Regional Education Board, 1340 SpringStreet, N.S-, Atlanta, GA 30309 ($6.00).Collected Works Conference Proceedings (021) --Reports - Research/Technical (143) -- Reports -Descriptive (141)

EDRS PRICE MF01/PC07 Plus Postage.DESCRIPTORS Administrators; College Faculty; *College Programs;

Continuing Education; *Cooperative Programs;Diseases; *Ethics; Grade Inflation; Health Services;Higher Education; Interdisciplinary Approach; Models;Nursing; *Nursing Education; Plagiarism; ProgramAdministration; *Research Methodology; *ResearchProjeCts

IDENTIFIERS Nursing Research

PUB. DATENOTE'AVAILABLE FROM

PUB TYPE

ABSTRACTCollaborative research in nursing is discussed in

five papers from the 1983 conference of the Southern Council onCollegiate Education for Nursing. lso inzluded are 32 abstracts ofnursing research, focusing on clinical practice, as well as nursingeducation and research models. Paper titles and authors are asf611cws: "Building a Climate for Collaborative Research" (JoanneHorsley); "Establishing and Implementing CollaborativeInvestigations" (Carolyn A. Williams); "Ethical Aspects ofCollaborating with Clients" (Shirley P. Damrosch); "The Ethics ofInter-Discipline Collaboration" (Oscar C. Stine); and "Ownership ofData and Plagiarism in Research" (Carolyn F. Waltz). Abstracts ofclinical research cover topics such as head injuries and risk ofcoronary heart disease. Abstracts on nursing education cover: gradeinflation in an associate degree nursing program, power sharing inuniversity nursing programs, perceptions of top academic nurseadministrators' on their job requirements, collaboration incontinuing education for nursing, and nursing faculty perceptions ofan integrated curriculum. Additional abstracts cover topics such as:models for multivariate effect size estimation in clinical nursingresearch, Q methodology in nursing research, and a collaborativemodel for developing nursing theory. (SW)

BESI COPYAVAIL AML. HE

`RESEARCH IN NURSING PRACTICE,-r-4 EDUCATION, AND ADMINISTRATION:14) COLLABORATIVE, METHODOLOGICAL,4) AND ETHICAL IMPLICATIONSC)

is.'Proceedings of theThird Annual SCCEN Research Conference

(L$. DEPARTMENT OF EDUCATIONNATIONA INSTITUTE OF EDUCATION

EOUCAT Al RESOURCES INFORMATIONCENTER IERICI

his documint has been reproduced asreceived from the person or orginizationoriginating it.

CI Mow changes have been made to improvereproduction quality.

Points of view or opinions stated in this document do not necessarily

repreSent 0441 MEposition or policy.

December 2-3, 1983Baltimore, Maryland

"PERMISSION TOREPRODUCE THISMATERIAL HAS BEEN GRANTED BY

TO THE EDUCATIONALRESOURCESINFORMATION

CENTER (ERIC)."

Southern Council on Collegiate Education for Nursingand

The University of Maryland at BaltimoreSchool of Nursing

2

,

RESEARCH IN NURSING PRACTICE,EDUCATION, AND ADMINISTRATION:COLLABORATIVE, METHODOLOGICAL,AND ETHICAL IMPLICATIONS

Proceedifngs of theThird Annual SCCEN Research Conference

December 2. 3,1983Baltimore, Maryland

Edited by:Ora L. Strickland, RN, PhD, FAAN

Shirley P. Damrosch, PhD.,

Southern Council on Collegiate Education for Nursingand

The University of Maryland at BaltimoreSchool of Nursing

1340 Spring Street, N.W,Atlanta, Georgia 30309

1985$6.00

TABLE. OF CONTENTS,

FOREWORD

BUILD1NG.A CLIMATE FORCOLLABORATIVE RESEARCH-JOahne Horsley

ESTABLISHING. AND IMPLEMENTINGCOLLABORATIVE INVESTIGATIONS- Carolyn A. Williams

SYMPOSIUM ON ETHICAL ISSUESIN COLLABORATIVE RESEARCH

ETHICAL ASPECTS OFCOLLABORATING WITH CLIENTS-Shirley DaMrosch

THE ETHICS OF INTER- DISCIPLINECOLLABORATION-Oscar C. Stine

OWNERSHIP OF DATA ANDPLAGIARISM IN RESEARCH-Carolyn F. Waltz

ABSTRACTS OF STUDIES

MODELS FOR MULTIVARIATEEFFECT SIZE ESTIMATION INCLINICAL NURSING RESEARCH-Ivo L. Abraham, Samuel Schultz II

Pate

ix

1

15

25

-29

35

47

NURSING ADMINISTRATION IMPLICATIONS 51OF THE RELATIONSHIP OF TECHNOLOGYAND STRUCTURE TO QUALITY OF CARE-Judith W. Alexander

iii 4

GRADE INFLATION: A STUDYWITHIN ONE,ASSOCIATEDEGREE NURSING PRO-GRAM-Betty L. Battenfield,Marianne Neighbors

A SURVEY TO DETERMINE WHAT'NUTRITIONAL INFORMATIONMIDDLE-CLASS WOMEN RECEIVEDURING PREGNANCY-Marilyn Bayne, Phyllis L. Neff

STRESS, TRUST, AND MISTRUSTIN NORMOTENSIVE, HYPERTENSIVE ANDHEART DISEASE PERSONS IN ANEXERCISE PROGRAM-Margaret T. Beard, Anne Gudniundsen, andMargit N. 3ohnson

PERCEPTION OF NURSES' EXPERTPOWER BY NURSES AND PHYSICIANS-Gerry Cadenhead

DIRECT OBSERVATION OF PARENT--INFANT RESPONSIVENESS IN NURSINGRESEARCH: THE NURSING CHILDASSESSMENT FEEDING SCALE-Nayna D. Campbell, Elizabeth A. D. Rankin,and Karen Soeken

A COMPARISON OF ADOLESCENTMOTHERS' PERCEPTIONS OF INFANTSIN "ROOMING IN" VERSUS TRADITIONALHOSPITAL SETTINGS-Verna Carson, Marilyn Winklestein

THE EFFECT OF SUPPORT ON LONG-TERM SUCCESS ATZREASTFEEDING-Regina M. Cusson

5 iv

55

61

65

69

71

75

79

Q METHODOLOGY IN NURSING RESEARCH: 85PROMISE AND PROBLEMS-Karen E. Dennis

POWER SHARING IN UNIVERSITIES:THE CASE OF NURSING-Jacqueline Dienemann

CANCER AND NUTRITION: DERIVINGA THEORY OF ADAPTATIONFROM A NURSING EXPERIMENT-Jane Dixon, John Dixon

THE EFFECT OF A DEATH EDUCATIONWORKSHOP ON NURSE ATTITUDESAND BEHAVIORS; PHASE I: DEVELOP-MENT OF AN OBSERVATIONAL TOOLTO DOCUMENT NURSE BEHAVIORWITH TERMINALLY ILL PATIENTS-Judy A. Donovan,Marianne L. Dietrick-Gallagher

89

93

95

'TOP ACADEMIC NURSE ADMINISTRATORS' 101PERCEPTIONS OF ESSENTIAL BEHAVIORSFOR THEIR POSITIONS-Shirley A. George, Carol Deets

A CASE STUDY OF THE IMPLEMENTATION 103-

OF LEGISLATION PROVIDING DIRECT THIRDPARTY REIMBURSEMENT FOR THE SERV-ICES OF CERTIFIED NURSE PRACTITIONERSIN THE STATE OF MARYLAND-Hurdis Griffith

COLLABORATION IN-CONTINUINGEDUCATION FOR NURSING: TRAININGCOMMUNITY HEALTH NURSES TOCARE FOR DISABLED CHILDRENAND THEIR FAMILIES-Carolyn Huard, Connie Petit,Marjorie North, Melissa Van Wey,and Jennie Austin

6

107

EPIDEMIOLOGICAL SURVEY OFCORONARY HEART DISEASE RISKFACTORS IN A_BLACK COMMUNITY- Margie N. Johnson, Margaret T. Beard

VARIATIONS ON AN OLD THEME:AN UPDATE ON HANDWASHING-Elaine Larson, Marcia Killien,and Eric Lation

111

115

DEATH EDUCATION AND DEATH ANXIETY 117- Bonnie Elam Lockard

RELIABILITY ESTIMATES INPUBLISHING NURSING RESEARCH:CURRENT USE AND DISUSE-Mary R. Lynn

WOMEN'S LITERATURE AND THE 123CLIENT BASE OF NURSING-Myrtle P. Matejski

RELATIONSHIP OF POLICY TOPRACTICE. IN THE DELIVERY OFMEDICARE HOME HEALTH SERVICES-Mary O. Muni:linger

THE EFFECTS OF MODIFYINGTHE ENDOTRACHEAL TUBESUCTIONING/MANUAL HYPERVENTILATIONPROCEDURE ON SEVERE CLOSED HEADINJURED PERSONS-L. Claire Parsons, Dorothy Vol!mer,Leanna J. Crosby, and Pam Holley-Wilcox

PARENTAL RESPONSES TO EARLYCHILDHOOD CHRONIC ILLNESS:CONGENITAL HEART DISEASE-Lesley A. Perry

THE POLICY AND POLITICSOF FLORENCE NIGHTINGALE-Barbara Pillar

vi

127

131

137

139

A COMPARISON OF EMOTIONALINDICATORS ON HUMAN-FIGURE'MbSELF-FIGURE DRAWINGS OF LEUKEMICAND HEALTHY CHILDREN-Joanne B. Scungio

HOT FLUIDS FOR THE FEBRILEPATIENT:: WARMING UP TO ANON-TRADITIONAL NURSINGINTERVENTION-Jeanne Sorrell

THE RELATION OF LOCUS OFCONTROL, SEX-ROLE IDENTITY, ANDASSERTIVENESS IN BACCALAUREATENURSING STUDENTS

. -Kathlien R. Stevens

NURSING FACULTY PERCEPTIONS OFAN INTEGRATED CURRICULUM ANDIMPLEMENTATION OF THE CURRICULUM- Corrinne Strandell

PROGRAMMATIC RESEARCH: ACOLLABORATIVE MODEL FORDEVELOPING NURSING THEORY- Patricia Tomlinson, Joyce Semradek

PARENTAL PERCEPTIONS OF INFANTBEHAVIOR AND RHYTHMIC PATTERNSOF THE PARENTS AND THE INFANT-Sharon A. Wilkerson

BLOOD PRESSURE AND HEART RATECHANGES DURING CARDIACCATHETERIZATION- Fran Wimbush, Ellen Sappington,Erika Friedmann, Sue Ann Thomas,and James J. Lynch

141

145

149

153

157

159

163

PAPER REVIEW COMMITTEE 1'67

CONFERENCE PLANNING COMMITTEEvii

171

FOREWORD

At the annual research conference of the SouthernCouncil on Collegiate Education for Nur Sing, held De-cember 2-3, 1983, in Baltimore, Maryland, five papers oncollaborative research offered practical pointers on struc-turing a climate for collaborative research, describedcharacteristics of successful collaborative groups, sug-gested collaboration with clients and other professions aswell as- nurse researchers, and identified ethical prob-lems 16 collaborative research.

Also presented were thirty-two abstracts of nursingresearch, chosen through a refereed process of blindreview by the paper review committee. The majority ofstudies pertain to clinical practice--ranging from spe-cialty topics such as head injuries to general, such as theeffect of hot liquid drinks on febcile patients. Nursingeducation studies range from a study on grade inflationin an associate degree program, to faculty perceptions ofan integrated curriculum. Researchers offer models formultivariate meta-analytic procedures that allow fortranslating principles of holistic health approaches toholistic research paradigms, discuss the use of Q meth-odology, and warn against the neglect of reliabilityestimates in nursing research.

The University of Maryland School of Nursing hostedthe conference. Members of the SCCEN Research Com-mittee,.who assisted in general planning, are: William E.Field, Jr., University of Texas at Austin; Margaret T.Beard, Texas Woman's University; Mary Colette Smith,University of Alabama in Birmingham; and Ora L. Strick-land, Uhiversity of Maryland, Baltimore.

Audrey F. Spector'Executive DirectorSouthern Council on Collegiate

Education for Nursing

#9

BUILDING A CLIMATE FORCOLLABORATIVE RESEARCH

Jo Anne Horsley, RN, PhD, FAANProfessor and Research.Facilitator

The Oregon Health Sciences University, Portland

Building a climate for collaborative research is adifficult topic to address. Difficult because on the one.hand, it is viewed by many as a very desirable approachto clinical research and experience shows it holds muchpromise, while on the other hand, little evidence can befound of its occurrence in nursing nor are there easyanswers as to how we might support its occurrence on apermanent basis.

What is the meaning of the phrase collaborativeresearch? It is important to recognize that two distinctand different processes are delineated in the phrase.The first process, that of collaboration, is used to carryout a second process; research, in order to produce aproduct,*knowledge. It is important to note that, for thepurposes of this paper, the collaboration is to occuramong faculty members and clinicians -- professionalnurses who are usually employed by different organi-zations. And, we must recognize that the facultymembers and clinicians, and their respective employingorganizations, may (most probably do) have both similarand dissimilar reasons for engaging in nursing research.

The Promise of Collaborative Research

Collaborative research in nursing is not new. In fact,it has been an essential component of a number ofnursing research development projects for at least thepast 15 years. In these projects collaborative researchhad the following characteristics: research groups with

1 0

. .

two or more members who represented different organi-zations and/or roles; group members who were interestedin the same general research topic and collectivelyrepresented a more diverse point of view regarding thetopic and how to study it than generally occurs with anindividual investigator; and finally, group members whoworked together throughout the conduct of tip researchfrom the" initial conceptualization of ..the study to thefinal report or publication.

A review of some of the research development proj-ects readil als the benefits of engaging in a col-

, laborati ,research .process. The lar est single projectincorporating a collaborative r process was con-ducted under the auspices of the Western InterstateCommission for Higher Education (WICHE) and is knownas the Regional Program for Nursing. Research Develop-ment. Approximately 350 participants who representedapproximately 250 different organizations worked inover 75 separate collaborative. research groups. Groupsize ranged from two to ten members (Krueger, Itlson do

Wolanin, 1978). In describing the project, Lindeman andKrueger (1977) citeda number of beneficial effects fromthe use of a collaborative research approach. Thebenefits can be classified inytwo categories: benefitspertaining to the research Nd benefits pertaining to theto the investigators.

Benefits pertaining to the research included:

group members who provided multiple viewpointsregarding all aspects of the research proCess;

potential for larger sample sizes resulting fromthe use of multiple settings; Nincreased flexibility in approaches used to in-vestigate different types of research questions, ,

for example, concurrent replication in multiplesettings.

112

Benefits pertaining to the investigators included;

an apprenticeship approach in which noviceslearned from those more experienced, and, wherethe apprentice and master roles switched at dif-ferent points in the research process;

division of labor among group members thatresulted in maximum utilization of group mem-bers' skills;

the potential fOr continuous peer review;

group pressure to complete assigned work onschedule.

In their report of the Nursing Research Developmentin the South project, Mauger and Huggins (1980) cited twoadditional benefits of the collaborative research process:(1) when individual participants and their institutionsJacked some of the, -resources necessary to conductresearch, the missing resources usually became availablethrough other group members and their institutions; and(2) it provided more initial points from which dissem-ination of results could occur. Krone and Loomis (1982),in a report describing the Collaborative Research Pro-gram of the Conduct and Utilization of Research inNursing project, proposed that when clinicians and re-searchers engage in collaborative research the outcomewill be more relevant to practice than when clinicians,are not involved.

Three types of benefits of collaborative researchhave been identified: .research that is scientifically'sound and practice relevant; professional growth andshared labor for the investigators; and increased access totechnical resources. I think we would all agree that '

these benefits represent a substantial contribution to ourresearch efforts.

3 12

It is difficult to determine the extent to whichcollaborative research actually occurs in nursing. Onesource of available data is published reports of cc. Elab-orative research studies. However, research publicationsprovide only indirect evidence of the use of a collab`-orative-research process through the examination of thetitles end employing organizations of the authors of thearticles. These data are subject to erroneous inter-pretation arising from the fact that authors are notnecessarily the only investigators and the pUblishedauthor titles and organizational affiliation may not rep-resent the actual role and employing organization of theauthors when the study was conducted. In spite of theseproblems, this method was used to determine an esti-mate of the frequency of collaborative research duringthe ,past five years. A convenience sample of 36 issuesof Nursing Research Research in Nursing and Health,and the Western Journal of Nursing Research containing297 articles was reviewed. vOnly 7 percent or 22 articleswere 'co-authored by investigators who were faculty inschools of nursing and clinicians from clinical agencies.Four of the 22 co-authored articles reported researchinitially sponsored by the WICHE research developmentgrant. With one exception, articles describing the use'ofthe collaborative research process were reports of fed-erally funded.research development'grants,

On the basis of 'this indirect evidence, it wouldappear that relatively little collaborative research hasoccurred outside the context of large externally fundedreseah development projects. Why is this the casewhen the benefits are so obvious? The first, and.believe the foremost, reasonf,is that the ,research devel-opment projects provided_ an organizational structurespecifically developed to support collaborative researchactivity. These projects assumed many of the- charac-teristics of organizationsthey provided purpose. andgoals, a meant of identifying other investigators withsimilar research interests, resources (travel costs, re-search expertise, pilot project funding), and commu-nication mechanisms (workshops, teleconferencing). The

13 sO

participants became, collaborators because the projectsbrought them together; they did not start as self-identified group members. The project provide_ d thepurpose and mewling for the existence of each group.When the "organizational structures" ceased to exist, sotoo, to a-large extent, did the ongoing operation of theollaborative research groups.

Is an organizational entity of some kind reallynecessary? Collaborative research does occur, but itusually involves researcher-to-researcher collaborationrather than faculty-clinician collaboration. There areseveral reasons ..for this. First, researchers are morelikely to know, or kncw of, one another than are facultymembers and clinicians. In other words, familiaritybreeds collaboration. Second, researchers have existingmechanisms, such as research conferences, that bringthem together from time to time. These, mechatiismsare less readily accessible to clinicians. For thesereasons I submit that we need to replace the organi-zational structure or system functions provided by re-search development grants ii we want to support the useof a collaborative_research process. Further, the newstructure must, in some way, bring faculty members andclinicians together through mechanisms provided by their_employing organizations.

An examination of the forces that serve to facilitateor hamper research collaboration between educationaland clinical organizations should provide some guidanceregarding the nature of the organizational arrangementsthat will be needed. The facilitating forces include:

I. Both types of organizations employ nurses whoare interested in engaging in research activity.

2. Both types of organizations need to engage inresearch activity in order to attain their re-.spective goals.

145

x,4

3. Both types of organizations have a positive ori-entation toward nursing research; it _may bemore pronounced in academic institutions but it isclearly evident in clinical agencies.

4. The research endeavors of each organizationwould be enriched through the use of the humanand technical resources available in the otherorganiiation.

5. Within the profession, there is a generalmovement toward reuniting these organizationsfor the pursuit of common goals. (Reunitingdoes not necessarily mean unification as it ,iscurrently used.)

The forces that impede the occurrence of collab-orative research are more complex than those thatfacilitate it. Four inhibiting forces have been identified :

1. Although both organizations are positively ori-ented toward the use of research as a means ofattaining goals relating to practice and theory,the relative emphasis each places on such goalsvaries greatly. This variance in -emphasis in-fluences both the nature of the research questionto be addressed and the way in which the processof scientific inquiry .occurs. The variance inemphasis occurs in three areas:

Immediate applicability. Clinicians and theirorganizations engage in research activity inthe hope of solving practice problems in animmediate time framethe generation andaccumulation- of knowledge is a long-rangegoal. Academic institutions view the timeframe for these goali in a different manner,with knowledge generation and, accumulationbeing the immediate goal and utilization ofthe knowledge in practice being a long-termgoal:

15 4.-

6

How the relevance of the study is judged.The basis for determining the value of theresearch question to be addressed is alsodifferent for clinician and facultyresearchers. The clinician tends to evaluatethe research question in terms of itsrelevance for practice, while the facultymember tends to evaluate the research ques-tion in terms of its potential for advancingthe state of the knowledge in the area ofinterest.

Relation to the generalizability of researchoutcomes. Clinical systems press for imme-diate answers to local problems, while aca-demic systems focus on the generation ofknowledge that can be generalized for allrelevant situations.

2. To provide a basis for collaboration; both orga-nizations must recognize the value of the other'sview of the goals of research and be willing toprovide some of their resources for use in attain-ing those goals.

3. Inherent in a collaborative process is the need toresolve different points of view and reachcommon understandings and agreement on howto proceed. This increases the time required toconduct a study and acts as an inhibitor to theuse of collaborative research.

4. Faculty members and clinicians have limitedaccess to information regarding one another'sresearch interests and freedom to engage inresearch as a part of their position. This clearlylimits the possibility of identifying appropriateresearch collaborators.

To overcome the factors that inhibit collaborativeresearch, two different levels of intervention become

16

apparent. The first is based on the assumptions that themajor inhibitor is the lack of a means for facultymembers and clinicians to become aware of each other'sresearch interests and that individuals can overcome theother negative forces without overt formal inter-organizational support. If these assumptions are true,some type of information exchange mechanism is neededthat would link schools of nursing and -nursing caresystems within geographical boundaries to allow rela-tively easy access among relevant faculty members andclinicians. The information to be exchanged shouldinclude the types of technical resources. (for example,computer access to subjects) available in each organi-zation and relevant information about the clinicians andfaculty members who are interested in and available toparticipate in collaborative research studies.

If one assumes that individuals cannot overcome theother forces that inhibit collaborative research, and thatformal interorganizational agreements are necessary,then a more elaborate intervention is heeded. At anorganizational level, two factors in addition to individualclinician-faculty linkage become important and must beexplicitly handled -in the formal arrangements.

The first factor is the commitment of each organi-zation to provide the resources necessary for collab-orative research. Of particular importance is theamount and duration of faculty or staff time that will bereleased for collaborative projects. This is particularlyimportant because any specific study may fit the goalsand needs of one organization better than the:other. Theformal commitment of time provides assurance thatstudy will be completed.

The second factor that requires explicit negotiationand agreement is how collaborative research projectswill be selected. This is important because of potentialdisagreement regarding the objectives of proposed col-laborative studies. Interorganizational agreement re-garding the purposes of a given study and the resources

8

committed foi the conduct Of the study provides* thefaculty and clinician investigators with assurance thattheir collaborative work represents the interests of theirrespective organizations and the requirements of theiremployment contract. These agreements could poten-tially take the form of the .student clinical experience.contracts already in use in schools of nursing.

What conditions within schools of, nursing and nursingservice agencies might be associated with research pro-ductivity?

During the 'early Sixties two scientists, Donald Pelzand Frank Andrews (1966) conducted a study to identifyhow two 'types of human factors in the environments ofvarious kinds of research and development organizationswere associated with research productivity: those asso-ciated with the working conditions that prevailed in thesetting, for example, freedom, communication, and di-versity; and those associated with the scientist, forexample, motivation, dedication, age, and creativity.

The subjects in the study were 1,311 scientists andengineers representing 11 different research labora-tories. Included in the sample were 641 industrialscientists, .144 university scientists, and 526 governmentscientists. The scientists represented a wide range offields engaged in research and development, includingthe social, biological, and health sciences.

Research productivity was defined in four differentways. Each scientist was rank-ordered with his labora-tory peers both for his contribution to the general ortechnical knowledge in his field and his overall useful-ness in helping the organization carry out its respon-sibilities. The judges were supervisory personnel fromeach laboratory who knew the work- of each individualscientist they evaluated. Productivity was also rneasureoby counting the number of scientific productspublishedpapers, technical reports, and patent applications- -produced by each scientist over a five-year period.

9 18

For the purposes of this paper I will discuss only thefindings related to the environmental conditions.

The first variable of interest was freedom or auton-omy to determine the goals and objectives of the variousProjects conducted by the subjects/scientists. Freedomor autonomy was measured in terms of the number ofdifferent sources of decision-making influence broughtto bear on each scientist's work, the extent to which thescientists themselves could influence the other decision-makers, and who the decision-Makers were, for example,self, colleagues, supervisor, or executives. The findingsshowed that increased productivity occurred whenmultiple sources of decision-making influence were usedto determhe project objectives and when the scientistsbelieved they could influence the other decision-makingsources. PhDs in research settings were most productivewhen the decision-making sources were self., and col-leagues. Non-PhDs in research settings were most produc-tive when they and/or their colleagues determin,e4,theproject objectives. Of particular note is the144,the majority of the PhD-prepared scientisteVhd` hadtotal freedom to set project objectives were less pro-ductive than those who did no have such freedom.

The second variable of interest was communicationamong colleagues. In attempting to measure communi-cation patterns among scientists, Pelz and Andrews tookthe view that a research laboratory is a system ofinteracting scientists (and other components), in whichthe scientists stimulate each other to produce highquality work. In other words, the laboratory is' morethan a facility which simply provides services and equip-ment so that its scientists can conduct research (PelzAndrews, 1966, p. 35).

Four different measures of colleague communicationwere positively associated with research productivity forthe doctorally-prepared scientists and engineers: fre-quency of contact with most important colleagues,

19 10

amount of time spent in contact with important co!-leagues,, the number of people in their immediate workgroups, and the number of people from other groupswithin the organization with whom they had contact. Ingeneral, the most productive scientists were those who(1) had the most frequent contact with colleagues (dailyor several times per week); (2) spent the most time incontact with important colleagues (6 to 10 hours perweek per colleague for PhDs, slightly_ more forengineers); (3) worked with larger numbers of immediateco-workers (10 to 20); and (4) had more contacts withcolleagues outside their immediate work group butwithin the organizatin.

The last variable of interest was diversity in the typeof scientific teaching or administrative functions carriedout by the subjects. Diversity of function was measuredin terms of percent of time spent in administrative,scientific/technical, and teaching activities. Scientificfunctions were further defined as being research, devel-opment, or technical service functions. In general,research productivity was higher for scientists who spent25 to 50 percent of their time in non-research functions(for example, teaching and administration), and forscientists whose research activity included both researchand development functions.

An analysis of the relationship suggested by Pelz and,Andrews' work provides an answer as to how to build aclimate for collaborative research.

In relation to freedom or autonomy, the collaborativeresearch process requires the faculty and clinician re-searchers to reach mutually satisfactory decisions re-garding all aspects of the research processa conditionthat is similar to the self and colleague condition iden-tified by Pelz and Andrews. It is important to have inoperation an administrative review regarding projectselection so that it does not unduly interfere with thedecisions made by the. collaborative research groupmembers. .

11 201 'I

In relation to communication, it should be apparentthat the collaborative research process provides most ofthe conditions identified by Pe lz and Andrews. First, itrequires ongoing contact among the members of a col-laborative research group. Although the typical col-labbrative research group is small (2-10 members), thiswould still -represent a gain in number of colleaguecontacts for the members of the group. While colleague-to-colleague contact is an integral part of the col-laborative research process, it is difficult to imaginehaving the luxury of 6 to 10 hours of contact time perweek per colleague in the foreseeable. future. Nonethe-less, the collaborative research process does provideincreased communication by legitimizing the expenditureof time spent in contact with larger numbers of researchcolleagues.

In relation to diversity, the collaborative researchprocess clearly provides investigators with an opportu-nity, if not an expectation, to work on projects with bothresearch and development objectives. The collaborativeresearch process would have little influence regardingthe proportion of time spent on research versus. othertypes of job responsibilities. Clearly it will be a longtime before the base budgets of schools of nursing and

. clinical agencies can, provide the financial resources tosupport multiple investigators who spend 50 .percent oftheir time on research. However, some academic andclinical organizatioris are currently able to support mul-tiple investigators for as much as 25 percent time for .

research. Perhaps what is most important is the under-standing that research productivity is inhibited if lessthan 25 percent of an investigator's time is spent inresearch activity.

This analysis of the relationship between the threeenvironmental conditions positively associated with re-search productivity and the collaborative research proc-ess leads to the realization that the collaborative re-search process provides its own climate for the conduct

21

of research. The task that confronts us is not to build Aclimate for collaborative- research, but instead, to find away to stimulate and maintain its, existence as anapproach to scientific inquiry. If we can solve thatproblem, the collaborative research process will in andof itself provide a productive climate for nursingresearch.

References

Krone, K. P., do Loomis, M. E. (April 1982). Developingpractice-relevant research: A model that worked.The Journal of Nursing Administration, 7(4), 38-41.

Krueger, J. C., Nelson, A. FL, & Wolanin, M. D. (1978).Nursing Research. Germantown: Aspen SystemsCorporation.

Lindeman, C. A., & Krueger, J. C. (July 1977). Increas-ing the quality, quantity, and use of nursing research.Nursing Outlook, 25(7), 450-454.

Mauger, B. L., & Huggins, K. (1980). Developing andimplementing collaborative nursing education re-search in the South. Nursing Research, 1980,189-192.

Pelz, D. C., & Andrews, F. M. (1966). Scientists in orga-nizations. New York: John Wiley and Sons, Inc.

13'2

ESTABLISHING AND IMPLEMENTINGCOLLABORATIVE INVESTIGATIONS

Carolyn A. Williams, N, PhD, FAANProfessor ;rid DeanCollege of Nursing

University of Kentucky, Lexington

Overview

In an effort to be systematic in addressing theimportant topic before us, preparation of this paperbegan with a search of the nursing literature; however,little was found on the topic acollaborative research.Of the few items located, three f references proved to beparticularly helpful: Krueger Nelson, and Wolanin'sbook (1978) which documents their experience with theWestern Commission on Nursing Education's- ResearchDevelopment project and two papers, one by Felton andMcLaughlin (1976) and another by Dishrow (1982). Afterconsidering the nursing research literature, attentionwas directed to the social science literature whereseveral papers which were tangentially related to thetopic were located. What was impressive about theliterature review was the limited attention this vv.)+,important issue has received. Therefore, many of mycomments are based entirely upon my own experience incollaborative research and conversations with colleagues.

What is Meant by Collaborative Research?

Whether collaborative research is rare or relativelycommon depends on how it is defined. One approach isto refer to collaborative research as research developedby two or more individuals who pool their knowlt 4e andresources to pursue a research question. However, toleave the definition here is inadequate. It is importantto make a distinction between collaborative research

15 23

viewed as a cooperative endeavor among professionalswho consider each other as peers and who also sharevalues of democratic governance and group research.While group research involves multiple investigators, itssimilarity with true collaborative research ends there.In fact, group research may be handled in a veryhierarchial manner, and a high level of competition maybe the rule. In distinguishing between group researchand collaborative research, it may be useful to considera continuum with collaborative efforts as an ideal at oneend and group research, meaning simply several inves-tigators working together on a project, at the other.Projects can be distributed along dimensions such asparity among investigators and democratic governance.Further, it may be useful to point out that a givenproject may shift on these dimensions during its course.

Another characteristic of collaborative research isthe collaborative projects may or may not infolve col-laborators from two or more discipline's. This issue isparticularly confusing when discussing nursing researchbecause nurse researchers who are working together mayhave had very different programs of doctoral preparationand may approach research questions in very differentways. While a casual observer may see two nurses andargue that the project is really not interdisciplinary, itmay, in fact,, have the "feel" and some of the charac-teristics of interdisciplinary research.

Why is Collaborative Research Considered Important?

Many share an implicit assumption that collaborativeresearch is a good thing. The Division of Nursing seemsto take this positionthe Nursing Research EmphasisGrants for schools with doctoral programs are designedto- stimulate collaborative research both among facultymembers and between faculty and doctoral students.

24 16

Why is collaborative work being encouraged? Severalpossibilities can be suggested. Collaborative work isfrequently projected as a good use of resources becauseit can be argued that: (a) it increases the chances ofdeveloping information of broad significance; (b) itincreases the chances of generating a sequence ofstudies producing a body of information in a given area,and; (c) it provides a framework or structure for thecrucial socialization of doctoral students and post-doctoral fellows into active research roles. In my view,more collaborative work is indicated if nursing is goingto develop the productive research efforts necessary tocontribute to the knowledge base for health promotion,disease prevention, and nursing care. I further arguethat such collaboration should not be limited to the"borderlands" mentioned by Stine during this conference(SCCEN, 1985), but it is needed in areas of, centrconcern to the discipline..

When Is Collaboration Indicated?

I would propose that there are three major condi-tions, any of which might lead to the conclusion that oneshould engage in collaborative research. When all threeare present, collaborative research. is clearly indicatedand has a high probability of success. These conditionsare: (1) a strong mutual interest in the researchquestion(s); (2) the question(s) under investigation demandresources beyond those of one investigator to be ade-quately handled; and (3) the investigators' desire to workwith each other as colleagues.

Group research and collaborative research are moreof a tradition in some fields than others. For example,in epidemiology the research question frequently in-volves several levels of analysis. When considering_associations between various environmental charac-teristics, personal characteristics, and a health (or

17 25

disease) state, variables may be conceptualized at thesociological, the physical, and the biological level. An-other reason is the large numbers of individuals beingstudied. Epidemiologists ftequently ,undertake largecommunity prevalence surveys of various disease states,such as hypertension. And, there are cohort studieswhich observe people over long periods of time whichMay, indeed, span the careers of several researchers.The Framingham heart study (Haynes, Levine, Scotch,Fienleib C Kannel, 1973) is an example. Finally, thereare large intervention trials involving multiple clinicalsites and multiple investigators. Such group efforts haveconsiderable payoff in the creation of large data sets andthe generation of opportunities for many investigators,including doctoral students, to be engaged in the pursuitof research questions. Other fields have similar tradi-tions; in the biological arena, laboratories are pursuingparticular lines of inquiry and a number of investigatorsand their doctoral-postdoctoral fellows work together.

While there is less of a tradition of group andcollaborative work in nursing, we are moving in thatdirection. In fact, 48 percent of the papers presented inthis conference may, have been collaborative efforts.

program indicates that 16 of the 33 papers meet thec teria of having multiple investigators. Whether or notth investigations are truly collaborative in that theymeet the criteria set forth earlier cannot be determinedfroin simply looking at an abstract.

It is important to mention that nurses have had a longtradition of being involved in group research, but theroles they have assumed have frequently been quitelimited. These include being data collectors or havingbeen involved in the intervention. Participation in the

'Intervention may entail either providing part of the carethat is being assessed or being the subject of observa-tion. Nurses also frequently assume the role of coor-dinator of the care processes. Much less frequently

26

nurses have been identified as paid consultants, as prin-cipal investigators, or as co-prinicipal investigatorsroles where they provide expertise or are active indevelopment of the design and conduct of the inquiry.

For the last several yeirs I have had the opportunityto review a large number of research grants as a memberof various governmental and foundation peer reviewpanels. Unfortunately, in my experience, it is rare tosee nurses listed as co-principal investigators in projectssubmitted by physicians or principal investigators fromother disciplines. Likewise, in projects with nurses asprincipal investigators, it is rare to see co-principalinvestigators from other disciplines. We can, however,take some comfort in an increase in the number ofprojects in which- a nurse is a principal investigator andanother nurse is identified as a co-principal investigator,suggesting that group research _and perhaps collaborativeresearch may be increasing within the discipline. I am,of course, only one observer and do not have data toprovide a firm estimate of the proportion of all projectssubmitted, that would meet the criteria for group orcollaborative research.

Characteristics of Successful Collaborative ResearchGroups t,

Based on their experience in working with collab-orative research groups, Krueger, Nelson, and Wolanin(1978) have described such groups as posssessing thefollowing characteristics: well-stated goals; adequateresources; r attain the .goals; a division of labor thatercoordinates and integrates members' activities; andmechanisms to provide for the satisfaction of members'needs for tension management, status, and self-respect.In addition to such structural characteristics, I would addthat the social psychological climate of the group isextraordinarily important.

27.19

Research groups may be composed of- very strongindividuals who are used to making their own decisions.Such individuals may have difficulty in a group situationdue to a lack of experience and/or an unwillingness toconcede to group level decision making. In this contestpower struggles may develop. To establish and maintainfunctional groups, several conditions are important.First, investigators should have mutual trust in oneanother as professional colleagues. Second, it is essen-tial to have open patterns of communication amongmembers of the group. Third, it is important to haveconsensual decision making on: the specification ofgoals, the setting of priorities, a division of responsi-bility that utilizes the strength of group members, andthe provision of rewards. Fourth, there need to bemechanisms for the reaffirmation of shared values andinterests.

One reason for entering into a collaborative arrange-ment is to have people with different but comple-mentary skills working together in an effort that couldnot be achieved alone. Thus, there is a delicate balancebetween the specialized interests and abilities of indi-vidual group members and the group's overall focus andgoals. In a paper on the division of labor in small groups,Baker (1981), a social psychologist, hypothesized that tothe extent the tasks in a group are specialized, themembers become more dependent on each other and thelevel of cohesion increases. But, if there is too muchtask specialization, group members may feel isolatedfrom each other and cohesion is lower. He furtherhypothesized that isolation in a group where individualsare specialized can be offset by fostering a strongcollective identity or a high level of interaction. WhileBaker's hypotheses need further testing, the messagesare that group identity is an important issue; groupmembers need to see a need for each other; and groupmembers need to want to work with each other.

8:(20,

What to Think About in Considefing Initiation of aC6llaborative Research Project

There are three matters to which bne should give atten-tion when considering the possibility of a collaborativeendeavor. They are: the nature of the research prob-lem, one's own characteristics, and the characteristics ofpotential collaborators. With regard to the nature of theresearch problem, it is important to ask whether it isnecessary to work !with another person because of theneed for their conceptual or methodological expertise.Or, is working together deemed desirable simply becauseof mutual interest in a particular question and thebenefits of colleague stimulation? Those of us who havehad an opportunity to work in successful collabOrativerelationships have experienced situations in which theteam has been able to be creative in identifying prob-lems and arriving at understandings and solutions thatprobably would not be possible for an individual workingalone. Krueger, Nelson, and Wolanin (1978, p. 24)described this process in the following way: "A challeng-ing Idea presented to an individual at a particularmoment may elicit a creative thought that normallywould be dormant. In turn, this expressed thought mayspark an idea in another, and then another, in whatbecomes a chain reaction. The stimulation may comefrom reinforcement of similar beliefs or from beingforced to consider different viewpoints held by memberswith diverse experience, knowledge and temperaments."

In considering one's own characteristics the followingquestions, several of which were suggested by "Disbrow(1982), are important to ask before entering into acollaborative relationship. First, am I open to the viewsof others? Am I able to expose my ideas and work toothersto take criticism and to profit from it? This isCommonly referred to as having a thick skin. While sucha characteristic may be, difficult to acquire, it is essen-tial. Do I have a sense of humor? And, am I willing toshare with others in a democratic mode?...

21 i9

In considering the characteristics of potential col-laborators, it is important to determine how such indi-viduals measure up on two sets of attributes. Potentialcollaborators should be perceived as sharing similar valuesand interests, being trustworthy, being capable of opencommunication, having initiative, and being reliable andable to meet commitments. It is also desirable that theybe creative and enthusiastic.

Tradeoffs

In conclusion, I think it is important to indicate thatthere are tradeoffs involved in engaging in collaborativework. On the one hand, engaging in a collaborativeactivity may allow one to address questions one couldnot handle alone. On the other hand, it is necessary togive up some independence in decision making. Forexample, if one is working with several investigators in alarge community survey each may have a set of ques-tions they wish to ask. To have a manageable ques-tionnaire, it may be necessary for each investigator tocOmpromise and drop several treasured questions. Thus,while one may have the stimulation of group interactionon tie positive side, one must also abide by groupdecisions.

Other decisions involve who will have priority in theuse of resources, who presents the research and where,who publishes what and .where, and the order of author-ship. These are all important issues and their successfulresolution is highly dependent upon The presence ofappropriate conditions for collaborative work, the struc-tural arrangements for working together, and the inter-personal climate. In the final analysis, each participanthas to contribute to the group and be perceived as makinga valued contribution. When group members are seen asredundant, problems will develop in the allocation ofresources and in meeting with the mutual needs ofindividual;members.

u 22

Collaborative research is not for everyone. Most ofthose with whom I have discussed this topic concludethat the benefits outweigh the difficulties. They add,however, that on subsequent projects they will be morecautious as they make decisions about with whom to,work and the arrangements that will guide their jointactivity,y.

In my view it is imperative that nurse researchersaccept the challenge to develop truly collaborative re-search projects that exemplify healthy modes of inter-action and produce quality research. The disciplineneeds the knowledge generated by, such projects; thediscipline also needs the opportunities for the social-ization and mentoring of developing researchers suchprojects can provide.

References

Baker, P. M. (1981). The division of labor: Independence,isolation, and cohesion in small groups. Small GroupBehavior, 12(i), 93-105.

Dishrow, M. A. (1982). Conducting interdisciplinaryresearch: Gratifications and frustrations. in N.

'Chaska (Ed.), The nursing profession: A time tospeak. New York: McGraw-Hill, 289-297.

Felton, G., & McLaughlin, F. E. (1976). The collaboratingprocess in generating a nursing research study. Nurs-ing Research, 25(2), 115-120.

Haynes, S. G., Levine, S., Sc9tch, N., Feinleib, M. doKannel, W. B. (1978). The relationship of psycho-logical factors to coronary heart disease in the Framing-ham study, I: Methods and risk factors. AmericanJournal of Epidemiology, 107(5), 362-383.

Krueger, 3. C., Nelson, A. H., & Wolanin, M. 0. (1978).Nursing research: Development, collaboration, andutlization. Germantown, Maryland: Aspen SystemsCorporation.

23 311.

ETHICAL ASPECTS OF COLLABORATING WITH CLIENTS

Shirley P. Damrosch, PhDAssistant ProfessorCenter for Research

School of NursingUnitisrsity of Maryland at BaltimOre

Gadow (1980) identified the nurse as uniquely quali-fied among health professionals to serve as a clientadvocate and discussed the importance of advocacy indetermining nursing's future directions. According toher concept of advocacy, the nurse and client "freelydecide" .(p. 81) the exact nature of the nurse-clientrelationship, which can include the option of having theclient serve as colleague.

Just as nursing practice involves nurse-client inter-active roles, nursing research can be conceptualized interms of researcher-subject interactions and relation-ships (Denzin, 1978). Both types of interactions areguided by ethical standards guaranteeing rights of clientsor subjects to make informed decisions about their careor their participation in research studies (Damrosch doLenz, 1984). It is reasonable, therefore, to extendGadow's (1980) idea of client advocacy to the researchcontext. This report outlines methods by which nurseresearchers can serve as advocates by collaborating withclient-experts in making decisions about important as-pects of clinical research.

Clients who experience illness or other aversive lifeexperiences may gain a perspective that nurses and otherhealth professionals may lack. For example, see studiesby Lauer, Murphy, and Powers (1982) and by Adorn and

This_effort was supported by Grant No. 1 R21 NU-00829-03from the Division of Nursing, Health Resources Ad-ministration, U. S. Public Health Service, Departmentof Health and-Human Services.

Wright (1982) on how nurses as "outsiders" misperceivedthe needs and priorities of cancer and ophthalmologicalpatients, respectively. Collaboration with clients whoare uniquely qualified because of their perspective as"insiders" can help researchers avoid such misperceptions.

Active involvement of such clients in the planning ofresearch is encouraged to accomplish the five-fold pur-pose of (a) tapping this source of client expertise, (b)allowing client-experts to serve as advocates of theresearch rights of their respt.-tive groups; (c) enhancingthe social relevance of nursing research, (d) insuringgreater respect for clients and subjects by making theresearcher sensitive to their unique needs and require-ments, and (e) increasing the likelihood that researchfindings will be implemented by and/or with respect forthe target population. This kind of involvement would beparticularly appropriate when research involves sensitiveissues, _ or studies subjects at particularly vulnerableperiods in their lives.

One example of a research effort which activelyincorporated clients as advocates and advisors is a studyof parents' psychological reactions to the loss of theirinfant due to Sudden Infant Death Syndrome (SIDS)directed by Dr. Camille Wortman at the University ofMichigan. A unique feature of the project was theestablishment of a research advisory board consisting ofparents who had suffered a SIDS loss. The researcherssolicited input from these parents during the earliestplanning stages of the study. As a result, the researchquestions for the SIDS project stemmed not only from areview of the literature, but also from conversationswith parents who were consulted as exper thepsychological aspects of the SIDS trage parentaladvisory group served as a conduit of regular feedbackduring all important phases of the research process.

-

A group of researchers (Lesley Perry, Lynn Nalven,and Shirley Damrosch) at the University of MarylandSchool of Nursing have successfully collaborated with apanel of client-experts in planning and implementing a

26

study of coping among parents of infants with Down'sSyndrome (DS). These experts are parents of childrenwith DS. (See Damrosch and Lenz, 1984, for details onthe recruitment of this panel.) The major accom-plishments of this collaboration include:_

Validation of the study focus. The parents' encour-agement and enthusiasm reinforced the researchers' de-cision to focus the study on parental coping.

Increased sensitivity to the neeLs and emotions ofsubjects. The input of the parental advisors was instru-mental in making the researchers sensitive to possibleproblems in a number of ways. First, the advisors' earlycriticism led to the avoidance of wording in the inter-

- view that was in any way offensive. They aided theinvestigators in determining optimal sequencing of itemsin the interview schedule, as well as the time of theinterviews. It was their unanimous view that the parentswould be ready to be interviewed as early as two monthsafter the birth of their infant.

Substantive contributions to instrument construction.The advisors not only critiqued items in the variousinstruments,, but contributed items of their own. Theseitems addressed dimensions of coping behaviors whichthe, researchers had not previously considered and whichwere not included in previously published instruments.

Assistance with ethical considerations. The input of_parent-experts helped assure that the rights of subjectswere protected, benefits maximized, and risks mini-mized, because an additional perspective was added tothe human subject review process. Moreover, theirsuggested additions- served to enhance the potentialusefulness of the study.

As a result of actually working with a panel ofparental experts in the area of coping with the demandsof rearing children with Down's Syndrome, the re-searchers became even stronger advocates for col-laborating, with- client-experts, and make the followingrecommendations:

27

34

1. More researchers, especially In sensitive areasof investigation, should seriously consider int-clusion of a client-advisory panel in the planninjgand implementation,of their studies. Such panelsshould be useful in improving the. quality ofresearch, enhancing its social relevance, andinsuring that subjects are interviewed in themost sensitive and ethical way possible.

2. Input from client-advisors on the ethical aspectsof proposed research could be an invaluablesource of information for institutional reviewboards.

3. Client experts who perform services for re-searchers should be given honoraria asrecognition of their contributions in much thesame manner as professional consultants.

References

Adorn, D., & Wright, A. S. (1982). Dissonance ill nurseand patient evaluations of the effectiveness of apatient-teaching program. Nursing Outlook,132-136.

Damrosch, S., & Lenz, E. (1984). The use of client-advisory groups in research. Nursing Research, 11,47-49.

iDenzin, N. K. (1978). The research act (2nd ed.). NewYork:' McGraw-Hill. 0.

Gadow, S. (1980). Existential advocacy: Philosophicalfoundation of nursing. In S. F. Spicker & S. Gadow(Eds.), Nursing images and ideals. New York:Springer.

Lauer, P., Murphy, S. P., & Powers, M. 3. (1982).Learning needs of cancer patients: A comparison ofnurse and patient perceptions. Nursing Research, 31,11-16.

35;285 ,

THE ETHICS OF .INTER-DISCIPLINE COLLABORATION

OsCar Stine, MD, DrPHChief of Ambulatory Care

Greater Baltimore Medical Center, Maryland

The Definition of Ethics

I ascribe to the proposition that there is no ethicsunless there is a, decision. The decisions that we havebeen asked to consider are whether or not, and how tocollaborate with members of anoter discipline to under-take a project. This requires anrithical analysis of thedecision by identifying conflicting values and making adecision on the basis of the priorities we assign to eachof these values.

Ethics of Professional Competence

Many of the values we hold are related to theprofessional competence that we recognize in ourselves,or in the person with whom we may collaborate. Each of tus has a set of competencies that is unique to us asindividuals, and a broader set that is unique to ourprofession. To be competent in the problems and sub-jects of your profession means that you are prepared, orexperienced, in dealing with all the combinations andpermutations of measurements, assessments, and prob-kif solving within that area.

Each profession has a larger body of information andskills that it shares with other professions; we callthese borderlands of the profession. In these border-lands, we have not solved all the combinations andpermutations. We recognize the opportunity for col-laboration with another professional to obtain his or herhe in the borderland in which there are shared skillsand information. The valuet \that should be explored

29 36

Includes What are the unique skills and Informationavailable from the potential collaborators?' How nec-essary are these skills and information to the completionof the task? If the probleth lies in the borderlands oftwo or more professions, Is it susceptible to solution bycollaboration?

Professional Relations Within a Profession Or Agency

There are values achieved by working within a singleprofession that are not as accessible in collaborationoutside of a single profession or organization. Within asingle profession there tends to be team work organizedin a hierarchy of functions. Tasks are delegated verti-cally to members of the same discipline or specialty.Budgets are prepared with the priorities held in commonby the participants. Prestige an4awards are grantedaccording to the s ndards of the specialty. Consid-eration of causes usually limited to those of interestto that specialty. Communication-between members ofthe same discipline uses the same jargonand proceeds withaccepted assumptions. Procedures, measurements, andrecords are carried out and completed by the newestmember of the profession.

These characteristics of work within one depaArt entsupport the professional person's unique store of infor-mation and skills. With all of these benefits derived fromwork within a department, why should a professionalperson venture into work in which the information andskills are shared by other disciplines; specialties, ordepartments? Why accept delegation 'f responsibility bysomeone who is equal to you in their training on thesubject, or by someone who is not part of your ownprofession? Why deal with multiple causes that are theinterest of multiple disciplines when a necessary orsufficient cause is well known by your own specialty?Are you willing to sacrifice your dependence upon yoursupervisor to gain autonomy in a democratic process?Do you strive for creativity that comes from "cross

3 7 30

fertilization" with someone' different from you? Theseare the vulnerabilities and the oppollunities of investingyour efforts in collaboration with,a professional petsonwho shares an information base and skills that are nottypical ofthe cores of either profession.

Collaboration Between Members ofDifferent Professions Within Borderlands of the Two

Collaboration represents an activity in which pro-fessional persons venture into subjects or fields forservice, education, or research in which they were notfully trained in their profession. In these areas, thereare ideas, methods, and facts still to be found to whichother professions can contribute. The habilitation ofretarded children, the developmental problems of teen=age mothers, the multiple problems of impoverishedfamilies, the sudden infant death syndrome, the problemsof family violence and child abuse are characteristic ofthe problems that have been submitted to collaborativeefforts in the past 25 years to the immense benefit ofthe clients served and the collaborating professions. Toaccomplish this, physicians, nurses, social workers, stat-isticians, psychologists, and others had to venture outinto services and subjects not fully delineated, much lessunderstood. This suggests that there are values that Weuse to define our roles as professionals, and there areseparate values to define our roles in collaboration onsubjects and methodologies that are shared Ilith otherprofessions.

Valuthgy ed Qualities in a Collaborator

The process of collaboration does not deny the impor-tance of the core training and experience of eachprofessional. Competence is needed from each memberof the team to contribute information and *ills. If oneperson in one profession simply adds his or her own work

31 38

to that of another, it is like putting the pieces of a puzzlc:together. I would submit that this is only the first ravelof collaboiation because it allows each IndiviNal toremain in his or her core competency and does not4requirereorganization of decision-raking hierarchies to whichthe participating-professionals belong.

The 'greatest opportunity for collaboration lies withinthe borderlands between,the defined competencies of theseparate professions. The qualities of professionalethatare valued most in the borderlands are Initiative, crea-tivity; clarity of communication, scientific discipline,and enthusiasm. These are personal characteristics thatare particularly useful in combined efforts for problem-solving or discovery. Since the nature, of the eventualsolution or of the future discovery may- not be fullyknown, you cannot clearly predict which of the partic-ipating professions is going to make the greatest con-tribution. These personal characteristics of the partic-ipants provide the greatest assurance that progress to-ward the solution or discovery will proceed smoothly.

Ethical Goals to be Sought in Collaboration

Autonomy is highly valued in collaboration. The self-governance of the participants is the basis for demo-cratic decision making in which each collaborator con-tributes according to his or her preception of the mosteffective or efficient means of reaching a shared goal._Autonomy is the basis for negotiation of a fair contractand democratic governance of the collaborative project.

Fairness is a major value for assigning tasks and fordistributing rewards in the collaborative, process. Thebasic test for fairness is to determine that a contributionor distribution is equally acceptable when it is given orreturned to the other collaborator(s). Their collab-oration requires that the rewards are proportional to thecontribution, but further ethical analysis will reveal thatthis giving and receiving may be according to ability,investment, productivity, or need. Obviously, there is no

0 3,9 32

formula for fairness, but because it has multiple dimen-sions, the parameter chosen to determine fairness is verysusceptible to negotiation among collaborators.

The concept of "relative advantage" is boriowed fromeconomics. It states that when a commodity can beproduced in many locations, the producers who candeliver the commodity most efficiently will receive thegreatest reward. When tasks are being assigned in acollaborative project, the collaborator who can completethe task successfully with the least effort is the obviousperson for the assignment.

Equality is a value related to autonomy and alsocontributes to the democratic process. In this instance,equality is determined by the goal of the project. It is anequality of opportunity to contribute, initiate, plan,coordinate, write, edit, or to convene colleagues for anyof these activities. Rotation of these responsibilitiesbetween collaborators is a sure test of equality. By this'test, we can isolate the collaborative fun tions and theconsultative function from the professional core of theunique contribution of each professional.

Summary,

Collaboration can be undertaken by autonomous pro-fessional persons who are oriented toward a task inservice,, teaching, or research. Their work has jointownership, which includes the contribution of the expe-rience, skills, motivation, and discipline of each indi-vidual, as well as the core of professional informationand skills in which each is trained. The process ofcollaboration occurs in the borderlands of the two ormore specialities involved and requires individual auton-omy for democratic decision making in these borderlandswhere it is necessary to be creative for problem solvingand discovery. The personal attributes of collaborators

33 46

Must be those that will move the ,group effort towardtheir shared,goals. Fairness in the assignment of tasksand in the distribution of rewards may be related toability, effort, productivity, or need, but should benegotiated as part of the collabOrative process. Effi=ciency is achieVed by-undertaking that task according tothe relatiye advantage of one or more collaborators.The functiObs of initiating, coordinating, writing, edit-ing, had convening may be fulfilled by any of theparticipants in the shared borderlands, and the rotationof these fundtions among participants is a test of theirequality. Collaborative projects are sought.by those whodesire to solve problems or make discoveries in areasOutside of the domains of single professions or special-ties, who desire to assume autonomous roles in a dem-odratic group in which the chain of command, verticaldelegation of tasks, budgeting, awards, and prestige oftheir specialty are temporarily laid aside for a col-laborative experience.

41 34

OWNERSHIP OF DATA AND PLAGIARISM IN RESEARCH

Carolyn F. 'Waltz, RN, PhD, FAANProfessor and Coordinator of Evaluation

School of NtirsingUniversity of Maryland at Baltimore

In the health professions, collaborative research proj-ects involving individuals with a wide range of skills,training, and roles are rapidly increasing in number andcomplexity. Given the proliferation of research con-ducted by teams, and the fact that records of researchand publication are critical variables in determiningone's status and position in the professional community,the ownership of data and the. assignment of credit whenpublishing manuscripts have become salient issues.

Unfortunately, there is little clear-cut agreementamong or within the various health professions regardingwhat is appropriate and inappropriate assignment ofcredit, especially in collaborative research situationswhere there are multiple and varied contributions madeby individuals with different educational backgrounds,organizational status, and experiences in research.

This fact is reflected in the results of a study,conducted by Waltz, Nelson, and Bond, in which theysurveyed a random sample of 400 health professionals inthe fields of nursing (N), dentistry (D), medicine (M),pharmacy (P), and social work (SW), who had published inrefereed journals in their respective fields between 19)8and 1980, to ascertain their views regarding the assign-ment of publication credit. Using a modified version ofan instrument developed by Spiegel and Keith-Spiegel(1970), respondents were presented with 36 collaborativeresearch situations, and suggested possibilities for author.,ship assignment. Scenarios addressed issues-regarding:

(a) using someone else's research idea;

.(b) the paid consultant;35 42

(c) the continuing research team;

(d) the highly qualified sub-doctoral assistant, stu-dent assistant, and volunteer assistant;

(e) professor-student collaboration;

(f) multiple publications from a single study;

(g) the research program administrator;

(h) the pre-study agreement;

(1) ordering of credits baied on functions performed;

(j) possible solutions to the multiple authorshipproblem.

Six areas in which agreement was evidenced amongthe various professions were, for the most part, relatedto instances in which credit for sub-doctoral and/ortechnical contributions come into play. Those areas inwhich, agreement was not evident were more closelylinked to research in the area, to assignment ofcredit when respective contributions are equal, whenstatus factors come into play, and when authorshiporder, other than first author, must be determined.

To illustrate the nature of the issues to be addressed,it is useful to examine responses to selected scenarioswith particular relevance to the ownership of data and

First, consider the question: What credit, if any,should be given to the source of an idea which leads to aresearch study? Reactions were sought to the follow-ing situation by respondents:

Dr. Brown informally suggested a research idea tohis colleague Dr. Smith. Dr. Smith then designed

43 00JJ

0

an experiment based on this idea, added ideas ofhis own, collected data, analyzed the results, andwrote a paper for publication.

A. He should give Dr. Browna co-authorship.

B. He should give Dr. Browna footnote, acknowledgingthe source of the idea.

C. Dr. Smith Is under noobligation to acknowledgeDr. Brown-as the source ofthe idea for the study.

D. He should have askedDr.-Brown to collaborateon the study.

Published Sample (n= 185)

Percent by Discipline PercentD MN P SW Total

5 8 4 0 3

31 42 30 35 66 41

8 3 22 17 6 11

56 48 44 48 28 45

In line with "asking Dr. Brown to collaborate," weremany comments about the possible nature of the rela-tionship between the two men and suggested possibilitiesfor not collaborating. "Why hadn't" or "Why didn't Dr.Brown initiate the research?" were often asked. If Dr.Brown decided not to collaborate, comments suggestedthat it would still be courteous to acknowledge hiscontribution in a footnote. `

Many comments elaborated on types of circum-stances that would affect the credit that Dr. Brownshould be given. The completeness of the original ideaand the way Dr. Brown presented the idea were impor-tant considerations in determining the credit to be given.

If Dr. Brown had suggested in passiiii7rwohder what_happens if . . then his suggestion would be less

4important than ithe had a more concrete plan of action.Some of the diverse comments about ideas included:

"The idea is a rare and valuable thing. Conductingand reporting research are commonplace skills."

, ."Ideas come from many sources and many ideasrequire maturation and the source of the originalimpetus is not known.",

"It is hard for me to imagine a research idea, thatdoes not change in development."

"Ideas are a dime a dozen."

Next, consider another scenario that addresses essen-tially the same question in a slightly different manner.

Dr. Jones consults with Dr. Frank regarding aresearch project Dr. Frank wishes to conduct atDr. Jones' clinical agency. Dr. Jones informs Dr.Frank that he will explore the possibility with hisstaff and get back to him. A year, later, Dr.Frank, who has not yet received Dr. Jones' re-sponse, -finds a study in the literature by Dr.Jones which investigates the problem and employsthe design proposed by Dr. Frank. When a formalcomplaint is lodged by Dr. Frank, Dr. Jonesargues that Dr. Frank's thinking contributed toideas he was already formulating, but not suffi-ciently enough to warrant credit or a collab-orative offer. Furthermore, he contends that thestudy plan proposed by Dr. Frank had a number offlaws and thus required a great deal of effort -onhis part to make it viable.

45 38

-

A. Dr. Jones is under noobligation ethically toacknowledge Dr. Frank.

B. Dr. Jones should have givenDr. Frank a footnote,acknowledging him asa source of the idea.

C. Dr. Jones should have asked-Dr. Frank to collaborateon the s.ady.

D. Dr. 3Ones should have givenDr. Frank a co-authorship.

Published Sample (n 176)

Percent.by DisciplineM t1 , P SW

PerceiitTotal,

0 3 0 13 0 4

16 13 11 11 19 14

82 71 85 70 81 71

3 6 4 6 0 4

The majority of respondents agreed that Dr. Frankshould have been asked to- collaborate. However, therespondents felt that if Dr. Frank had not followed up onthe initial proposal within a year, he really had norecourse. It was also suggested that the original pro-posal should have been made in writing. Several respon-dents also indicated that if Dr. Jones had really beengeeing up for such a study, he should have indicatedthese preparations at their meeting.

Yet another dimension of issues regarding using someone else's ideas, where agreement was more evident,relates to the question: Who holds the ultimate respon-sibility for assuring that collaborative research is con-ducted in an ethical manner?

Professor Bright requests that Mr. Smith, agraduate student, conduct a literature review.Mr. Smith delegates portions of the task tothree of his- undergraduate students. ProfessorBright publishes the resulting article with Mr.Smith. Subsequently, Professor Ross lodges a

39

4-6

formal complaint stating that the work publishedby Professor Bright and Mr. Smith containsverbatim passages from his work and that hiswork -is not acknowledged anywhere in thearticle. Professor Bright faults Mr. Smith,who, in turn, faults the three undergraduatestudents.

A. Profe;sor Bright, as seniorauthor, is ultimately ac-countablcand responsiblefor assuring that the work,was completed in an ethi-cal manner, even thoughhe delegated a portion ofhis work to Mr. Smith.

B. Professor Bright and Mr.Smith as co-authors areaccountable and responsiblefor assuring that the workreported by them was corn-pleted in an ethical manner,

C. Professor Bright andMr. Smith were ethicallyresponsible for their workand thus should not have

- delegated tasks but rathershould have conducted theirown literature.review andwritten their own report.

D. It is unreasonable to expectauthors to assume respon-sibility for all tasks com-pleted during the conductof a- study and accountabilitymust be placed with thoseresponsible for a task.

Published Sample (n z 171)

Percent by Discipline Percent.M .N P SW Total

-19 36- 15 32 27 26

70 33 70 54 64 62

8 0 13 t 9 3

3 6 0 6 0

4 7 40

Respondents agreed that the authors are responsiblefor the ethical conduct of the work; 70 percent extendthis responsibility to both authors, while 26 percentwould limit it to the senior author.

Reading of the literature for a review article bythose who wrote the article was considered an importantdeterminant in this situation. The researchers should becognizant of important researchers in the area and thenature of their work and, thus, are able to acknowledgeresearch appropriately.

Literally, by definition, plagiarism means to steal orpass off the ideas- or words of another as one's own; touse a created production without crediting the source; tocommit literary theft; to present as new and original anidea or product deriving from an existing source(Webster, 1977).

It is interesting to note that while agreement wasevident in response to the scenario relating to usingverbatim passages from another work in an article, i.e.;to commit literary theft; responses regarding what con-stitutes plagiarism during the formative stages of aresearch project, i.e., using someone else's ideas, wereless in agreement and subject to a number ofqualifications.

Now consider the following questions regarding own-ership of data: Should a principal investigator whoadministers a grant comprised of many individual studiesclaim ownership of data and receive authorship on re-sulting manuscripts? Is it ethical for professors toestablish a policy whereby they will not chair a student'sthesis or dissertation committee unless given authorshipon any publication resulting from the effort? Whatcredit, if any, should be given to a professional aclinical agency without whose assistance subjects mightnot have been readily available?

41 48

Dr. Ebal, a well known researcher whose careerhas advanced to the point where -he acts primarilyas the adMinistratively responsible investigatorfor many research studies land who has severalresearchers and technicians working under hissupervisionV-demands that all publications arisingfrom his program bear his, name as an-author eventhough. he does no work on the projects and- neverdoes more than read the-manuscript.

A. Dr. Ebal should not evenreceive a footnote, becausehe never actually makes anycontribution to the studies.

B. Dr. Ebal should receiveonly footnote recognition.

C. This policy is warrantedas longas Dr. Mal isthe lastauthor listed.

D. Dr. Ebal is entitled tosenior authorship on aUmanuscripts If hedemands It.

Published Sample (n = 176)Percent by Dithipline PercentDMNPSW TO*

30 39 33 21 33 28

32 23 19 37 61 33

32 43 41 42 6 34

3 3 7 0 0 3

Respondents were about equally divided on no credit,footnote, or last author for an administrator who made .

no substantive contribution to a particular manuscript.

Comments speculated on the nature of the admin-istration and funding. If Dr. Ebal secured funding as theprincipal investigator, then respondents felt he shouldreceive author credit. If the funding was secured by-thereseachers responsible for the project, then no creditwas deemed necessary and footnote credit was con-sidered a courtesy.

42

49

Some consideration of Dr. Ebal's contribution to theresearch efforts of the lab in the way of ideas and.efficient management was also mentioned.

A policy of being, named second author on all publica-tions stemming from ,students' theses or dissertationswas considered unethical by 63 percent of the respon-dents to the following. situation.

Dr. White has as one of his job responsibilities thedirection of graduate students' research. Dr.White has established a policy whereby he will notchair a student's thesis or dissertation committeeunless the student agrees to name him- as secondauthor on any publication resulting from thestudent's research efforts.

Published Sample In = 176)

A. This is an unethical practice

Percent by DisciplineD M .N P SW

PercentTotal

that needs no qualification. 49 31 50 3$ 71 47

B. This is an unethical'practice since the direc-tion of student researehis a specified function inDr. White's'job description. 16 22 '31 6 12 16

C.' '?his practice Is ethicalonly If Dr. White assumesresponsibility for the designand conduct of the studybeyond the involvementexpected by the jobdescription. 27 4i 15 42 1$ 30

D. This is an ethical practicethat needs no qualification. S 6 4 15 0

The extent of the professor's contribution is again thedetermining factor in extending authorship credit.

43 50

Many respondents commented that a practice such asthis was an abuse of the professor's power and expressedconcern for students who might have no alternative inseledting an-advisor.

If Dr. White assumed responsibility for the design andconduct of the experiment (Answer C), then 30 percentof the respondents would consider the practice an ethicalone. Several comments mentioned research in thephysical sciences where student dissertations are oftenthe outgrowth of a professor's laboratory research; insuch situations, second authorship seems to be a routineand accepted practice.

A researcher'sethical obligation to a colleague whoprovides assistance and access to subjects for datacollection is considered in the next situation.

Dr. Blue conducted clinical research at a localmental health agency. A staff member in theagency helped Dr. Blue gain entry to the agencyand assisted in the collection of observationaldata over a 6-month period. The staff memberdid not contribute to4the design-of the study.

A. Dr. 1314. should acknowl-edge the staff-member'sassists* In a footnote.

B. Dr. Blue sheiuld name thestaff me as juniorauthor.

\C. Dr. Blue has no obligatIon

to acknowledge the\assist-ance of the staff member.

Published 4ample (n = ISO

Percent by Discipline PercentP SW 241

74 69 71 S2 SS 71

26 21 15 14 9 IS"

0 3 7 K 3 3

1 k4

)

,The majority of respondents agreed that footnotecredit would be appropriate. Comments were aboutequally divided between footnote credit and authorcredit. The uniqueness of the situation and the avail-ability of other sources of subjects for the research werementioned as factors in such a decision. "Courtesy ,to-colleagues is always a reason for at least acknowledgingtheir contributions."

In general, responses indicated that many factoincluding, but not limited to, status, position, nature, andextent of one's contribution to the research determineone's views on ownership of data. It appears that issuesto be resolved in this area focus primarily on questions, -

regarding who, and tinder what circumstances, should begranted ownership of the types acknowledged by juniorauthorships, and footnotes.

The obvious lack of agreement among the respectivehealth professionals represented in this study, suggeststhat much work needs to be done before inter-professionalresearch collaboration can occur in a manner satisfactoryto all:involVed.

references

Patton, M. Q. (1980). Qualitative evaluation methods.Beverly Hills, California: Sage Publications.

Spiegel, D., Sc Keith-Spiegel, P. (1970). Assignmentof publication credits: Ethics and practices of psy-chologists. American Psychologist 25, 738-749.

Waltz, C., Nelsiin, B., ac Bond, S. Assignment of publica-tion credits: Views of nurses and other health profes-sionals. Accepted for publication in Nursing Outlook.

Webster. (1977). Webster's New Collegiate Dictionary.New York: G. ac C. Merriam Company.

52

MODELS FOR MULTIVARIATE EFFECT SIZEESTIMATION IN CLINICAL NURSING RESEARCH_

No L. Abraham, RN, MS.Teaching Fellow

Center for Nursing Research

Samuel Schultz II, PhDProfessor and Director of Computing Facilities

Center for Nursing ResearchUniversity of Michigan, Ann Arbor

With the current trend in nursing to approach humanphenomena from a holistic perspective, there is a con-comitant need to develop multivariate statistical proce-dures that allow for the full incorporation of this pen,spective into the various types of nursing research. Thisissue_ is particularly pertinent in light of the currentproliferation of nursing research findings, and the con-sequent exigency to summarize and synthesize thesefindings into distinct bodies of empirical nursing knowl-edge: The purpose of this paper is to present and discussmodels for multivariate meta-analytic procedures Innursing research.

One of the major problems confronting nursing re-search today may be the challenge of translating theprinciples of the holistic approach to human phenomenaof health and illness into holistic research paradigms.Regrettably, nurse. researchers, while supporting a ho-listic perspective and operating within such framework,have for the most part employed univariate statisticalprocedure& to analyze univariate designs and answerunivariate questions. In other words, whileacknowledging that multiple factors are operative in theoccurrence or non-occurrence of human phenomena ofhealth and illness, nurse researchers have failed toaddress these factors as such. Studies have eithertended to focus on one isolated factor,_ or to collapse the

47 53

multivariate entity of multiple 'factors into the singu-larity of one, univariate factor. Such strategies, whilenoble in cause, nonetheleis ,reduce the multidimen-sionglity of nursing and, health to the unidimensionalsphere.

A similar univariate orientation is discernible in thestatistical paradigms proposed for summarizing andsynthesizing research studies in concept formation andtheory building (Glass, 1976; Hedges, 1982, 1983;Rosenthal, 1983; Rosenthal & Rubin, 1982; O'Flynn,1982). Yet Abraham and Schultz (1983), notingthe discrepancy between the holistic framework and theunivariate models, called attention to the need fornursing (as well as for the other health and socialsciences) to develop multivariate statistical models formeta-analytic effect size estimation.

Three models for such meta-analysis of nursing re-search are presented, the differentiation between modelsbeing based upon the number of methods/instruments

,i used to measure factors .j. For the purpose ofillustrating each model, examples from a hypotheticalset of studies on the post:mastectomy well-being ofbreast cancer patients are used.

(a) Model 1: Multifactor-Unimethod, in which in kindependent studies, each factor f_i is measuredby one method Lan,i. Assuming well-being con-sists of three factors (psychological, physical,social well-being), this first model would beappropriate in the effect size estimation ofstudies that utilize one specific method/instru-ment for each factor, thus yielding a designcomprising three ethods/instrumentsand rilo,3 to measure factors fl, f2,,and f3,respectively.

(b) Model 2: Balanced Multifactor- Multirnethod inwhich, in k independent studies, each factor h ismeasured by an equal number of methods r-tli

5448

(with i = 1,...,p). Whereas the first model appliedto k multivariate studies with univariate mea-surements of factors this second model de-fines procedures for estimating the effect sizeof the vector of factor effects (ti,9-2.,23) when afixed number a Of measures is utilized.

(C) Model 3: Unbalanced Multifactor-Multimethod,in which, in k independent studies, each factorL is measured by an unequal number of methodsm (with i = 1,..".,n). Designs to which thismodel applies comprise at least one factor L1measured by multiple methods/instruments(thus, I.> 2), and at least one factor .f.± measuredby a sole method/instrument i .

The paper concludes with a discussion of the advan-tages of multivariate meta-analysis effect size estima-tion in nursing theory development, and with a delinea-tion of the cautions that are necessary in applying theproposed strategies.,

References

Abraham, I. L., & Schultz, S. (1983). Univariate modelsfor meta-analysis: Comments on O'Flynn. Manu-script under editorial review.

Glass, G. V. (1976). Primary, secondary, and meta-analysis of research. Educational Researcher, 5,3-8.

Hedges, L. V. (1982). Estimation of effect size from aseries of independent experiments. PsychologicalBulletin, 92, 490-499.

Hedge,s, L. V. (1983). A random effects model for effectsizes. Psychological Bulletin, 93, 388-395.

O'Flynn, A. (1982), Meta-analysis, Nursing Research,31,314 -316.

49

55

Rosenthal, R. (19$3). Assessing the statistical andsocial importance of the effects of psythotherapy.Journal of Consulting and Clinical Psychology, SI4-13.

/-Rosenthal, R., & Rubin, 13. B. f1982Y. Comparing effect

sizes, of independent studies. Psychological Bulletin,92 500-504.

5.6 50

NURSING ADMINISTRATION IMPLICATIONSOF THE RELATIONSHIP OF TECHNOLOGYAND STRUCTURE TO QUALITY OF CARE

Judith W. Alexander, RN, PhD/Assistant. Professor, College of Nursing!University of South Carolina, Columbia

For years organizational theorists_have proposed thatthe fit between the technology of an organization and itsstructure can be linked to organizational effectiveness(cf, Child, 1972; Harvey, 1968). This study investigatedthe linkage between the diinension sot technology andstructure on pursing subunits to the ,effectiveness ofthese subunits in terms of quality of nursing care.

Thirty-four nursing subunits in three general hospitalsparticipated in the study with complete dada obtained on27 subunits. A total of 187 nursing personnel and 340patients were included in the study. Nursing personnelincluded the head nurse and one to five nursing personnelfrom each subunit, representing an 81 percent responserate. Patients were randomly selected from those whomet the criteria established for the study with 7 to 20patients participating on each subunit,

Technology. -was defined as the acts employed bynursing personnel to change individuals from being hos-pitalized to being discharged and was measured using theOverton, Schneck, and Hazlett (1977) 34-item, 5-pointLikert-like instrument. Three dimensions of technologyemerged through factor analysis: instability, variability,and uncertainty. Reliability coefficients for the dimen-sions in this study were .86; .77, and .81, respectirlf.

Structure was defined as the allocations of worlt,\rolesand administrative mechanisms tr control and integcatework activities. It was measured-by the 21-item, SpointLikert -like instrument employed by Leifer and Hubei,(1977) to measure flexibility of structure. This instru-ment was developed from the empirical findings of

51S7

Duncan (1971). Factor analysis of this data producedthree dimensions of structure: vertical participation,horizontal participation, and formalization. Reliabilitycoefficients for the dimensions were .74, .63, and .60,respectively.

Effectiveness was defined as the quality of caredelivered on nursing subunits and was measured using theRush-Medicus Nursing Process Monitoring Methodology(Hegyvary, Haussmann, Kronman and Burke, 1979).This instrument measures quality in six dimensions: thenursing care plan, attention to physical needs, attentionto non-physical needs, evaluation of care, protection ofpatients, and -administration of the delivery of care.Data were obtained from a variety of sources on thenursing subunit on different days of the week and dif-ferent hours of the day. The two pair of RegisteredNurse observers who collected the data achieved interra-ter xellabilities of .88 and .92.

Data were analyzed by calculating fit variables,which were the absolute value of the differences be-tween the combination of the technology and structurefactor scores. Regression analysis, using the three mostsignificant differences as independent variables, pro-duced a significant model (F = 8.28, P .001) thatexplained 52 percent of the variance in quality of care.Interpretation of this model suggested that nursing unitswhich have high technological instability and requirefrequent nursing observations and technical monitoringshould use a structure that has low vertical participa-tion in which superiors do not often seek- out theirsubordinates in decision making. Nursing units withgreater technological variability, characterized by vary-ing patient diagnosis and problems, should develop astructure that allows more horizontal participation de-noted by positive attitudes toward the involvement ofpersonnel in decision making and in the definition oftasks. Finally, nursing units with more technologicaluncertainty, suggested by complex nursing problems,

58 52

should employ a structure containing more rules andregulations to produce higher quality nursing care.

These results indicate that nursing administratorsshould look at the technological aspects of a nursingsubunit prior to making decisions concerning thestructural characteristics of that subunit. These struc-tural characteristics could include staffing patterns, co-ordination with other departments, and _decision-makingpatterns.

References

Child, J. (1972). Organizational structure, environmentand performance: The role of strategic choice.Sociology, 6, 1-22.

Duncan, R. G. (1971). Multiple decision making structuresin adapting to environmental uncertainty: The im-pact of organizational effectiveness. Working paper,Northwestern University Graduate School of anage-ment, Paper No. 54-71.

Harvey, E. (1968). Technology and the structure oforganizations. American Sociological Review, 11,247-259.

Hegyvary, S. T., Haussmann, R. K., Kronman, B., dcBurke, M. (1979). Users' manual for Rush-Medicusnursing process monitoring methodology. (Appendomto monitoring quality of nursing care: Part 4.) NTIS# HRP0900638.

Leifer, R., & Huber, G. P. (1977). Relations among per-ceived environmental uncertainty, organization struc-ture, and boundary spanning behavior. Administra-tive Science Quarterly, 22 235-247.

Overton, P., Schneck, R., do Hazlett, C. B. (1977). Anempirical study of technology of nursing subunits.Administrative Science Quarterly, ag, 203-219.

53 59

GRADE INFLATION: A:STUDY WITHIN,ONE ASSOCIATE.DEGREE NURSING PROGRAM

Betty L. Battenfield, RN, EdDProfessor and- Associate Dean

Marianne Neighbors, RN, MEdAssistant Professor ,

Associate Degree Nursing ProgramUniversity of Arkansas, Fayetteville

This study sought to determine if there is gradeinflation in one associate degree nursing program, and ifthere is, to compare it to grade inflation found oncampus in the College of Arts and Sciences. It has beenfrequently reported in the literature that academicgrade inflation has occurred in the higher educationclassroom during the past two decades (Dietman, 1981;Eldridge, 1981; Ferritor, 1981; Handleman, 1980; Juola,1979; Prather, 1979; Suslow, 1977).

The study used three time-line groups of grades andtwo divisions of courses. The three time-lines (T-L)established were:

T-L Group I: 1971-1973T-L Group II: 1974-1977T-L Group Ill: 1978-1981

The division of courses were: nursing (individual coursesand total courses) and, arts and sciences (total courses,science courses, and upper division courses). Comparisonwas made by individual courses, grouped courses and T-Lgroups. Variables examined were:

1. Number and percentages of grades in nursingcourses by T-L group and by course.

2. Percentage of grades given in each nursingcourse (all T-1. groups).

6055

3. Total number and percentage of nursing gradesby T-L groups.

4. Grade point Overage by T-L .group and nursingcourse.

5. Grade point average by T-L group and non-nursing courses.

To compare the Associate Degree Nursing Programgrade inflation with the College of Arts and Sciencesgrade inflation, mean grades on nursing courses and non-nursing courses by T-L groups were utilized. Variables.examined were:

6. Grade point average of non-nursing courses by T-L group.

7. Pre-nursing and graduate grade point average byT-L groups.

g. Selected general education courses grade pointaverage by T-Lgroup.

9. Nursing student T-L groups by: (a) non-nursinggrade point average; (b) nursing grade pointaverage; (c) College of Arts and Sciences:- allstudent grade point average; (d) College of Artsand Sciences: upper division grade pointaverage.

A comparison was also made of:

10. American College Testing Assessment (ACT)(Lindquist, McCarron, & Tyler, 1959) compositescores by T-L groups of all entering universitystudents and associate degree nursing students.

The population was the 556 graduates from theAssociate Degree Nursing Program from 1972 through

6 1, "

1931. The Cci liege of Arts and Sciences study used asampling from three T-L groups to examine three cate-gories of courses. The results were accepted as repre-sentative of the arts and sciences- student population andcompared with data from the total population of nursinstudents.

The number and percentages of grades (A, B,\D, F, WP) for each nursing course in the three T-L

groups were calculated. From the data the followingcharacteristics emerged:

As students progressed through the nursing se-quence there were more Bs in all semesters; moreCs than As in the first semester; more As than Csin the fourth semester. The second semestertended to' have a higher number of As and Bs thanthe third semester.

Frogiessively through the T-L groups nursingcourses had more As, more Bs, and -fewer Cs.

T-L Group I had the lowest nursing grade pointaverage of 2.70, T-L Group II had an intermediategrade point average of 2;35, and T-L Group III hadthe highest grade point average of 2.94.

In non-nursing courses T-L Group I had the lowestgrade point average, and X-L. Groups II and IIIwere progressively higher. Nursing course gradesincreased, by a total of .24 grade points; non-nursing course grades increased by a total of .15grade points.

Graduate grade point averages of all three T-Lgroups showed an increase over pre-nursing gradepoint average by group.

Comparison was made of sample data from theCollege of Arts and Sciences study using grade point

" 62

averages of all students, grade point averages of upperdivision students, grade point averages of pre-nursingstudents, and grade point averages of nursing students.The results were as follows:

Grade point aveiages in all categories increasedprogressively by group.

Non-nursing courses taken by nursing studentsshowed the least increase, probably due to theheavy science component which also showed lessinflation in the arts and sciences study.

College of Arts and Sciences grade point averageof all students showed the greatest increase.

Comparison of ACT composite scores of T-L Groupr, !!and III (ACT was not used during the Group I T-L) showeda drop of nursing students' ACT composite scores fromabove the total university average to below theuniversity average.

'Summary,

The study revealed that from the first group tothe last T-L group there was consistent moderate nursingclassroom grade inflation. Throughout the T-Ls there wasconsistent moderate inflation In every nursing course.Within each T-L there was grade ihflatitIn from the firstthrough the last nursing course with the exception of agrade point drop in the third semester nursing course.Nursing grade inflation tended to parallel the College ofArts and Sciences grade inflation.

Further studies are needed to determine whethernursing course grade inflation is present throughout thenation. If this is true, research should be conduCted todetermine if there is inflation of the national licensureexamination, scores.

63 58

References

Dietman, M. (1981, November). Truth or consequences.Exchange.

Eldridge, R. (1981). Grading in the 70's: How we changed., College English, 43, 64-68.

Ferritor, D. E. (1981). (Study of grade inflation in theCollege of Arts and Sciences. Unpublished manu-script. Fayetteville: University of Arkansas.

Randleman, C. (1980)a, Teaching and academic stan-dards today. Community College Review, z, 40-46.

Juo la, A. (1979). Grade inflation in higher educa-tion-079, #88. East Lansing: Michigan StateUniversity, ,Learning and Evaluation Service.

Lindquist, E. F., Mc Carrell, T. H., & Tyler, R. W. (1959).American College Testing Assessment.

Prather, J. E., Smith, G., & Kodras, J. (1979). A longi-tudinal study of grades in 144 undergraduate courses.Research in Higher Education, 10, 11-24.

Sus low, S. (1977, March). Grade inflation: End of atrend? Change, 44-45.

64

.59

A SURVEY TO DETERMINE WHATNUTRITIONAL INFORMATION MIDDLE-CLASS

WOMEN RECEINE DURING PREGNANCY

Marilyn. Bayne, RN, MSAssistant Professor

Phyllis, L. Neff, RN, MSInstructor

School of NursingUniversity of Maryland at Baltimore

This study investigated the type of .nutritional infor-mation. and methods being used to counsel middle-classpregnant women. Generally, clinic patients attend spe-cial classes in which nutritional information is provided.A review of the literature revealed that middle and

. upper socioeconomic groups of women may not be re-ceiving adequate nutritional counseling (Feigenberg, &Schiller, 1977). Concern about excessive weight gainmay lead some women to severely restrict food intakeduring pregnancy at a time of rapid fetal growth andincreased maternal and fetal energy requirements(Shearer, 1980). Since there is substantial evidenceindicating that there is a relationship between maternaldietary habits and pregnancy outcome (Winick, 1969), thepurpose of this study was to examine the methods ofnutritional counseling, the time during pregnancy whencounseling, was initiated, and the effectiveness of dietarypractices as Measured' by maternal weight gain and birthweight.

The sample consisted of 155 private maternity pa-tients who were studied within 2-6 days followingdelivery in two suburban hospitals in the Baltimore-Washington area. _ Data were obtained-by administering aquestionnaire consisting of two parts. Part I concernedgeneral demographic information and information abouttheir pregnancies, such as weight gain, medical problemsand type, as well as methods and sources of nutritional

6 1

counseling. Part II addressed dietary practices duringpregnancy, knowledge of I optimal nutritional consump-tion, Instruction received aboutli recommended nutri-tional consumption, and in, s truction received about rec-ommended nutritional reqtiirements during pregnancy.

Frequency distributio4 and summary statistics on thedemographic data and other variables on Part I of thequestionnaire were obtained. Pearson correlations as-sessed patterns of relationships. Chi-square andCramer's V were used to' assess relationships betweennutritional practice, knowledge, and Informationreceived.

Analyses revealed thatl the majority of respondents(139) indicated they did receive information about nutri-tion in pregnancy from their physicians. Most respon-dents received information, during the first month (42.6percent) and the second month (27.1 percent) of preg-nancy. The average weight's gain of the sample was 31.2pounds; the average birth weight was 7.6 pounds. Thedata revealed a correlation of .17, (Q = .Q3) betweenWeight gain of the mother ,and weight, of the newborn.There was also a significant correlation between totalweight gained during pregnancy and weight gain rec-ommended by physicians (r .30, 2. = .01). A series ofchi-square computations employed to test relationshipsbetween practices, knowledge, and information revealedthat, on the average, subjects followed directions onnumber of servings within each food category when giventhe information.

Some implications from this study are that nursesshould be alert to the importance of nutritional counsel-ing early in pregnancy and should indicate this to clientsand physicians. The physician's nurse or nutritionistcould assume the role of patient educator during earlypregnancy. Childbirth education classes should be ex-panded so that early pregnancy classes could be con-ducted with emphasis on nutrition.

626 6

References

Feigenberg, M., & Schiller, R. (1977, November/Decem-ber). Nutritional counseling for middle-class gravidas.Journal of Obstetrical and Gynecological Nursing,

19-22.

Shearer, M. W. (1980). Recommended dietary allowances(9th ed.).. Washington, D. C.: The Food and Nutrition.Board of the National Academy of Sciences.

Winick, M. 0969). Malnutrition and brain development.The Journal of Pediatrics, 74, 667-679.

63

67

. STRESS, TRUST, AN MISTRUST tki NORMOTENIVE,HYPERTENSIVE, AND HEART DISEASE PERSONS

IN AN-EXERCISE PROGRAM

Margaret T. Beard,,RN, PhDAssociate Piofessor

Anne Gudnuiridsen, RN, PhDProfessor and-Dean.

Margie N. Johnson, RN, PhDAssociate Professor

. College of NursingTexas Woman's University, Denton

The concept of "stress" has come into increasedprominence since World War II. The term is broad andcovers phenomena at the physiological ($elye, 1956),psychological (Lazarus, 1966), and social (Smelser, 1963)levels of analysis which may be described in terms ofetiological intervening processes and effects. One of theeffects of stress is coronary heart disease.

Coronary heart disease is one of the most seriousthreats to the nation's health as well as to its economictoll. A precursor to heart disease is hypertension. Theworld created by the Industrial Revolution necessitates

, reliance on others for many needs. Many decisionsinvolve the necessity to trust others which is a potential

i source of stress. The Industrial Revolution also createda world in which it was potsible to lead sedentary lives.Coronary heart disease, hypertension, and anxietyt, areamong the problems arising from sedentary life styles.Only recently has the realization that exercise is nec-essary to maintain a high level of, health become evident.The purpose of this investigation was to determine, if arelationship exists between non-specific stress, trust,mistrust, and selected demographic variables in nor-motensive, hypertensive, and heart disease persons par-ticipating in an exercise program.

A non - probability sample* of 64 volunteers was usedwhich consisted of 20 subjects with heart problems, 18with hypertension, 22 with normal blood pressures, and 4with both heart problems and hypertension. There were30 males and 34 females in the sample.

The design was ex post facto. Instruments used werethe Psychiatric Epidemiology Research. Interyiew (PERI)developed by Dohrenwend and co-workers (1980) andRotter's (1967) Interpersonal Trust Scale (ITS). The PERIhas .90 reliability; assessments developed by Vernon andRoberts (1981) were used for additional reliability andvalidity determinations. The ITS is a 40-item additivescale which utilizes a Likert-type format and has a test,retest reliability of .68. The ITS has been validated inlaboratory settings, with questionnaires, self-reports,and peer ratings (Wright and Maggied, 1975).

Participants signed in each day as they arrived . forexercise in a university gymnasium. 'Th_ere was a 10-minute warm-up period. Blood pressure and heart ratewere checked before and immediately after warm-up.Walkers and joggers began in a clockwise direction andchanged to counterclockwise at half-time. Post-exerciseblood pressures were taken at the end of exercise, be-fore cooling down. The period of exercise was onehour. , Subjects completed the PERI and ITS at onesitting. Pearson's correlations were utilized to determinerelationships.

Alpha reliability coefficient for the PERI with, thissample was .83. A statistically significant relationshipwas found between the scores of subjects on the distrustscale and nonspecific psychological distress (stress)162 =.85 (72 percent). A statistically significant relationshipwas found between trust and education 162 =percent), between trust and -retirement 162 =percent), and between heart disease and sex 162=

.36

.34.45

(13(11(20

percent). A statistically significant inverse relationshipwas found between heart disease and education 162 = .36(13 perdent), and between trust and age 1:62 =' -.34 (11percent).

69 66

P

If mistrust is a- source otstress, the efforts to fostertrust should ,be taken. If an exercise program is bene-ficial, the benefits could be evaluated in terms_ ofmedical costs averted, future earnings saved, and in-creased life expectancy. An exercise program as de-scribed with concomitant social support could increaselevels of trust among participants.

References

Dohrenwend, B. P., Shrout, P. E., Egri, G., & Mendel-sohn, F. S. (1980). Non-specific psychological dis-tress and other dimensions of psychopathology.Archives of General Psychiatry, 27, 229-1236.

Lazarus, R. N. (1966). Psychological st ess and thecoping process. New York: McGraw-Hill BookCompany.

Selye, H. (1956). The stress of life. ,Iew York: McGraw-Hill Book Company.

tSmelser, N. J. (1963). Theory of collective behavior.

. New York: Free Press. -

Rotter, J. B. (1967). A new scale foir the measurementof interpersonal trust. Journal df Personality, 21,651-655.

Vernon, S. W., & Roberts, R. E. (1914). Measuring non-specific psychological distress and other dimensionsof psychopathology. Archives of General Psychiatry38 1239-1247. \

Wright, T. L., et Maggied, P. (1975). n unobtrusivestudy of interpersonal trust. Journal of Personalityand Social ilyclllogy,12, 446.

67

PERCEPTION OF NURSES' EXPERT POWERBY NURSES AND PHYSICIANS

Gerry Cadenhead, RN, MSChair, Baccalaureate Program

School of NursingUniversity of Southern Mississippi, Hattiesburg

Nurses make up the largest _group in the health caresystem; however, nurses traditionally have not beenidentified as having power in the system. The lack ofpower nurses hold in the health care system -limits therole nurses have in the system and influences the effec-tiveness of the delivery of health -care.

This study concerned the expert power base of.nursesas perceived by_nurses and physicians. The purposes ofthe study were to: (a) explore the expert power base ofprofessional nurses as perceived by physicians andnurses; (b) determine the relationship of level of nursingeducation on nurses' perceptions of professional _nurses'expert power; and '(c) determine the differences inphysicians' and nurses' perceptions of the expert powerbase of professional nurses. French and Raven (1959)defined expert power as_the special skills and knowledgepossessed by an individual.

The study was descriptive and data were collected bysurveying 200 nurses and 200 physicians randomly se-lected from those licensed to practice in a Southeasternstate. A questionnaire, "Measure of Skill and Knowledgeof Professional Nurses," developed by the researcher,was mailed to the subjects. The instrument measuredperceived skills and knowledge of professional nurses infour areastechnical skills, psychosocial knowledge,communication skills, ana community planning skills. Sixnurses and five physicians participated in a pilot study toestablish content validity for the instrument. Constructvalidity was established for the scales by computing afactor analysis on the study data. Scales were, revisediby

69 71

including items which loaded > .40 after' varimax rota-tion. The revised instrument had 52 _ items-13 fortechnical skills, ,11 for community planning skills, 13 forpsychosocial knowledge, 9 for communication skills, and6 atypical items were included to insure reader relia-bility. The reliabilicies far the scales ranged from r =.87- to r = .92.

Expert power scores were analyzed using a two-tailedt test. Both technical and professional nurses identifiedan expert power base for professional nurses on all fourscales of the instrument; however, professional _nursesperceived they had more expert power in the psycho-social knowledge and communication skills areas than didtechnical nurses (p< = .05).

A significant difference (p .< .05) was found betweenthe perceptions of both nurses and physicians of theexpert power base of professional nurses. Nurses per-ceived that professional nurses had more expert poweron all four scales than did physicians. Physicians per-ceived that professional nurses had expert power in theareas of technical skills, psychosocial knowledge, andcommunication skills.

The perception by nurses and physicians that profes-sional nurses had an expert power base should improvethe utilization and-retention of professional nurses-in thehealth care system. _The role professional nurses have inmaking decisions regarding the health care system shouldincrease. If nurses are perceived as having power, theywill use power and resist power attempts frqm others( Jacobson, 1972).

References

-French, J., & Raven, B. (1959). The social bases ofpower. In Dorwin Cartwright (ed.), Studies in socialpower. Ann Arbor: The University of Michigan.

Jacobson, W. D. (1972). Power and interpersonal relations.Belmont, California: Wadsworth Publishing Company.

70

DIRECT OBSERVATION O1 PARENT-INFANTRESPONSIVENESS IN NURSING RESEARCH:

THE NURSING CHILD ASSESSMENT FEEDING SCALE

Nayna D. Campbell, RN, PhDAssociate Professor, School of Nursing

. The University of North Carolina at Greensboro

Elizabeth A. D. Rankin, RN, PhDAssistant Profesior, School-of Nursing

Karen Soeken, PhDAssistant Professor, School of Nursing

The University of Maryland at Baltimore

In, clinical practice, teaching, and in conducting _re-search, the goal of the maternal-child nurse is to achieveimproved health care for mothers, infants, and children.One assumption for the goal is, the "expressions of healthand illness that can be °operationally defined have iden-tifiable behaviors, lend themselves to modification andmeasurement, (and) form the core of nursing practice"(Nea1,1980).

Nursing and related fields have increasingly rec-ognized the relationship between parent-infant inter-

-----action-and_health. outcomes, especially_for_the child(~Brazelton, Kodowski, do Main, 1974; Klaus dc Kennell,-1976; Barnard Sc Eyres, 1979). Maternal-child cliniciansfrequently come into contact with parents and children

. in the delivery of health care service. An analysis oftheir interactive patterns as part of child health assess-ment would assist in meeting the goal of early iden-tification of problems, when intervention could be mosteffective for the support of child health omcomes.Ideally, such a tool would also be an effective research

One tool that has been developed for use by cliniciansis the Nursing Child Asse7ment Feeding Scale (NCAFS),

71 73.

out of The School of Nursing at the University ofWashington (Barnard, 19781 This is a 78-item binaryscale that defines the interaction based on six con-structs; (a) sensitivity to infant's cues by the parent, (b)parent's response to infant's distress, (c) social-emotionalgrowth fostering, behaviors of the parent, (d) cognitivegrowth fostering_ behaviors of the parent, (e) Clarity_ofcue's given by the infant, and (f) responsiveness of theinfant to the parent. These are measured- by a series ofdichotomous items to determine the presence or absenceof specific behaviors in the interaction.

This tool was developed for use by clinicians doingdirect observation in both in-patient and out-patientsettings. It was used to describe local Washington statefamilies. The sdales showed moderate test-retest reli-ability across ages, likely influenced by developmentalchanges. In examining predictive validity, the NCAFS at12 months was correlated with the 24-month Bayleymental score (r = .55), with =Behar Preschool BehaviorQuestionnaire (r = .61) and with the Caldwell HOMEscore (r = .79) at 36 months, and with the Stanford-Binet(r = .59) at 48 months (King, 1979). A longitudinal studyof 846 parent-infant dyads- found that the NCAFS scoreis a better predictor of developmental outcome thanmother's education (Barnard, 1979A). Utilization of theNCAFS requires a nine-week certification training pro-gram and the establishment-of inter-rater reliability.

The purpose of the present study was to examine theefficiency of this tool in describing another population,and to compare that population with the Barnard sample(1979B). This would further establish the validity of thetool, and its potential usefulness in research as well as inpractice.

Researdhers acquired certification, and an inter-rater reliability of .86 was established ,prior to, and re-evaluated during, observations. The feeding episode was.chosen ,because it is a common, frequent, and essential

. 74 72

caregiving activity, central in early parent-Child inter-action. It is an often observed and easily defined unit,with a beginning and end not controlled by the re-searcher, and is a natural situation in which to _assessattachment and responsiveness.

The 'bottle enviraitheTit, rather than the clinic envi-ronment, was chosen for observations because this iswhere the parent and child normally interact, and it iswhere they have the most control. Data were collectedin 119 home visits to parents with their first-born four-month-old infants. As established by Barnard's guide-lines, researchers observed either or both parents withthe infant during a feeding episode.

Participating parents were predominantly motherswho volunteered to be in the study. Descriptive statis-tics were generated for comparison with Barnard's find-ings. Comparison was made between the feeding scalescores and scores on Caldwell's Home observation forMeasurement, of the Environment Inventory (1978), scoreson the NCAST Teaching Scales, and scores on the DyadicAdjustment Scale (Spanier, 1976). The interactions ofmothers and fathers with infants in the present samplewere also compared. Findings suggest the validity andusefulness of the Nursing Child Assessment Feeding

, Scales both in,practice and research.

References

Barnard, K. E. (Ed.)(1978). The nursing child assessmentfeeding scales. Seattle: University of Washington.

Barnard, K. E. (1979A, August). Memo to certifiedNCAST II learners: Unpublished, University of Wash-ington, Seattle._

Barnard, K. E. (1979B). Summary of the validity ofthe NCAST feeding and teaching scales. Seattle:UniverSity of Washington.

7375

Barnard, K. E., dc Eyres, S. 3. (Eds.) (1979, June). Childhealth assessment part 2: The first year of life.DHEW No. HRA 79-25.

Brazelton, T. B. Koslowski, B., dc Main, M. (1974).The origins of reciprocity: The early mother- infantinteraction. in Lewis, M. dc Rosenblum, L. (Eds.).The effect of the infant on its caregiver. New York:John Wiley and Sons.

Caldwell, B--.1 dc Snyder, C. (1978). Home observationfor measurement of the environment. Seattle: Univer-sity of Washington.

King, D. W. (1979). Psychometric characteristics ofthe nursing child assessment teaching and feedingscales. Unpublished paper, University of Washington,Seattle.

Klaus, M. H., dc Kennell, J. H. (1976). Maternal-infantbonding: The impact of early-separation or losson family development. St. Louis: C. V. MosbyCompany.

Neal, M. V. (1980). Expanding maternal-child profes-sional nurse role, HHS Grant #932-08.

Spanier, G. (1976). Measuring diadic adjustment: Newscales for assessing the quality of marriage and simi-lar dyads. Journal of Marriage and the Family, 28,15-28.

7674

A COMPARISON OF ADOLESCENT MOTHERS'PERCEPTIONS OF INFANTS IN "ROOMING IN"VERSUS TRADITIONAL _HOSPITAL SETTINGS

Verna Carson RN, MSAssistant Professor, gchool of Nursing

Marilyn Winkelstein, RN, MSInstructor, School of Nursing

University of Maryland at Baltimore

The adolescent mother and her baby are at risk for avariety of complicating health conditions before, during,and after delivery (Cran-Elsberry 45c Malley-Corrinet,1979; Merritt, Laurence, dr. Naeye, 1980). Improving theprenatal _health care provided to the adolescent motherhas been the focus of much research, innovative clinicalprograms, and federal monies. Much less effort has beenexpended in identifying ways to assist the adolescent inthe postpartum period.

"Rooming in" is one technique which has beenstrongly supported in the literature as a positive meansof enhancing bonding and mothering skills (Kennell,Jerauld, Wolfe, Chesler, Kreger, McAlpine, Steffa, doKlaus, 1974; O'Conner, Vietze, Sherrod, Sandler, do

Altemeier, 1980). However, research done on "roomingin"' has not controlled for the variable of maternal ageand few studies have compared "rooming in" to tradi-tional methods of postpartum care.

The purpose of this study was to compare adolescentmothers who received one of two different types ofpostpartum care (combined mother/infant care or tradi-tional care with a centralized nursery) in terms of theadolescent mother's (1) perception of her infant, (2)satisfaction with nursing care, and (3) degree of comfortin caring for her infant.

75-

The study was conducted in two different hospitalsettings. One setting utilized a traditional unit, and theother setting utilized a combined mother-infant care unit.In each setting, 32 adolescent mothers between 12 and19 years of age were interviewed at 1-2 days postpartum.Each mother was asked to complete the Neonatal Per-ception Inventory (NPI I) developed by Broussard andHartner (1971). Test re-test reliability of this tool was0.22 for primiparas (Freese & Thoman, 1978). In addi-tion, two tools designed by the study investigators (a"Satisfaction with 'Care" question and a "Degree ofComfort Scale," which measured material, comfort inperforming five mothering tasks) were administered.

At four weeks -postpartum, the mothers completedBroussard and Hartner's Neonatal Perception Inventory(NPI II) and the Degree of Bother Scale (Broussard doHartner, 1971). Test-retest reliability for the NPI II was0.82 for primiparas (Freese & Thoman, 1978).

Summary statistics for the demographic variables forthe total sample of 64 mothers were as follows: meanage of 17.6 years; mean educational level of 10.6 yearsof schooling; 80 percent of the mothers were single orseparated; and 80 percent had an annual income of$6,000 or less. No statistically significant differenceswere found between the rooming in and the non-roomingin group with reTect to any of the demographic var-iables except education and race. Mothers in the non-rooming in group_ had completed more years of schooling(p < .05) and there were more Caucasian adolescents inthe non-rooming in group (p < .05).

There were no significant differences in the totalscores of the two groups on the NPI I and Degree ofComfort Scale administered at 1-2, days postpartum,using t-tests. There were also no significant differencesin the total scores of the two groups on the NP1 IIadministered at four weeks. However, adolescents withrooming in received more teaching regarding cord care

78 76

(p< .01) while in the hospital,.and mothers with roomingin were also more satisfied with their nursing care thanthe mothers without rooming in (p < .05).

At four weeks postpartum, adolescents without room-ing in perceived the average baby as crying more thanthe rooming in group (p < .05). At this time, the non-rooming in group was also more bothered by- their owninfant's crying than the rooming in group (p <.05).

Since an experimental design was not used, causalinferences cannot be readily made. However, the resultsof this descriptive study suggest that rooming in mayinfluence the amount of teaching provided to mothers aswell as the degree of maternal satisfaction with nursingcare. In addition, rooming in may alter perceptions asrooming in mothers were less bothered by their infants'crying at four weeks of age. The investigators believethat the findings of this study suggest that there may beadvantages to the use of rooming in as an interventionfor adolescent mothers.

References

Broussard, E. R., & Hartner, M. S. S. (1971). Furtherconsiderations regarding maternal perception ofthe first born. In J. Hellmuth (Ed.), Exceptionalinfant: Studies in abnormalities (Vol./). New York:.Brunner/Mazel, Inc.

Cran-Elsberry, C., & Malley-Corrinet, A. (1979). Theadolescent parent. In S. H. Johnson (Ed.), High riskparenting. Philadelphia: x.13. Lippincott.

Freese, M. P., & Thoinan, E. B. (1978). The assessmentof maternal characteristics for the study of mother-infant interaction. Infant Behavior and Development,1 95-105.

77 79

Kennel!? J. H., Jerauld, R., Wolfe, H., Chesler, D., Kreger,N. C., McAlpine, W.,..Steffa, M., 8c Klaus, N. H. (1974).Maternal behavior one year after early and extendedpostpartum contact. Developmental Medicine andChild Neurology, 1.6j 172-179.

Merritt, 3. A., Laurence, R. A., Naeye, R. L. (1980).The infants of adolescent mothers. Pediatric Annals,2, 32-33.

O'Connor, S., Vietze, P., Sherrod, K., Sandler, H., dt,,

Altemeier, V. (1980). Reduced incidence of parent-ing inadequacy following rooming in. Pediatrics,66 176-182.

8078

THE EFFECT OF SUPPORT ON LONG-TERMSUCCESS. AT BREAST FEEDING

. -Regina M. Cusson, RN, MS

Instructor of Pediatric NursingSchool of Nursing

Universq of Maryland at Baltimore

Breast feeding is on the rise in this country, with thecurrent rate now at 50 percent (Filer, 1978); this is incontrast with the rate of 18 percent found a generationago (Meyer, 1968). Many new mothers have nowhere toturn for support and encouragement. The emphasis onchildbirth as a healthy process has removed nursingmothers from contact with the.health care system atprecisely the time when they most need assistance withbreast feeding. Breast milk, on the average, begins tocome in on the second to fourth day--the period whenmothers are discharged from the hospital.

The rate of long-term success at breast feedingseldom exceeds 50 percent (Hall, 1978). Notable excep-tions include studies in which an attempt was made toprovide information on breast feeding and support tomothers during hospitalization for delivery and in thepost-partal period following discharge (Peterson et Bock,1978; Hall, 1978). There are few controlled studies toreplicate those findings.

A qualified experienced nurse can provide the kind ofassistance a mother needs to achieve success at breastfeeding.

Prior studies indicate that breast feeding declinesrapidly upon discharge from the hospital (Fomon, 1974;Coles, Cotter, & Valman, 1978; Sloper, McKean, acBaum, 1975). This study tested an intervention designedto prevent this decline.

/ 81

The purposes of this study were (1) to test the effectsof support by the nurse on long-term success at breastfeeding by comparing mothers who received hospitaltreatment (i.e., breast feeding information and support,during hospitalization) with mothers who, in addition tohospital treatment, received follow-up telephone callsabout breast feeding for six weeks after discharge, and(2) to compare the attitudes about breast feeding ofmothers 'who were successful or unsuccessful at. long-term breast feeding.

Subjects were 80 first-time breast feeding mothers-selected from the post-partum unit, of a suburban Bal-timore hospital. All subjects were married; their ages.ranged from 18 to 43 years. The study was conductedover a period of several years, with data from the first20 subjects collected three years ago, and the remainingdata collected latt year.

All subjects received in-hospital teaching and supportconcerning breast feeding. Each mother was seeg dailyuntil discharge to have her questions answered and toprovide her with assistance with breast feedingtechniques.

Prior to discharge, the subject was randomly assignedo to the follow-up or no follow-up group. The follow-up

group received phone calls until six weeks after dis-charge; the no follow-up group did mot., All subjectswere telephoned at six weekt by a research assistant, foran interview about their breast feeding status at thattime.

Each subject signed a consent form, and completed ademographic data form and an Attitudes Toward BreastReding scale. This scale, developed for a preViousstudy, contains 18 Likert-type items plus eight generalquestions concerning the subject's previous exposure tobreast feeding. Split-half reliability (odd- even), cor-rected by use of the Spearman-Brown adjustment, was.83; coefficient alpha was .71.

82: 80

Data were analyzed by means of t-tests and Pearson'scorrelations.

_Hypothesis 1: There is no difference in successbetween subjects who received follow-up for six weeksand' those who did not.

The t-test results were non-significant. The overallsuccess rate was 83 percent, so there were few un-successful subjects with whiLL to compare. Most of thesubjects (69 percent) stated that they received assist-ance and support from other sources.

_Hypothesis 2: Subjects who are still breast feeding atsix weeks have higher attitude scores than do subjectswho have ceased breast feeding by six weeks afterdelivery.

This hypothesis was rejected since the t-test showeda .significant difference. Moreover, successful breastfeeders were more likely to have attended breast feedingclasses, done breast preparation, and planned to breastIfeed for a set period of time.

Hypothesis 3: There is no relationship betweendegree of satisfaction among subjects who are stillbreast feeding at six weeps and those who are not.

Using the Pearson Product Moment Correlation,there was a significant correlation between success andsatisfaction. This null hypothesis was thereforerejected.

Hypothesis 4: There Is no difference in succ "ssbetween subjects who decided to breast feed prior .opregnancy and those who decided during or afterpregnancy.

The results were non-significant, demonstrating that,for this study, the timing of the decision was notsignificant in relation to success.

81

Hypothesis There is no difference in attitudebetween subjects who decided to breast feed prior totpregnancy and those who decided during or afterpregnancy.

Using the t-test, there was no significant difference.

Hypothesis 6: There is no difference in Successbetween subjects with complications and subjects withnormal vaginal deliveries.

Using the t-test, there was no significant. difference.HoWever, there were other differences: , the complica-tions group was older and started breast- feeding later7than the normal group.

'This study presented data about maternal attitudes,and the influence of support and information on breastfeeding success. Probably the most important findingswere the 83 percent success rate, and the mothers'reports that breast feeding support was becoming more-readily available in the community.

References

Coles, E. C., Cotter, S., Sc Valman, H. G. (1978). Increas-ing prevalence of breast feeding. British MedicalJournal, 2, 112.

Filer, L. 1. (1978). Early nutrition: Its long-term role.Hospital Practice, 1.3_, 87-95.

Form), S. J. (1974). Infant nutrition (2nd ed.). Phila-delphia: W.B. Saunders.

Hall, J. M. (1978). Influencing breast feeding success.Journal of Obstetrical and Gynecological Nursing,7 28-32.

84

Meyer, H. F. (1968). Breast feeding in the U. S. Re-__ports of a 1966 national survey with comparable

1946 and 1956 data. Clinical Pediatrics, z, 708-715.

Peterson, J. C., 4c Bock, W. (1978). Educating nursing,mothers. Perinatal. Care, 2 (7), 44-47.

Sloper, K., McKean, L., & Baum, J. D. (1975). Factorsinfluencing breast feeding. Archives of Diseasein Childhood, 50, 165-170.

8583

Q METHODOLOGY IN NURSING RESEARCH:PROMISE AND PROBLEMS

Karen E.Dennis, RN, MSN, PhD CandidateSchool of Nursing,

University of Maryland at Baltimore

Q methodology was founded on the epistemologicalpremise that it is possible to study subjectivity in anobjective, orderly, and scientific manner (Stephenson,1978, 1953). An approach to the study of subjectivity, Qmethodology has commonalities with both quantitative andqualitative methodologies in an intriguing combination ofthe strengths of both research traditions.

While different approaches to the discovery of scien-tific knowledge are valuable, Q methodology, with itsfocus on the individual and on subjectivity, a partic-ularly viable approach that holds a great deal of promisefor nursing research. The uses of Q methodology in-

, clude, but are not limited to, the study of opinions andattitudes, groups, roles, culture, personality, decision-making, and values (Brown, 1980). A few substantiveareas in which nurses might use Q methodology areparental perceptions of and reactions to their newborns,coping behaviors of patients with long-term illness, andbelief systems influencing health-seeking behavior.

This paper presents the promise of Q methodology innursing research through the explication of the techniqueand its many strengths.

Rather than focusing on the interindividual differ-ences of traits within large sample sizes, Q methodologyunderscores intraindividual significance and requires a,limited number of subjects, usually no more than 40. IhQ technique, persons rate a group of 40-60 statementsalong a specified continuum of significance, such asmost-like-me to most-unlike-me. This technique resultsin an array of items called a Q sort. Data obtained from

85 86

Q sorts are factor analy- zed; the factors reveal thedimensions of the phenomenon under investigation aswell as clusters of persons who hold similar viewpoints.

Q methodology is one solution to the problem ofsmall sample sizes which plague many studies withinnursing. Its approach to subjectivity is more penetratingthan other forms of instrumentation and, since thesubject is his/her own point of reference, it is partic-ularly amenable to interorganizational use. Problemswith missing data and response sets are virtually non-existent, and Q studies are, usually not costly to conduct.Data tend to demonstrate high reliability, and are condu-cive to both deductive and inductive interpretations.

Despite its promise for nursing research, Q meth-odology has its own set of problems with which theresearcher must deal. The administration of Q sorts canbe time-consuming, since subjects usually are not al-ready familiar with the procedure to follow. Validity isat issue if subjects are not clear on how to respond.Although specifying the distribution of the Q sort is seenas a strength, since it requires subjects to make finediscriminations among items, subjects may make me-chanical rather than conceptual choices in order tocomplete the process. Patients' limited energy resourcesare particularly important to consider.

Nurse researchers use a variety of methods, such assurveys, experiments, and a 'number of descriptive ap-proaches to explore phenomena from a nursing per-spective. Adding Q methodology to that repertoirewould enhante the nature and the richness of the designalternatives for developing nursing knowledge.

References

Brown, S. R. (1980). Political subjectivity: Applicationsof Q methodology in political science. New Haven:

- Yale University Press. ._

Stephenson, W. (1978). The shame of science. Ethicsin Science and Medicine, .5_, 25-38.-

Stephenson, W. (1953). The study of behavior: 0-tech-nique and its methodology. Chicago: Universityof Chicago Press.

87

4

88

\

POWER SHARING IN UNIVERSITIES:THE CASE OF NURSING

Jacqueline Dienemann,,RN, MSNAssistant Professor

George Mason UniversityFairfax, Virginia

Power involves shaping bureaucratic rules and influ-encing discretion by the administration in organizations.The. Strategic Contingency Theory (Pondy, 1969) focuses onorganizational-environmental interaction and immediacyto workflow as a means of identifying powerful subunits.This study focused on the nursing schools in comparisonto other subunits within their respective institutions.

The subunits that compete for power and resources inuniversities are disciplines organized as professionalschools or academic departments in the Arts and Sci-ences. This study was an extension and replication ofPfeffer and Moore's 1980 causal model of resourceallocation. It identified four power determinants andtested the effect of power and workflow need on theallocation of budget, faculty, teaching support, and fa-cility support. Ten hypotheses were tested with datafrom 582 subunits in 88 universities. Subunits werechosen using a typology of professional schools developedfor this study and Salancik, Staw, and Pondy's (1980)measure of the paradigm development of disciplines.The national sample of universities include four typesidentified by the Carnegie Commission on Higher Educa-tion: elite (35), doctoral (21), comprehensive (17), anhealthsciences (15),

The national sample of universities was chosen toinclude all 1980 National League for Nursing-accreditedgraduate programs in schools of nursing. Nursing schoolsare the only subunits on university campuses in whichfaculty and administration are essentially only women.

0

89 89

This study examined faculty gender as a power deter-.

minant. It also compared and contrasted nursing schoolsto other professionAtschools of education, library science,social work, medicine, pharmacy, engineering, and busi-ness administration, as well as the academic disciplinesof chemistry and sociology, on characteristics-of hori-zontal power identified in the Power Sharing Model andshares of resource allocations across university types.

Data were collected from the National Center forEducation Statistics, American Council on Education,and 2700 mailed questionnaires. Questionnaires, sent todeans and department chairs, asked them to rate thepower of subunits using Arnold Tannenbaum's (1968)control in organizations questionnaire and a reputationalmeasure of power.. Supplemental questionnaires seeking

.data not available frcm the National Center for Educa-tion Statistics and the American Council on Educationwere also sent to institutional research officers anddeans and chairs of subunits being examined. Theresponse rate of usable forms was 42 percent for insti-tutional research officers and 58 percent for deans andchairs.

Path analysis supported the extension of the model ofresource allocation, titled the Power Sharing Model.Multiple regression analysis was used to test the modelacross university types, professional school types, anddisciplines. Testing homogeneity of types supported thetypology of professional schools. Results supported theaddition of sex mix as a predictor of power.

-

Based on the analysis, and theoretical framework, itwas recommended that nursing school administratorsurge faculty to apply for research grants, increase thevisibility of their school's contrihutions to universitycritical resources, seek strategic placement of tacultyon key university committees, and creatively reducefaculty teaching ,loads to enhance the power and re-source allocations of schools of nursing on- all types of

campuses. The nursing profession needs to develop aresearch career path, perceive women's issues as nursingissues, and continue to support theory development andhigher education for nurses.

References - 6i

Pfeffer, J., ae Moore; W. (1980). Power in universitybudgeting: A replication and extension. Administra-tive science- quarterly, 21, 637-53.

Pondy, L. R. (1969). Varieties of organizational conflict.Administrative science quarterly, 14, 499-506.

Salancik, G., Staw, EC., ec Pondy, L. (1980). Administra-tive turnover as a response to unmanaged organi-zational interdependence. Academy of ManagementJournal, 23 422-37.

Tannenbaum, A. (1968). Control in organizations. NewYork: McGraw-Hill.

CANCER AND NUTRITION: DERIVING A THEORYOF ADAPTATION FROM &NURSING EXPERIMENT

, Jane Dixon, PhDAssodiate Professor

Yale University Scho Ol of NursingNew Haven, Connecticut_

John Dixon, PhD- Director

Adaptation Research Institute.Branford, Connecticut

Cancer is often accompanied by extreme weight lossand nutritional deficiency. A longitudinal, clinical ex-perimentperiment was conducted to evaluate the effectiveness oftwo nursing intervention protocols, used singly and incombination, on nutritional status cid performance sta-

. tus of persons' with cancer.

Eighty-eight cancer patients, assessed as nutri-tionally at-risk, were followed for up to 16 months.Subjects were randomly assigned to a control group or toone of four intervention groups: nutritional supple-mentation; relaxation training; both nutritional 'supple-

and relaxation training; or nurse's visits notinvolving either nutritional supplementation of relax-ation training. Subjects in any intervention group werevisited at home every other week for the first fourmonths. ThoSe subjects, as well as control subjects;---received, follow-up visits at 4, 7, 10,and 16 months.

Nutritional changes, and patterns` of activity wereasst....ied. Initial statistical equivalence between ran-domly assigned experimental groups was established. Inrepeated measures analysis of variance, significant groupby time interactions were obtained for weight (as per-centage of ideal weight) and for arm rritiscle circum-ference (a meaw.e Of protein status) indicating that forthese variables, the groups changed differentially over

93 92

time during the intervention period. With the Karnofskyscale (Karnofsky do Burchenal, 1949) measure of per-_

formance status as dependent variable, the group. bytime interaction approached significance ( p = .08). Forall three variables, gain was greatest for the relaxationgroup, and most severe loss occurred in the controlgroup. Relaxation, combined with nursing visits, seemedto lead to improved nutrition for this sample of poten-tially malnourished cancer patients, while dietary sup-plements, supplied by the program, did not. This may bedue to the fact that supplements were often stigmatized,and, on occasion, called "cancer food."

Collection of extensive personal information on thesepatients led to observations on the connection betweenpersonality styles and response to illness. The factors ofactivity and flexibility appeared to combine in shapingeach person's unique reactions to cancer and ability torespond creatively to a health. crisis. Development ofassessment and intervention approaches related to suchbehavioral manifestations of Creative adaptation wouldbe a fruitful area for nursing research.

References ,

Karnofsky, D. A., (3c Burdienal, 1. H. (149). The clinicalevaluation of chemotherapeutic agents in cancer.In C. M. MacLeod (Ed.), Evaluation of chemothera-peAcagents. New York: Columbia UniversityPress, 191205. .

93

94

THE EFFECT OF A DEATH EDUCATIONWORKSHOP ON NURSE ATTITUDES AND BEHAVIORS

PHASE I: DEVELOPMENT OF AN OBSERVATIONALTOOL TO DOCUMENT NURSE BEHAVIOR

WITH TERMINALLY ILL PATIENTS

Judy A. Donovan, RN, MSNurse Coordinator, Palliative Care

University of Pennsylvania Cancer Center, Philadelphia

Marianne L. Dietrick-Gallagher, RN, MSNStaff Developinent Instructor

University of Pennsylvania Hospital, Philadelphia

Providing direct patient care to terminally ill pa-tients has been identified as a source of stress to thenurse (Quint, 1966; Kubler-Ross, 1969). This supposition,coupled with an identified need of the staff nurses on a25-bed gynecologiCal oncology unit for information aboutdealing with terminally ill patients, Was the impetus todesign a workshop for staff addressing death and dying.The purpose of the workshop was to familiarize nurseswith (a) the psychological experience of-the dying per-son; (b) sociological perspectives on death; (c) ethicalissues surrounding death and dying; and (d) grief counsel-ing techniques for the nurse. The investigators' premisewas that education about death is a developmentalprocess that transmits valid death-related knowledge andits implications for subsequent change in attitudes andbehaviors (Seidel, 1981). While it is inappropriate to tryto restructure students' attitudes, educators are entitledto demand that health professionals become more awareof the strength of their own attitudes and how suchattitudes may affect their behavior towards patients andinfluence clinical and ethical decisions (SimpsOn, 1979).It was hypothesized that nurses would: (1) _increase thenumber of behaviors directed toward non-physical needs;

95 9 4

(2) increase the number of open, verbal responses; (3)increase the frequency of eye contact and touch; (4)decrease their proximity to the patient; and (5) evidencea .change in attitude.

A literature review suggested attitudinal change ofvarying duration after death education programs(Hopping, 1977; Murray, 1974; Miles, 1980; McDonald;1981), but provided little evidence to support behavioralchange on the part of the nurse in nurse/patient inter-

- actions. The Winget questionnaire "Tool for Under-standing the Dying" (Ward (3c Lindeman, 1979) was chosento measure attitudes; however, no tool to documentbehavioral charige was available. It became apparentthat the overall study would be comprised of a number ofphases. The purpose of Phase I was to design anobservational toolito measure behavioral change.

Initially, a hiOly structured behavior checklist, basedupon experiential knowledge and literature review wasdeveloped. Behaviors included interactions addressingnon-physical needs, frequency of closed versus openverbal responses, and frequency of eye contact, type oftouch, and proxiMity to the patient. The tool contentwas validated by a panel of four master's and doctorallyprepared nurses with expertise in psychiatric nursingand/or caring fort terminal'y ill patients.

Inter-rater re,liability was tested on a separate unitwith a high proiDortion of terminally ill patients. Aconvenience sample of 11 nurses was observed con-

- currently by the two co-investigators during nurse/patient interactions. Each nurse was followed by two_observers for six, randomly chosen 10-minute intervals.Inter-rater reliall,ility. was 90 percent within the ',non-verbal section$ tut was virtually nonexistent fqr theverbal behaviors. There were three hypothesized reasonsfor this disparity!: (1) The verbal behavior categories'

s, were not mutually exclusive. (The observers saw thesame behaviors, nterpr ed them in a similar manner,

955 96

but documented them differently on the tool.) (2) In-vestigators interpreted nurse/patient verbal interactionsdifferently at times. (Verbal behaviors must be viewedwith consideration given to the context in which theyoccur. The tool was, designed to document frequency ofopen verbal responses versus closed verbal responses;however, situational variables determined whether aresponse was open or closed more often than did thecontent of the verbal response, Therefore, consistentdocumentation of a nurse's verbal behavior could not beaccomplished utilizing this toll.) (3) The tool wascumbersome. (The number of categories was too largeto permit accurate documentatio)

The investigators then' restructured the tool to re-flect1 ,betaViors which were considered clinically impor-

tant ,an which had a high degree of inter-rater reli-abilityt. Based on a literature review, and previous pilotjtesting`cthe following behaviors were 'selected as appro-priate* evaluate: eye contact, touch, proximity to thepatient; and nurse initiation of discussions of psycho-social concerns. A semi-structured approach was se-lected to measure nurse initiation of discussion of psy-chosocial concerns and open responses to patient initi-ation of the same. The four questions addressing thesebehaviors were answered by the observer immediatelyafter the observation period.

Instrument testing was performed on a 40-bed medi-cal unit which housesia high proportion of terminally illpatients. A total of 15 nurses were observed for two-hour time intervals simultaneously by the two investi-gators. Inter-rater reliability ranged from .75Sitem 1)to 1.00 (item, 7). The lower reliability coefficient for,item one (semi-structured) is related to differing inter-pretation by observers of the operational definition ofpsychosocial/ need. Analysis of descriptive responses toitems 1 through 4 indicates that all discussion of psycho-social need involved information exchange. No nurse/pa-tient interactions addressed the patients' needs for hope,emotional 'expression, or concern% related to death anddying (all of which are included in the definition ofpsychosocial need).

97A.: ' 96

.Touch. behaviors occurred in 55 percent, of all en.,counters; only 5 percent were not task oriented. Nursesmaintained eye contact throughout most interactions;however, in two interactions nurses avoided eye contactwhen the patient posed a difficult question.

While the many problems inherent in observationcannot be overcome, the use of both highly structuredand semi-structured sections of the tool reduces a num-ber of the limitations. The format of the non-verbalsection reduces the halo, leniency, severity, and centraltendency effects. The Verbal response questions permitthe observers to consider contextual components. Anec-dotes documented in response to the verbal behaviorquestions will be used to clarify nurse behaviors forfurther study.

,.A number of additional considerations for obser-

vational experiences arose during the study. (I) Separat-ing physically oriented behaviors from psychosociallyoriented behaviors was difficult. Nurses address bothsimultaneously. Without a clear definition of psycho-

-. social need, delineation of behavior is impossible.During initial pilot testing, psychosocial heed was re-fined based on the work of Young-Brockopp (1982). (2) Instructured tools, the number of categories must belimiteda tool with more than 15 categories is difficultto- use. (3) Our experience has indicated that observa-tions are most fruitful at certain times of the day. Forthe purpose of documenting nurse/dying patient inter-actions, the ideal time was mid-morning when nurseswere involved in direct care with a limited number ofpatients. Observations at other times of the day yieldeda low percentage of the desired behaviors. (4). Observerlocation was of paramount importance. If positioningwas not ideal, facial expressions, eye contact, and touchwere missed or misinterpreted. When observing beha-vior, one must clarify the observer's position in refer-ence to the nurse and patient. (5) Some ethical questionsarose during the study. What should the observer dowhen the patient and nurse are behind closed curtains?

What response does the observer make when the patientinitiates a discussion of psychosocial issues with theobserier?

The Observation Tool to Document Nurse Behaviorwith Terminally 111 Patients developed in Phase I will bevaluable to other investigators hoping to studynurse/patient interaction as it is grounded in the experi-ential component of nursing. Additionally, the knowl-edge gained and shared by the investigators about obser-vational skills and tool development may be helpful forother nurse researchers.

References

Hopping, B. L. (1977). Nursing students' attitudes towarddeath. Nursing Research, 26, 443-447.

Kubler-Ross, E. (1969). On death and dying. New York:MacMillan Co.

Laube, J. (1977). Death and dying workshop for nurses:Its effect on their death anxiety level. International

-Journal of Nursing Studies, 11 ji 111-120.

McDonald, R. T. (1981). The effect of death educationon specific-attitudes toward death in college students.Death Education, 1, 59-65.

Miles, M. S. (1980). The effects of a course on deathand grief on nurses' attitudes toward dying patientsand death. Death Education, 4 245-260.

Murray, P. (1974). Death education and its effects ondeath anxiety of nurses. Psychological Reports,2.2, 1250.

Quint, J. C. (1966). Awareness of death and the nurse's-.composure. Nursing Research, 1.5.1 4945.

99 98

Seidel, M. A. (1981).- Death education: A continuingprocess for nurses. Topics in Clinical Nursing, a,87-96.

Simpson, M. A. (1979). Death education: Where isthy sting? Death Education, 3 165-174.

Ward, M. 3., dc Lindeman, C. A. (Eds.) (1979). Instrumentsfor measuring nursing practice and other healthcare variables. Hyattsville, Maryland: U.S. Departmentof Health, Education, and Welfate.

Young-Brockopp, D. (1982). Cancer patients' perceptionsof five psychosocial needs. Oncology Nursing Forum,9(4), 31-35.

99100

TOP ACADEMIC NURSE ADMINISTRATORS'PERCEPTIONS OF ESSENTIAL BEHAVIORS

FOR THEIR POSITIONSI

Shirley A. George, RN, EdDAssociate Professor

School of NursingMedical College of Georgia, Augusta

Carol Deets, RN, EdDProfessor, School of- Nursing

Indiana University,. Indianapolis

The- purposes of this exploratory study were to iden-tify behaviors that are essential to the successful admin-istration, of a university nursing education unit as per-ceived by current top-level nurse executives, and toidentify whether certain institutional and educationalvariables are related to selected role behaviors.

A complete, up-to-date listing of baccalaureate andhigher degree programs throughout the country, whichincluded deans, chairpersons, directors, or heads of the'nursing education unit was obtained from the AmericanAssociation of Colleges of Nursing. Of the 374 potential'participants,- 60 percent (224) replied to the surveyquestionnaire. The questionnaire contained 53 behaviorswhich were identified from a literature review, indi-vidual and group discussions with several top nurseadministrators, and an in-depth discussion with a nurseresearcher engaged in decanal research. Additionalitems measured five institutional variables and two edu-cational variables.

The majority of the respondents (78 percent) haddoctoral degrees. Most reported to vice-presidents (45percent) or deans of disciplines other than nursing (37percent). One half of the respondents had been in theirpositions less than five years.

foi loo

From the 53 role-related behaviors, the sample iden-tified 18 essential (80 percent or more agreement) and17 non-essential behaviors (40 percent or more agree-ment). Essential behaviors were specific administrativeand managerial transactions in the educational unit andthe greater university, and activities that conveyed theunit's mission to significanrgroups. Non-essential be-haviors included policy negotiations between nursingservice and education, professorial role functions, andinterdisciplinary roles. Chi-square analysis identified asignificant relationship between title, enrollment, andadministrative structure and five behaviors: (a) developa coalition to meet the goals of the proiession, (b)establish reciprocal mechanisms with nursing serviceadministrators for jointly appointing faculty and nursingservice staff, (c) practice in their area of expertise, (d)develop strategies to meet _conflict, and (e) ensurebuilding maintenance and improvement.

The study provides clarification of the top nursingrole in academic administration and can be used by nurseexecutives to reevaluate their perceptions of essentialand non-essential behaviors. The findings also can beused by educators to plan programs that meet the needsof potential candidates for the position. In addition,potential candidates can benefit from the findings byplanning experiences which foster development of theessential behaviors.

101102

41

.

A CASE STUDY OF THE IMPLEMENTATION OFLEGISLATION PROVIDING DIRECT THIRD PARTY

REIMBURSEMENT FOR THE SERVICES OF CERTIFIEDNURSE PRACTITIONERS IN THE STATE OF MARYLAND''

Hurdis Griffith, RN, MSN, PhD CandidateSchool-of Nursing

University of Maryland at Baltimore

A Maryland , state law which mandated direct third.."-party reimbursement for the services of nurse prac-

titioners without direct physician supervision was en-acted in 1979. Maryland was one of the first states topass this type of legislation. There was no empiricalevidence available that addressed the extent to whichthe law had been implemented or the barriers which hadbeen encountered in the implementation process. Thepurpose of this study was to explore this implementation_process.

The research design was, a case study using a modi-fication of the Nakamura and Smallwood (1980) theory ofimplenientation as the theoretical framework. Thistheory views the implementation process as three "envi-ronments" which are connected by "linkages." The threeenvironments are: policy formation, policy irnple-Mentation, and policy evaluation.

Instruments included: (1) an open-ended interviewschedule which was modified for use with groups of nursepractitioners, insurers, legislators, physicians, midwives,and psychiatric/mental, health nurse specialists; (2).'aclosed-ended questionnaire which was mailed to- all cer-tified nurse practitioners with written egreements inMaryland. The interview schedules consisted of part I,which addressed broad-based economic and policy ques-tions, and part II, which was specific to the practicerole of the respondent being interviewed. The question-naire addressed the demographic, practice, and reim-bursement characteristics of the nurse practitioners.

103

102

Data collection was conducted in two phases; theexploratory phase and the descriptive phase. The sam-pling process and instrumentation varied within thesephases. In phase I, the exploratory phase, a purposivesample of 13 nurse practitioners, 11 insurers, 13 legis-lators, 10 physicians, 5-midwives, and 6 psychiatric nursespecialists were interviewed. Using official records, andupon the. recommendation of colleagues, respondents'were selected on the basis of their. inv-olvemcnt in theenactment or implementation of the legislation or theirknowledge of the topic. The sampling principle in theinitial phase- was on "richness," that is, how rich or-meaningful the data would tbe when supplied by thesample chosen (Selltiz, Wrightsman, dc Cook, 1976). Theusual concern about repre&entatiyeness of the sample forpurposes of generalizing to the larger population did notapply here. Open-ended interviews ranging from 30minutes to 3 hours in length were conducted, tape-recorded, and transcribed.

In the analysis of exploratory phase data, the re-sponses were independently placed into categories bytwo researchers. An average of the estimates of theinter-rater reliabilities on the independent placement ofthe responses into the categories on 20 randomly se-lected questions was 91.83 percent. Frequencies andpercentage tables were used to present the data.

In phase II, the descriptive phase, a questionnaire wasdeveloped using the categories of responses determinedduring phase I. It was pilot-tested on 10 nurse prac-titioners without written agreement in Maryland andrevisions were made on content and format. It, wasmailed in July 1983, to 211 nurse practitioners certifiedin the state of Maryland with written agreements ap-proved by a joint committee of the Board of Examinersof Nurses and the Board of Medical Examiners. Therewas a 74.3 percent return rate; 148 of the questionnairesmet the criteria for the study. Fifty-two nurse prac-titioners with written agreements in Maryland did not

respond to the study. Using a Chi-square procedure, itwas determined that there was not a significant dif-ference between the respondents and the nonrespondentson the following variables: level of education, type ofnurse practitioner preparation, age range, and type of,nurse practitioner certification. Inter-ra,ter reliabilityfor the coding of the responses, using a random sample?of the questionnaires, was 9933 percent. Frequenciesand percentage tables were used in descriptive phasedata.

Based on findings from the descriptive phase, nursepractitioners in Maryland perceived that the three great-est barriers to the implementation of the legislationwere: resistance by physicians, resistance by insurers,and the salaried status of nurse practitioners. Otherreasons for lack of implementation included: most ofthe nurse practitioners had not applied for providernumbers, the majority perceived that the law has had noeffect on them, few are reimbursed on a fee-for-servicebasis, and most of the third party claims are beingsubmitted by employers or institutions. Regarding pay-ment of fees for the services delivered by nurse prac-titioners, the majority of their patient contacts arecovered by insurers, few nurse practitioners sign theinsurance claim form for the services they provide, andthe charges for the services provided by nurse prac-titioners are highly correlated with the charges for theservices provided by physicians.

Implications for nursing are that prior to the enact-ment of state legislation providing direct third partyreimbursement for the services of nurse practitioners, itwould be advantageous to assess the reimbursementneeds and the most appropriate reimbursement mo-dalities for nurse practitioners in that state. Policydecisions regarding the reimbursement of nurse prac-titioner services must address the discrepancy betweenthe insurers' interpretation of services and the nurse

105 104

practitioners' perceptions of services offered. The in-surance package and marketability are crucial to theimplementation process; a mandated offering proilsioncould render nurse practitioner 'reimbursement legisla-tion unimplementable if unreasonable premiums are

/charged and/or it is marketed poorly. In making policydecisions regarding the rates for nurse practitioner serv-ices, strong consideration should be given by insurers toways of implementing the fixed-rate alternative that isclearly preferred by nurse practitioners. _If more atten-tion is not given to the implementation phase of directthird party reimbursement for nurse practitioners, theefforts . to provide that reimbursement may not berealized. ;-

References

Nakamura, R., & Smallwood, F. (1980). The politicsof policy implementation. New York: St. Martin'sPress.'

Selltiz, C., Wrightsman, W., & Cook, S. (1976). Researchmethods in social relations (3rd ed.). NeW York:Holt, Rinehart and Winston.

105106

COLLABORATION IN CONTINUING-EDUCATION FOR NURSING: TRAINING

COMMUNITY HEALTH NURSES TO CARE FORDISABLED' CHILDREN AN9 THEIR FAMILIES

Carolyn 'Huard, RN, PhDProgram Evaluator

Connie Petit, RN, MPHProgram Director

Marjorie North, RN, MSN-.Clinical Instructor"

Community Health Nurse Training ProjectChildren's Hospital Medical Center

Washington, D. C.

Melissa Van -Wey, MSNStaff Member

Jennie Austin, RN, MSNStaff Member

Georgetown University Child Development CenterWashington, D. C.

This study was an evaluation' of the CommunityHealth Nurse Training Project,* a collaborative cor(tinu-ing education project for nurses in the care of disabledchildren and their families, sponsored by Children's Hos-pital National Medical Center and Georgetown Univer-sity Child Development Center. The collaborative as--pects of the project allowed both Institutions to drawtogether an interdisciplinary team of 20 experts whoparticipated in the program. The program was unique-Inthat it included both 60 hours of didactic experience and60 hours of supervised practicum experience applying the

*Community Health Nurse Training project (NU23063), Department of Health and Human Services, Divi-sion of Nursing, Special Projects Branch.

107 106-

learned skills in the home setting under the supervisionof nurse preceptors. The program was presented a totalof four times during the funded period of 1981 to 1984.Results for two groups of participant nurses arereported. .

The study's purpose was to docUments-cognitive, af-fective, perforniance, and role changes associated withparticipation in a prototype continuing education projectfor nurses. Subjects were Registered Nurses caring, fordisabled children in. metropolitan Washington, DI C.community health agencies who volunteered for partic-ipation in the program. The study included pre-posttesting of 21 nurses comprising Group 2 participants, andan 8-month follow-up of 19 nurses, comprising Group 1.Affective assessment instruments included the AttitudesTowards Disabled Children (ATDC) scale adapted fromthe Attitudes Towards Disabled Persons scale (Yuker,Block, & Young, 1970). Other instruments measuringaffective, cognitive, performance, and role changes wereconstructed for this project and were subjected to con-tent validity assessment. Internal consistency reliabilitycoefficients for the measures ranged from an alpha of0.55 to 0.93. Pre-post comparisons were made by t-tests(for correlated means) and correlation analyses weremade by the Pearson r. All cognitive tests were groupadministered; all other instruments were self-admin-istered. The 8-month follow-up semistructured inter-view was submitted to frequency analysis.

Major results of the .post - testing included a signifi-cant in8r-ase in knowledge scores (p < .03), reportedconfidence levels (p < .01), and reported frequency of useof presented assessment and intervention skills (p <.01).The pre-test scores of the ATDC correlated significantlywith the post-test knowledge scores (p < .001).

Major results of the 8-month follow-up interviewwith Group 1 participants included a perceived increasein empathy for handicapped children (23 percent). Othernursing role changes included a continued increase in' the

108

107

use of standardized assessment tools (24 percent); devel-opment of innovative approaches in the care of disabledchildren (33 percent); and increased comfort and par-ticipation in interdisciplinary relationships (30 percent).The results document the expansion of nursing practicewhich can be achieved when nurses participate in in-tensive, specialized, continuing education programs.Networking skills among participant nurses, their agen-cies, and referral hospitals which were established duriiparticipation in the program have become an importantsource of support for nurses working with disabledchildren throughout the metropolitan Washington, D. C.area. It was concluded, that significant change can bemade in nursing knowledge, attitudes and skills by par-ticipation in a continuing education program; and thatthe ATDC may be helpful in predicting post-test knowl-edge scores.

Reference

Yul;er, I-1., Block, 3., (3c Young, J. (1970). Measurementsof Attitudes Towards Disabled Persons. Albertson,N. Y.: Human Resources Center.

109108

04 EPIDEMIOLOGICAL SURVEY OF CORONARY HEARTDISEASE RISK FACTORS IN A BLACK COMMUNITY

Margie44. Johnson, RN, PhDAssociate Professor

Margaret T. Beard,. RN, PhDAssociate ProfessorCollege of Nursing

Texas Woman's University, Denton

Blacks have a 50 percent greater chance of devel.,oping hypertension than do whites (American Heart As,sociation, 1982; Freis, 1973) and, when present, it istwice as severe' (McIntosh, et a1, ,1978). It is estimatedthat One out of four blacks has hypertension comparedto=one out- of six whites (Hosten, 1980). The purposes ofthis -study were to identify coronary heart disease (CHD)risk factors. present in a -black community, to examinerelationthips between hypertension and other psycho-

, social and physiological risk fact s, and to gather,baseline data for the development of a multiple -riskfactor preventive program. This study viewed hyper-tension both as a risk factor and as a response to OtherCHD risk factors (i.e., diet, stress,,exercise).

The total' non-probability sample consisted of 285subjects-179 females and 106 males, ages 18 and above.Subjects were surveyed through eight predominatelyblack churches located in a small-urban town Of 47,730,with a black population of 4,739. Approximately 10

_;_percent of the black adultpopulation participated in thestudy. _

The design was ex post facto. The Risk FactorQuestionnaire (RFQ) was developed and piloted by amultidisciplinary team. The RFQ consisted of: demo-graphic data; self-report of ask factors, e.g.) heredity,smoking, alcohol, stress, diet, exercise; and risk factor

111

a

10

measurements of height, weight, and body mass. Stresswas measured by a 5-item Type A Personality Scale, an 8-item Life Satisfaction Scale and a 40-item InterpersonalTrust Scale (Rotter, 1967). Reliability coefficients forthis sample were 0.69, 0.69 and 0.76, respectively.

Subjects completed the RFQ and physical measureswere taken. Hypertensive participants were referred totheir personal physician or the project physician con-sultant. Pearson's correlations were utilized to deter-mine relationships. Multiple regression was used todetermine the effects and magnitudes of variables aspredictors.

Significant relationships were found between dia-stolic blood pressure and Type A Personality r267 = 04.(p = .05), Life Satisfaction r2 9q = 0.11 (p = .03), andInterpersonal Trust C264 = .24 (p = .001). Systolic anddiastolic blood pressure were significantly associatedwith recreational activity with r 279 = 0.26 (p = .001) and

r279.= 0.13 (p = .01), respectively. Weight was significantlyassociated ,with elevated systolic blood pressurer279 =0'36 (p = .001). Multiple regression analyses indicated,that high blood pressure in self, age, and salty foodswere predictors of diastolic blood pressure (Multiple R =0.64, Adjusted R Squared 0.39), while age, high bloodpressure in self, and fat intake were predictors ofsystolic blood pressure (Multiple R = 0.73, Adjusted RSquared 0.50).

Significant relationships between blood pressure andstress variables have implications for stress reductionprocedures. The statistically predictive values of sys-tont.., and diastolic blood pressure in self with dietary saltand fat intake have implications for community-widehealth teaching programs regarding diet and blood pres-sure control.

110112

References

American Heart Association. (1982). Heart facts.

Freis, E. D..(1973). Age, race, sex, and other indicesof risk in hypertension. American Journal of Medi-cine 51, -275 -280.

Hosten, A. 0. (1980). Hypertension in black and otherpopulations: Environmental factors and approachesto management. Journal of the National MedicalAssociation, 72 (2), 11-117.

McIntosh, H. D., Dknoyan, G., ac Jackson D. (1978).Hypertensiona potent risk factor. Heart and lung,7 (1), 137-140.

Rotter, J. B. (1967). A new scale for the measurementof interpersonal trust. Journal of Personality, 15,651-656.

113

VARIATIONS ON AN OLD THEME:AN UPDATE ON HANDWASHING

Elaine Larson, RN, PhD, FAANDirector, Quality iissurance and Research

University HospiialSeattle, Washington

Marcia Kil lien, RN, PhDLecturer, Parent Child Nursing

University of Washington School of Nursing, Seattle

Eric/Larson, MD, MPHAssistant Professor

University of Washington School of Medicine, Seattle

This presentation summarizes the results of thieeresearch projects conducted during_1980 to 1982. Majorpurposes of the studies were:

1. To determine the prevalence of -gram-negativebacteria (GNB) on hands of hospital personnel.

2. To identify factors which influence handwashing(HW) behavior of patient care personnel.

3. To evaluate the influence of a role model on HWbehavior.

The hand flora of 103 hospital personnel and 50controls were studied over a mean of 35 days. One ormore species of GNB were found to be carried per-sistently on the hands of 21 percent of hospital personneland 80 percent of controls (Larson, 1981). Despite thefact that hands of health care personnel are a doc-umented source of transmission of nosocomial infections,compliance with acceptable levels of HW is poor (Albert& Condie, 1981). Therefore, 193 hospital personnel, aconvenience sample, were surveyed with a questionnairebased on a decision theory model (Edwards, 1954) toidentify factors important in influencing individuals towash or not wash their hands. One significant factoridentified as an important determinant of HW behavior

1,15_112._

was the HW practices of professional colleagues (Larson(Sc Killien, 1982). To test this possibility, a study wasconducted to assess the effects of attending physicianHW behavior on HW practices of medical house staff.

For two months one investigator accompanied a med-ical team (1) attending rounds during which the attendingphysician "role model" washed hands between each pa-tient contact (experimental team), (2) work rounds withthe same team in the absence of the "role model," and(3) work rounds with the same team in a succeedingmonth when the attending physician and house staff hadchihged (control team). None of the team membersexcept the "role model" was aware of the study. '°

During; the study period 87 patients were visited, and169 contacts when HW were indicated. HW occurred in47 (29.0 percent) of these instances. 11W practices ofexperimental and control groups during work rounds weresignificantly different: the experimental team washed48.1 percent of times indicated; the control team, 24.1percent, p < .05. If our findings are substantiated bystudies in various health care settings and observationswith other health care personnel groups, there are sig-nificant implications for the importance of the clinicalmentor in influencing 11W behavior.

-References

Albert, R. K., (5c Condie, F. .(1981). Handwashing pat-terns in medical intensive care units. The New En-gland Journal of Medicine, 221, 1465-1466.

Edwards, W. (1954). The theory of decision making.Psychological Bulletin, .21, 380-417.

Larson, E., do Killien, M. (1982). Factors influencinghandwashing behavior of patient care personnel.American Journal of Infection Control, 10, 93-99.

Larson, E. L. (1981).. Persistent carriage of gram-negativebacteria on hands. American Journal of InfectionControl,.9 112-119.

1.16113- --

DEATH EDUCATION AND DEATH ANXIETY

Bonnie Elam Loc.s.kard, RN,- EdDAssociate Professor, Division of Nursing

Mississippi University for Women, Columbus

The purpose of this study was to investigate theimmediate, residual, and long-term residual effects Ofdeath education on the death anxiety level of nursingstudents. Death anxiety level was also related to threecategorical variables: age, personal death experience,and previous nursing experience.

Two intact classes of female nursing students from asmall women's university were randomly assigned toexperimental or control groups. Seventy-four subjectscompleted the study through the post-posttest and 50subjects were retained through the follow-up study oneyear later.

This study utilized a pretest-posttest repeated mea-sures quasi-experimental design. Analysis of covarianceand stepwise multiple regression statistical models wereused for data analysis. Templer's (1970) Death AnxietyScale (DAS) was employed for all tests of the dependentvariable.

Data for the study were collected during a 15-monthperiod from October 1981 through December 1982. Af-ter completing the pretest and a questionnaire, subjectsin the experimental goup participated in a seven-hourdeath education instructional unit while subjects in thecontrol group continued with regular classroom instruc'tion. Both- groups were posttested_immediately after theinstruction and post-posttested four weeks later. Oneyear later a follow-up study was conducted by mail witha 68 percent response rate.

Results of analysis of covariance on all three post-treatment testings revealed a significant difference

117

114

(p< .001) between the death anxiety level of subjects whoparticipated in the death education and those who didnot. Results of stepwise multiple regression procedureindicated that there was no relationship between thecategorical variables studied and pretest death anxietylevel. On the follow-up study, however, there was anegative relationship between previous nursing_ expe-rience and the death anxiety level of subjects in theexperimental group.

The death education instructional unit utilized in thisstudy appeared to result in a reduction in death anxietylevel, which remained stable over a one-year period.There was also an indication that subjects in the ex-perimental- group with previous nursing experience weremore resistant to changing their death anxiety level thanthose without previous nursing experience.

The following implications emerged from this"- re-search: (1) Death education must receive high priority inursing education to reduce the death anxiety levo,preservice nurses. (2) Individuals with previous nag*/experience need more intensive death education pro-grams in order to effect a positive change in deathanxiety level.

Reference

Templer, D. I. (1970). The construction and validation ofa death anxiety scale. The Journal of GeneralPsychology, 82, 165-177.

115118

RELIABILITY ESTIMATES IN PUBLISHINGNURSING RESEARCH: CURRENT USE AND DISUSE

Mary R. Lynn, RN, PhDAssistant Professor, School of Nursing

University of Mississippi Medical Center, Jackson

Reliability estimates for an instrument or other datacollection strategy are nal sample invariant, as they arenot a property of the data collection strategy. There-fore, the reliability of a data collection strategy must bereestimated in each application. The purpose of thisstudy was to determine the extent of reliability esti-mates, for all types of data collection strategies, inrecently published nursing research. A second purposewas to see if there had _been any change (increase) in thereporting of reliability estimates over the study period.

An ex post facto survey design was utilized in thisstudy. A stratified random sample of studies published.during the past five years (1978 to 1982) in Nur2simiResearch and Research in Nursing and Health was se-lected. _the total number of research articles publishedin each journal during the live years was determined anda proportional allocation of articles was selected fromeach journal. A total of 70 articles were surveyed.Using an investigator-designed coding form, the follow-ing data were collected by the researcher from eacharticle:, (1) intended applications of the finding(s) of thestudyclinical, educational, administrative, or other ap-plication; (2) methods of data collection_ or strategyutilizedpreviously established, new, or a combinationof established and new data collection strategies; (3)report of previous reliability estimates of establisheddata collection strategies; and (4) report of reliabilityestimates for the data collection strategy or strategiesused in the current study. Using 10 non-sampled articlesas a pilot and a one-week interval, the intra-raterreliability for this data collection was found to be .98.

11.9 116

As the proportion of studies in which reliabilityestimates were reported was not significantly differentbetween the two journals (t = 0.204), combining thearticles across the two journals was considered appro-priate. For the two journals combined, 45 percent of thearticles were based on the use of established datacollection strategies, 38 percent on the use of new datacollection strategies, and 13 percent were based oncombinations of established and new data collectionstrategies. For 3 percent of the articles, existing datasets were used.

In the articles in which only established data col-lection strategies were used, 56 percent contained reli-ability estimates for the previous use(s) of the strategywhereas only 22 percent contained reliability estimatesfor the current use of the strategy. In the articles whereonly newly developed data collection strategies wereused, 52 percent contained reliability estimates.

When both established and new data collection strat-egies were used in combination, the extent of reliabilityestimation or reestimation was different from the sin-gular use of these strategies. In these strategy combi-nation articles, no reliability estimates were reportedfor the previous or current uses of the establishedstrategy, yet 89 percent of the new strategies hadreported reliability estimates.

When the proportion of articles per year with reli-abilty estimates was addressed, it was found that therewas no significant change in the reporting of reliabilityover the study period. This was found for the combina-tion of the two journals as well as for each journalseparately. It should be noted, however, that the lack ofsignificance is probably a function of the use of theproportion of articles per year which resulted in a smallsample size (n = 5 years), thereby decreasing the prob-ability of a significant finding. From a cursory analysisof that same set of data, there is a noted increase in the

117 120

proportion of articles in Nursing Research with reli-ability estimates and a corresponding decrease of thesame in Research in Nursing and Health.

It was concluded that the estimation of reliability ofmeasurement strategies is either not receiving sufficientattention in the teaching of the research process or isnot considered necessary for the acceptance of a nursingresearch article. Despite the increased sophistication innursing research, a fundamental principle of measure-ment, the reliability of the measures, is commonly beingignored. Reliability, like validity, is sample-dependentand requires estimation with each application of a datacollection. strategy. Knowledge of the current reli-ability of a data collection strategy is necessary foraccurate interpretation of the findings of a study.Therefore, guidelines for the estimation of reliabilityfor each type of data collection strategy were presented.

11

121

WOMEN'S LITERATURE-ANDTHE CLIENT BASE OF NURSING

Myrtle P. Matejski, RN, PhD.Assistant Dean and Associate =Professor

College of NursingUniversity of Rhode Island, Kingston

Historical studies of nursing usually describe theevolution of nursing as a profession, concentrating onnursing leaders, their concepts of nursing, and theiractivities. Such studies engender understanding of theirmotives, beliefs, values, and hopes. However, nursingliterature reveals a persistent' gap concerning appro-priate education, appropriate clinical functions, and atti-tudes as perceived by nursing leaders and as perceived bythe "rank and file."

Sigerist (1960), a medical historian, has suggestedthat history should provide a more complete picture ofthe development of civilization; it should make us awareof our-beginnings, of where we stand today, and thedirection in which we are heading. Shryock (1953) andClarke (1971) observed that the history of any science orprofession could best be interpreted in terms of itsinternal, trends and its relation with the social andcultural environment in which it evolved.

The purpose of this study was to identify the interplayof cultural values- and social norms with women's per-ception of role as revealed through popular women'sliterature prior to the emergence of professional nursingin the United States and its effect on the client base ofnursing. A situational approach was used, combiningpeople and the environment Into one analytic scheme(Berkhofer, 1969).

Relations between the internal aspects of events andthe external environment were anlyzed, focusing on what

123 119

people do and how they behave. Emphasii was on thecultural and social factors influencing situations andnursing behavior. Limited attention was given to con-.scions intention or unconscious motivation of the writersand no attention was given to their perceptions of

.physical environment or physiological factors.

Merton (1957) suggested that social structures mightbe analyzed by changing function from a mathematicalsense to an organic sense by using 11 steps for analysis.In modifying Merton's paradigm only three steps were.

retained:

1. The objects of analysis were those representingstandardized and repetitive social roles and cul-tural patterns. Socially acceptable women'sroles and attitudes as portrayed in women'sliterature were, then, analyzed against acceptednorms of behavior for nurses as revealed innursing literature, letters, editorials, and arti-cles in early nursing journals.

2. Subjective motives and purposes of women au-thors were separated from the objective resultsof their writings.

3. Role segmentation, hierarchical ordering of val-ues, and social divisions of labor were analyzed.

Examination of popular women's literature of the lastcentury revealed a centrality of mission. Focus was onheroines who surmounted heavy burdens and-were noble,virtuous, kind, sympathetic, helpful, cheerful, and loyal,no matter what the difficulty. These were also attrib-utes of the "good nurse" that are described repeatedly inthe reviews of letters, editorials, and articles publishedin nursing journals from 1889-1920. Obedience, silence,and anonymity were frequently cited as desirable atti-tudes. As with heroines in women's, literature, nurseswere expected to be morally superior beings.

120124

Women's literature, read by thousands of middle andlower socioeconomic classes of American women, influ-enced these women concerning role attributes and ac-ceptable- attitudes and status in society. The novels alsoinfluenced those who would become nurses in the twen-tieth century. The whole contributed to the disparitythat persists between leaders of nursing and, their per-ceptions of nursing role and the perception of theaverage nurse clinician.

References .

Berkhofer, R. F. (1969). Behavioral approach to histor-ical-an4ysis. New York: The Free Press.

Clarke, E. C. (Ed.). (1971). Modern methods in thehistory of medicine. London: The Athlone Pressof the University of London.

Merton, R. K. (1957). Social theory and social structureed.).. Glencoe, IlPnois: Free Press.

Shryock, R. Hr (1953, April). The interplay of social- and internal factors in the history of modern medi-

cine. The Scientific Monthly, pp. 221-229.

Sigerist, H. E. (1960). On the sociology of medicineH. I. Buemer (Ed.). New York: M. D. Publications.

121125

RELATIONSHIP OF POLICY TO PRACTICE IN THEDELIVERY OF MEDICARE HOME HEALTH SERVICES

Mary 0; Mundinger, RN, DrPHDirector, Graduate \Program

School of NursingColumbia UniversityNew York, New York

This was an exploratory study conducted to examineand analyze the relationship between Medicare policy forhome health services and the actual delivery of thoseservices. The purposes of the research were:

1. To reveal the correlates of professional nursedecision making in providing home carein par-ticular, how policy, patient characteristics, andcomponents of the scientific discipline of nursingare related to the care given;

2. To learn more about the nature of nursing auton-omyin particular, what kind of accountabilitynurses assume in delivery of home care services.

The methodology was primarily qualitative and in-cluded interviews, record review, participant observa-tion, and analysis of case studies developed for allpatients observed receiving care. Qualtitative analysiswas also used to suggest three hypotheses.-

The design of the study was to observe5O'home visitsto Medigare patietttk to collect data on all process andenvironmental components of those visits, and to analyzethe content of the visit and care given. ---

The patient, visits were selected -.1ndornly torepresent all geographic, ethnic, and so Itural ele-ments in the population served by the agt chosen ictstudy. Medicare policy criteria selected e,ate the threemajor eligibility criteria: homebound, skilled care, ar4

127 122

4

physician plan of care. Policy compliance was deter-mined using Medicare definitions for each criterion.Numerical data were collected regarding policy com-pliance; qualitative data were collected to describeactual service delivery and to explain policy non-compliance.

Of the 50 patients visited, 19 were not homebound,16 were not receiving, skilled care, and two had nophysician plan of care. Seven were doubly ineligible andone patient met none af the three criteria. Only 22, orless than half, of the patients receiving care reimbursed byMedicare were truly eligible.

Nurses consciously and deliberately manipulatedpolicy to give needed care. In comparing patients whowere ineligible for care with those who were eligible,important differences emerged. Ineligible-patients,moreoften: lived alone; needed adaptive assistance; andneeded multiple services not available in any other singlecare setting.

. Further analysis of physician plans of care revealedthe following:

1. Mi physician_ plans of care were amended &-rewritten by the nurse after the first visit.

2. In exactly half of the- cases (25, of 50)y thephysician441frins of care were inadequate, even'forthe very basic aspects of skilled care.

3. Specialist physician plans were more often inad-equate than plans written by primary care physi-cians.

Hypotheses emerging from the study include:

1. Nurses manipulate policy to give needed care.

2. . Nurses assume accountability for care planningand care delivery, even in the absence of policymandate.

128

_12-3

,-,

3. Nurses identify areas in which their scientificdiscipline operates, even when at odds withpolicy.

Implications are broad and important:

1. Nurse directed and delivered home care services,may be more legitimate and comprehensive thanphysician directed care; cost implications of thisfinding may 'also be positivg

2. Policy development and evaluation may requiredifferent criteria when dealing with behavior. ofprofessionals.

129124

, THE EFFECTS OF MODIFYINGTHE ENDOTRACHEAL TUBE SUCTIONING/MANUAL

HYPERVENTILATION PROCEDURE-ON SEVERECLOSED HEAD IN3U1tED PERSONS

L. Claire Parsoris,.RN, PhDProfessor, School of Nursing

Dorothy Vollmer, RN, MSNNursing Staff, Medical Center

Leanna 3. Crosby, RN, MSNResearch Associate, School of Nursirg

-Pam Holley-Wilcox, PhDProgram Analyzer, Computer CenterUniversity of Virginia, Charlottesville

*

Closed, head injuries constitute a serious and some-times life threatening condition to persons so afflicted -because of secondary injury associated with the rapidlydeveloping cerebral edema and/or intracranial bleeding(3ennett, 1977). These secondary injuries result in tie--vated intracranial pressure (ICP) and decrease cerebralperfusion pressure (CPP). While therapeutic interven-tions have provided , ways of medically managing theacutely. ill, head injured person (Marshall & Shapiro,1977), nursing interventions have not been systematicallyor quantitatively investigated except as pilot studies(Shalit & Umansky, 1977; Mitchell & Mauss, 197g). Thus,the safety of instituting nursing interventions remainsunclear. -

This paper reports findings of a follow-up, study re-lated to the effectt of modifying iLle nursing interven-tion- of endotracheal tube suctioning/inanual-hyperventi-lation (ETTS/MH)_ procedure. on severe closed head in-jured persons. An earlier study (Parsons & Shogan, Inpress) documented that the Procedure of ETTS/Mti re-sulted in significant (p <.05)-incremental increases in the

131125

physiologic measurements of heart rate (HR), meanarterial blood pressure (MABP), mean intracranial pres-sure (MICP), and CPP. These elevations were morepronounced iollowing the second and third catheteralpasses into the endotracheal tube. In the earlier study,MH was only 30 seconds in duration after each cathe-teral pass. Physiologic measurements remained elevatedabove baseline levels for at least one minute following-the termination of the procedures.

Research questions addressed included: (1) Whateffect does modifying the MH procedure from 30 to 60'seconds after the second and third catheteral passeshave on the cerebrovascular status of 'severe closed headinjured patients whose baseline (resting) MICP is < 20mmHg? (2) Do physiologic measurements of HR, MABP,MICP, and CPP recorded upon subjects during theETTS/MH_ intervention return to baseline (resting) levelswithin a five-mintue period following the termination dfthe intervention?

The sample consisted of 15 severe closed head injuredpatients who had been admitted to a Surgical IntensiveCare unit (SICU) in a University Medical Center, and whomet criteria for selection. No restrictions were placed'on the sample with respect to age, sex, or admissionblood alcohol concentration (BAC).

Subjects with. documented severe closed head injurywere selected who required intubation and mechanical-ventilation and who had in place a functioning subarach-noid bolt (SAB); an arterial catheter; and standardelectrocardiographic electrodes for purposes of record-ing mean intracranial_ pressure (WO, mean arterialblood pressure (MABP) and heart rate (HR). Cerebralperfusion pressure (CPP) was calculated using the equa-tion, CPP = MABP MICP. All subjects were required tobe without severe pulmonary-thoracic complications.Recording of the four dependent variables (i.e., HR,MABP, MICP, and CPP) were made using- a calibratedbedside Hewlett-Packard Patient Monitoring System

132

(H-PPMS). The research design was a partially between-subject and partially within-subject (repeated measures)quasi-experimental design. A median split of the samplewas done by eliminating one subject who had an un-usually high baseline (resting) MICP; the remaining 14subjects were divided into two equal groups of seven.Group I included subjects with a baseline MICP of < 9mmHg; group II contained subjects with a baseline of > 9mmHg. Specific comparisons were made between aildwithin groups, with 13 levels of repeated observations ineach of four qualitatively different sets: baseline,manual hyperventilation, endotracheal tube suctioning,and recovery. Data were analyzed using multivariateanalysis of variance (MANOVA), with the repeated ob-servations treated as multiple dependent measures. Sep-arate analyses were carried out for the phySiologicmeasures of HR, MABP, MICP, and CPP. Specialcontrasts and trend analyses were used where appro-priate to address specific research questions. The alphalevel of significance was set with p < .05.

Of the 15 subjects, eight were males and seven werefemales, ranging in age between 14 and 83 years. Eightsubjects were between 14 and 32 years old while sevensubjects were over 35 years of age. Only one subjecthad an admission blood alcohol concentration 100'> mg/dl.The range of the Glasglow Coma Scale (Teasdale andSennett, 1974) scores varied between 2 and 14.

Preliminary analyses showed that the group maineffect was not statistically significant for any of the fourphysiologic variables. However, the obseryation maineffect was significant relative to HR, MABP, MICP andCPP. Because significant differences existed amongobservations, further analysis of repeated measures wasjustified.

A weighted contrast compared baseline values on onehand to MI-11-4 on the other. Significant differenceswere found for the variables of HR, MICP, and CPP.When baseline values of the dependent variables were

133 127

compared to ETTS1-3 through another special contrast,significant differences were shown for HR, MABP, andCPP. Contrasts between MI-11_4 and ETTS 1_3 demon-strated significant differences for HR, MABP, MICP andCPP.

The main effect of the group was not significant forany of the physiologic variables.. The observation maineffect showed significant differences relative to MICPand CPP. This effect approached significance for MABP

= .07). The group ...by observation interaction wassignificant only for MICP.

To further explore the recovery data trend analyseswere performed for, observations 9-13 of the protocol.Significant linear trends were found for MABP, MICP,and CPP. A significant quartic trend was also found forCPP.

Demographic findings relative to age distribution andsex tended to agree with the findings of others(Caveness, 1979; Kalsbeck et al, 1980). The extension intime of,MH following the second and subsequent suction-ings demonstrated that MICP could be decreased tolower than baseline levels. Extending the recoveryrecording time from one to five minutes provided ade-quate time for dependent variables to reach baselinevalues. Thus, the modified ETTS/MH procedure can berecommended as a more effective procedure in thisMedical Center for suctioning closed head injured per-sons with resting MICPs of <20 mmHg.

The linear and quadratic trends during the inter-ventions of MH1-4 and ETTS 1-3 reflected the parallelhemodynamic relationship between MABP and CPP ob-served when the cerebrovascular is autoregulated. Thedecreasing linear trend in MICP reflects the brain'smicrocirculatory response, to decreased levels of Pa 02brought about by extending in time MHs three and four.

128

References

Caveness, W. F. (1979). Incidence of craniocerebraltrauma. in the United States in 1976 with trendsfrom 1970 to 1975. In R. A. Thompson dc J. R. Green(Eds.), Advances in Neurology, (Vol. 22). New York:Raven Press.

Jennett, W. D. (1977). Severe head injuries in threecountries. Journal of Neurology, Neurosurgery,-Psychiatry., IA 291-298.

Kalsbeck, W. D., 'Mc Laurin, R. L., Harrii, B. S., (3c Miller,3. D. (1980). The national head and spinal cordinjury survey: Major findings. Journal of Neurosurgery,53 519-553.

, Marshall, L. F., & Shapiro, H. M. (1977). Barbituratecontrol of intracranial hypertension in head injuryand other conditions: latrogenic coma. Acta Neuro-' logica Scandinavia, 56(Suppl. 64), 156-157.

Mitchell, P. H., & Mauss, N. K. (1978). Relationshipof patient-nurse activity to intracranial pressurevariations: A pilot study. Nursing_ Research,4-10,

Parsons, L. C., & Shogan, J. S. (In press). The effectsof endotracheal tube suctioning/manual hyperventi-lation procedure upon severe closed head injuredpatients. Heart and Lung.

Shalit, M. N., & timansky, F. (1977). Effects of routinebedside procedures on intracranial pressure.Journal of Medical Science,13, 881-886.

Teasdale, G., ac Jennett, it. (1974). Assessment of cimaand impaired consciousness: A practical scale.

- Lancet, 2 81-84.

135 129;

PARENTAL RESPONSES TO EARLY CHILDHOOD

CHRONIC ILLNESS: CONGENITAL HEART DISEASE

Lesley A. Perry, RN, PhDAssistant Professor, School of NursingUniversity of Maryland at Baltimore

The purpose of this study was to determine whatfactors were related to anxiety in parents of childrenwith congenital heart disease. The Lazarus (1966) theoryof psychological stress and coping was used as a theo--retical framework. The study examined the extent thatdegree of anxiety in parents was related to a number ofpersonal and situational factors, including: patentalself-esteem, coping style, and perception of severity ofthe child's illness; family demographic characterittics;the occurrence of surgery for the cardiac defect; and thelength of time since diagnosis. It was hypothesized thatparents who experienced greater anxiety would (a) per=ceive their child's illness to. be more severe, (b) havelower self-esteem, and (c) be more likely to use asensitizing, rather than a repressive, coping style.

Forty-one couples (41 mothers and 41 fathers) withchildren under three years of age who had been diag-nosed as having congenital heart disease within the yearproceeding this study and who were expected to needcontinuing treatment for the defect constituted thesubjects for this study. The study subjects were drawnfrom the pediatric cardiology clinic of a large universityhospital. A home visit was arranged in which thesubjects completed an Information Questionnaire,- a Per-ception of Severity, of IllneSs Rating Scale (Perry, 1982),

the Rosenberg Self-Esteem Scale (Rosenberg, 1965), andthe Modified Repression-Sensitization Scale (Epstein ScFenz, 1967). Anxiety was measured using the State forinof the State-Trait Anxiety Inventory (Spielberger,

137 130

1970). Forced order and step-wise multiple linear re-gression analyses were performed to determine pre-

- dictors of parental anxiety. The findings supported thehypotheses related to perception of severity of illnessand Coping style. The hypothesis related to self - esteemwas not supported. Of the other factors examined, onlylength of time since diagnosii was found to be a sig-nificant predictor of parental anxiety, with anxiety de-creasing as the length of time increased. The findings ofthe study have implications for identifying parents at

- risk for experiencing high levels of anxiety and fortargeting professional programs and resources to aid thisvulnerable group. Intervention strategies which takeinto account identified factors could assist in reducingparental anxiety.

References

Epstein, S., k`Fenz, W. D. (1967). The detection of areasof emotional stress through variations in perceptualthreshold and physiological arousal. Journal of Exper-imental Research in Personality, 2, 191-199.

Lazarus, R. S. (1966). Psychological stress and the cop-ing process. New York: McGraw-Hill Book Company.

Perry, L. A. (1982). Factors related to anxiety andstress in parents of children with congenital heartdisease. Unpublished doctoral dissertation, Univer-sity of. Maryland.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, New Jersey: Princeton UniversityPress. 0'

Spielberger, C. D., Gorsuch, R. L., 8c Lushene, R. E.(1970). The State-Trait Anxiety Inventory Test Manual.Palo Alto, California: Consulting PsychologistsPress, Inc.

131 138

THE POLICY AND POLITICSOF FLORENCE NIGHTINGALE

Barbara Pillar, RN, MSN, CSPhD Candidate, School of Nursing

University of Maryland at Baltimore

At times, it seems that Florence Nightingale is oldhat. More than 120 years after Notes On Nursing waswritten (Nightingale, 1860), one wonders what else canbe said about the woman or her ideas. There is,however, a renewed interest in Nightingale, and closescrutiny of her writings reveals that we have just begunthe inventory of her legacy to nursing. This paper usesthe approach of policy analysis to examine Nightingale'swork and her contribution to nursing. It is 'divided intothree Major parts: policy formation, policy imple-mentation, and policy evaluation.

Policy formation looks at the factors that led Night-ingale into the role of reformer. It proposes that themajor influences were her personal needs and life expe-riences, her education, and her powers of observationand affinity for research. It is held that these threeareas were integrated by Nightingale in such a way thatthe resulting ideas she developed for reform werenecessary, timely, realistic, and thorough. The areas ofreform that are examined are the practice of nursing,nursing education, and hospital management. Specif-ically, each area is described as it existed before Night-ingale and _is followed by a description of the changesinstituted by her.

Policy implementation examines the basis of Night-ingale's influence and political power that brought herproposed policy changes to fruition. Using Zey-Ferrell'sclassification (1979) of the major sources of power, it isargued that Nightingale's ability to institute change on anational level came from her knowledge and compe-tency, her personal attributes (charisma), and from her

139132

social position. The charge that she was overzealous andruthless in her drive to reform the delivery of nursingcare and hospital services is examined in the light of 'herreligious values.

, -Policy evaluation _looks at the accomplishments of

Nightingale during her lifetime and their continued im-pact on the health care system today. Gains and lossesfor the nursing profession, and other health care pro-fessions, since Nightingale's time are outlined. Finally,her values and methods are examined for possible appli-cation to two of the major underdeveloped areas ofpresent-day nursingthe need for well-defined, knowl-edge-based practice and the need for nursing to beincluded in the decision-making process of the healthcare system.

References

Nightingale, F. (1969). Notes on nursing: What it is andwhat it is not. New York: Dover PublicaVons,(Republication of the first American edition, 1860).

Zey-Ferrell, M. (1979). Dimensions of organizations.Santa Monica: Goodyear Publishing Co.

133

-140

A COMPARISON OF EMOTIONAL INDICATORS ONHUMAN-FIGURE AND SELF-FIGURE DRAWINGS OF

LEUKEMIC AND HEALTHY CHILDREN

Joanne-B. Scungio, RN, PhDAssociate Professor, School of NursingUniversity of Alabama in Birmingham

A drawing captures symbolically on paper many ofchildren's thoughts and feelings about themselves or ofsignificant others, which they cannot express in speechor writing. "It makes a portion of the inner self visible"(Klepsch 8c Logic, 1982, p. 6). More is revealed by theconetnt, which is determined by the way the child,consciously or unconsciously, perceives himself and sig-nificant others in his life.

This descriptive study sought to compare the emo-tional indictors leukemic and healthychildren included intheir human and self-figure drawings. Three sub-problems were investigated: (a) Are there significantdifferences in the number and/or type of emotionalindicators children draw on the human-figure and. self-figure drawings? (b) Are sex and/or age contributing-factors in number and/or type of emotional indicatorschildren include in their drawings? (c) Are the numberand type of colors used in drawings significantly dif-ferent between leukemic and healthy children?

The study sample was comprised' of 40 randomlyselected school-age children (8-12 years of age); 20children with acute lymphocytic leukemia from a pedi-atric oncology clinic, and 20 healthy children from aneighborhood in Southern New England:

Children's human-figure and self-figure drawingswere utilized as the data collection tools. The childrenwere asked to draw a person (someone other than them-selves) and then a picture of themselves. Reliability of

141 . 134

the investigator's scoring technique of children's draw-.,Ings was r =.96.

Comments and descriptions of children's drawingswere content analyzed. A 3-way ANOVA, was utilized todetermine if a significant difference existed between themean emotional indicator scores on leukemic and healthychildren's pictures. The data were analyzed for group,sex, and age differences. Chi-square analyses wereconducted to identify if significant differences existed inthe number and types of colors children used in theirfigure dtawings.

Findings from the study indicated that:*'

I. Leukemic children included significantly moreemotional indicators on both their human-figureand self-figure drawings than did the healthychildren.

2. The types of emotional indicators on leukemic andhealthy children's drawings differed significantly."

3. Sex and age were not contributing factors in thedifferences found between leukemic and healthychildren's drawings.

4. Leukemic children used significantly fewercolors in their self-figure drawings than did thehealthy children.

Types of emotional indicators identified on leukemicchildren's human-figure drawings included such items aspoor integration of body parts, shading of the licelyand/or limbs, gross asymmetry of limbs and teeth, handscut off, and arms clinging to sides. In addition, leukemicchildren included shading of the' face, hands, or limbs;tiny figures; slanted figures; omission of nose, mouth,hands, feet, neck, and/or hair; crossed eyes; and shortarms and/or legs on their self-figure drawings.

1 3.5 142

Results of the study suggest that the ramifications ofa chronic illness such as leukemia produce numerousemotional stressors in the school-age izhild. Drawingsprovide an excellent medium for children to expresstheir concerns, fears, and misconceptions about theirillness. Drawings can provide the nurse with an assess-ment of children's ideas regarding their own situation.Once the data have been obtained and interpreted, thenurse must use the material to guide future interventionand interactions_ with the child and the family, andprovide opportunities for further discussion and support.

Reference

Klepsch, M. & Logie, L. (1982). Children draw and tell:An introduction to the projective uses of children'shuman figure drawings. New York: Brunner/Mazel.

140T FLUIDS,FOCC THE FEBRILE -

PATIENT: WARMING UP TO A NON-TRADITIONALNURSING INTERVENTION

'Jeanne Sorrell, RN, MSNAssistant PrOfessor,:SchoOl Of Nursing

Shenandoah,College dc Contervatory of MusicWinchester, Virginia

Nursing literature contains few systematic studies ofthe effectiveness of various methods for reducing fever,even though clinical fever is a problem which frequently-confronts the nurse. The purpose of this study was todemonstrate selected physiological phenomena basic to arationale for a specific nursing intervention in the feb-rile patient. The intervention selected was that of -administering an internal heat load, in the form of a hotdrink, to the patient with fever. The 'rationale for thestudy, was that an ingested heat load would raise thetemperature of blood perfusing the hypothalamus, andthat the threshold temperature for activating heat lossmchanisms would be achieved more rapidly than bymetabolic heat alone.

Responses of temperature, sweating, and subjectivecomfort in afebrile and febrile subjects were measuredafter ingestion of a heat load. A convenience sample of10 afebrile and 8 febrile subjects was selected. Theafebrile group was composed of healthy volunteers withtemperatiwes ranging from 36.25-36.65 C. The febrilegroups included hospitalized patients with temperaturesranging from 36.80-38.65 C. A heat load consisting of300 cc of tea at 64 C was administered to each subject.Tympanic membrane and rectal temperatures were Mon-itored continuously until approximately 20 minutes aftercompletion of the heat load. Skin temperatures werenoted at one-minute intervals during the testing period.They were monitored with a surface temperature probe.Sweating responses were evaluated once before and afterthe heat load by the use of the starchiodine test and' the

145 1-3 7

plastic impression method. Subjective comfort responsesalso'were obtained for subjects once before and after theheat load.

Values for tympanic temperature before and at thegreatest evaluation following heat load were analyzed byt-tests for paired samples. Findings supported thehypothesis that there is a small, rapid increase of intra-cranial temperature in both the afebrile and febrile stateafter ingestion of a heat load. A comparison of tym-panic temperature increases between afebrile and febrilegroups utilizing a t-test for unpaired samples revealed nosignificant difference between groups in the amount oftemperature increase. Thc.- time interval to reach half ofthe _peak tympanic temperature elevation, however, wasfound to be significantly shorter for febrile subjects,indicating a more rapid initial increase of intracranial,temperature in these subjects. Rectal temperature data,analyze i by the t-test for paired samples, supported thehypothesis that there would be no significant increase inrectal temperature in either afebrile or febrile groupsafter the heat load. Values for individual skin tem-peratures before and at the greatest elevation followingheat load were analyzed by t-tests; resultant increases intemperatures were found to be significant for bothafebrile and febrile subjects. Sweating responses wereanalyzed with the Wilcoxon matched-pairs signed-ranks,test; a significant increase in the number of active sweatglands after heat load was shown for febrile subjects.Subjective comfort responses were analyzed by t-tests.Sensations of increased warmth and increased sweatingafter heat load were found to be significant for theafebrile, but not the febrile group.

Findings in this study demonstrate that a smallinternal heat load can raise intracranial temperature.The small amount of tympanic temperature increase wassimilar for both afebrile and febrile subjects, suggestingstrongly that ingestion of a moderate amount of hot fluid

138 146

by febrile patients does not precipitate a response mech-anism unique to febrile patients. Thus, data from thisstudy raise the question of the advisability of nursingpersonnel giving only cold fluids to the febrile patient.

A second important implication for nursing Care ofthe febrile patient relates to the finding that the sr .311ingested heat load did activate heat loss mechanisms.Data indicated increased skin temperatures for afebrileand. febrile subjects, increased number of active sweatglands for febrile subjects, and increased feelings ofwarmth and sweating for afebrile subjects. It is possiblethat the small increase of intracranial temperature aftera hot drink could constitute an input to the thermo-regulatory system that might otherwise be gained onlywith continued shivering. The aborting of shiveringcould be of considerable benefit, because the increase inmetabolism --as much as 7 percent per degree centi-grade (Kluger, 1980)during shivering is detrimental tomany patients. Data revealing activation of heat lossmechanismsz,after heat load, in conjunction with theknown interdependence of sweating and shivering(Shaver, 1982), support the contention that a hot drinkcan abort shivering.

This study advances a significant implication fornursing care of the febrile patient in relation to the sitefor monitoring internal bcidy temperature. The tern-perature measurement site should be selected in ac-cordance with the purpose for measurement. Results ofthis study and the literature related to temperaturemeasurement (Benzinger, 1969; Ilsley, Rutten, do

Runciman, 1983) strongly suggest that rectal temper-ature measurement is not an appropriate site in situ=ations where precise temperature changes need to beevaluated. Tympanic thermometry was found to be asimple, comfortable, and sensitive method for moni-toring small internal body temperature changes.

139

References

Benzinger, M. (1969). Tympanic thermometry insurgery and anesthesia. The Journal of the AmericanMedical Association, 209_, 1207-1211.

Ilsley, A. H,, Rutten, A. 3., & Runciman, W. B. (1983).An evaluation of body temperature measurement."Anaesthesia and Intensive Care, kb 31-39.

Kiuger, M. 3. (1980). Fever. Pediatrics, 66, 720-723.

Shaver, J. F. (1982). The basic mechanisms of fevers:Considerations for therapy. Nurse Practitioner,7 (10), 15-19.

140148

THE RELATION OF' LOCUS OF CONTROL,SEX-ROLEIDENTIT.V, AND ASSERTIVENESSIN BACCALAUREATE NURSING STUDENTS

Kathleen:R. Stevens, RN,,EdDClinical Nurse Specialist for Nursing Research

M. D. Anderson HospitalHotiston, Texas

The role of nursing is developing toward greater inde-pendence and responsibility, and nurses 'are being, urgedto exercise more self- and professional-determination(Ashley, 1976; Claus & Bailey, 1977; Donnelly, Mengel, &Sutterly, 1980; Heide, 1973; Herman, 1978; Matiksch,.1980; Sandford, 1978; Stuart, 1981). Self-confidence,assertiveness, and, competency are necessary if nursingstudents are to successfully fulfill this contemporaryrole in nursing.

The aim of this study was to increase the under-standing of the interrelations among the variables ofpersonal control, sex-role identity, and assertiveness innursing students. The objectives were to describe: (I)the profiles of lochs of control, sex-role identity, andassertiveness of beginning and_grailuating nursing stu-dents, and (2) hovil these attributes are related in anursing student sample.

Data were collected from 185 beginning and 125graduating female nursing students enrolled in-all five ofthe state-supported ;baccalaureate nursing programsfiliated with -major health science centers in Texas. TheLevenson (1972) locus of controLsCale -measured internalcontrol, powerful others, and chance. The PersonalAttributes Questionnaire (Spence, Helmreich, & Stapp,1974) measured sex-role identity on the Masculinity andfemininity scales. The Rathus (1973) Assertiveness Sched-Aile assessed assertiveness. Reliability of the instru-ments was moderate to high (r = .63-to .90).

149 141

The demographic data- supported the concluSion thatthe sample reflected the trends existing in the study

, population. The resulting profiles of locus of control,sex-role identity, and assertiveness serve as referentvalues in future studies.

Multiple correlation technique was used to determinethe relations between the assertiveness and locus ofcontrol subscales together with sex-role identity sub-scales. Internal control and masculinity were signif-icantly related to assertiveness. A canonical correlationtechnique revealed that locus of control and sex-roleidentity were minimally related. Both are necessary fora comprehensive explanation of assertiveness.

Multivariate analysis of variance (MANOVA) yieldedsignificant differences for the control variable, school.The nonsignificant Status-by-Schopl effect enabled test-ing of the main effect, status. The MANOVA and posthoc analyses yielded significant differences for status(beginning and graduating): graduating students scoredsignificantly higher on assertiveness than beginningstudents.

The significant relations among the study variableswarrant further attention. Experimental manipulation ofvariables to discover causality is recommended.

References

Ashley, J. A. (1976). Hospitals, paternalism, and therole of the nurse. New York: Teachers CollegePress.

Claus, K. E., & Bailey, J. T. (1977). Power and influencein health care: A new approach to leadership. St.Louis: The C. V. Mosby Company.

Donnelly, G. F., Mengel, A., & Sutter ley, D. C. (1980).The nursing system: Issues, ethics, politics. NewYork: John Wiley & Sons.

142150

Heide, W. S. (1973). Nursing and women's liberation:A parallel. American Journal of Nursing, 73824427.

Herman, S. 3. (1978). Becoming assertive: A guidefor nurses. New York: Van Nostrand Company.

Levenson, H. (1972). Distinctions within the conceptof internal-external control: Development of a newscale. Proceedings, 80th-Annual Convention, APA,261-262.

Mauksch,. I. (1980, November-December). Nursing;unique and essential, but powerless. Texas Nursing,pp. 7, 13.

Rathus, S. A. (1973). A 30-item schedule for assessingassertive behavior. Behavior Therapy, 11, 398-406.

Sandford, N. D. (1978). Identification and explanationof strategies to develop power for nursing. In Amer-ican Nurses' Association, Power: Nursing's-challengefor change. Papers presented at the 51st Convention.

Spence, J. T., Helmreich, R. L., ac Stapp, J. (1974).The Personal Attributes Questionnaire: A measureof sex-role stereotypes and masculinity-femininity.Journal Supplement Abstract Service Catalog ofSelected Documents in Psychology, 4 43-44.

Stuart, G. W. (1981). How professional is nursing? Image,13 (1), 1843.

143

NURSING FACULTY PERCEP IONSOF AN INTEGRATED CURRICUL M AND

IMPLEMENTATION OF THE CURRICULUM

Corrine Strandell, RN, PhIndependent Practice and Consultation

- Evanston; Illinois

The purpose of this study was to eiamine nursingfaculty perceptions of an integrated curriculum and theimpact that the perceptions have upon the actual imple-mentation of the integrated curriculum. The twofoldproblem investigated was:

1. To examine baccalaureate nursing faculty per-ceptions of the integrated curriculum in terms of(a) what their perceptions of the concept of theintegrated curriculum are, (b) what they per-,ceive their current integrated curriculum is, and(c) what, they perceive their current integratedcurriculum should be.

2. To examine the relation between the nursingfaculty's perception of their current integratedcurriculum and the actual implementation of thecurriculum.

The problem was further investigated using as correlateithe faculty members' professional experience, education,experience in curriculum planning, and involverrient inplanning the current integrated curriculum as they relateto faculty's perception of the curriculum.

Participants were faculty members from bacca-laureate nursing programs that were accredited by theNational League for Nursing. A survey was conducted toidentify the programs that had a fully integrated cur-riculum design. Torres' definition (1974, p. 2) of anintegrated curriculum ("an integrated curriculum refersto the unifying of the content in such a way that the

1" 144

parts or specialties are no longer distinguishable") wasused in the study. A stratified random sampling _pro-cedure was utilized to select 12 participating schoolsfrom the 111 respondents which indicated that they hada fully integrated curriculum.

The study included the development of the instru-ments for collecting the data and the administration ofthe instruments to the nursing faculty from the -12schools. The first instrument, "Integrated Curriculum inNursing Inventory," measured (a) nursing faculty per-ceptions of the concept, "integrated curriculum," (b)perceptions of what their current integrated curriculumis, and (c) perceptions of what their current integratedcurriculum should be. The reliability coefficients(Cronbach's alpha) for the three parts of the inventorywere: Part A, .929; Part B, .950; and Part C, .951 (N =186). The second instrument, "Integrated CurriculumImplementation Inventory," was used to measure thedegree of implementation of the integrated curriculum.This instrument consisted of criteria statements thatwere utilized by the_ investigator and two other raters :ndoing content analysis of the documents of both thecurriculum and curriculum implementation of the 12participating nursing schools. Each school was given aconsensus implementation score. All instruments werepilot-tested prior to use.

Based on the findings of the study, the followingconclusions were made:

1. There were significant differences in nursingfaculties' perceptions of the integrated curric-ulum, both by geographical area groups and byschool groups (p < .01). The dependent variablethat contributed most to this significance wasfaculty perceptions of what their current inte-grated curriculum is.

2. A significant direct relationship existed betweeneach of the parts of the "Integrated Curriculum

145 154

in Nursing Inventory:" Parts A and B, r = .624,p <.05; Pa_rts A and C, r = .819, p < .01; Parts Band C, r = .586, p <.05.

3. There were significant differences in nursingfaculty perceptions of their current integratedcurriculums and the actual implementation ofthe_curriculum (p <.01).

A significantdirect relationship (r = .685,_p <.05)existed between the actual implementation ofthe curriculum and the nursing faculty per-ceptions of what their current integrated cur-riculum is:

5. There were no significant differences in nursingfaculty perceptions of the integrated curriculumin relation to educational level, professional ex-perience, years of experience in planning nursingcurriculum, and degree of participation in plan-ning the current integrated curriculum.

6. The actual- number of years the integrated cur-riculum had been implemented was unrelated tothe degree of implementation of the curriculum.

Some recommendations include:

1. Graduate programs in nursing sho-Uld include the-ory on curriculum development, including theintegrated curriculum.

2. Nursing faculty understanding of the integratedcurriculum should be included when intek viewingfaculty to teach in these programs.

3. Time needs to be allotted for faculty dev*lop-ment and discussion of problems related to theimplementation of the integrated curriculum.

4. Qualitative studies might aid in describing theproblems that continue to exist in implementingthe integrated curriculum.

155 146.

Torres, C (1974). Educational trends and integrated cur-riculum approach in nursing. in Faculty-curriculumdevelopment, part 4. Unifying the curriculum--the

- integrated approach. Publication number 15 -1522.New York: National League for Nursing.

156

PROGRAMMATIC RESEARCH: A COLLABORATIVEMODEL FOR DEVELOPING NURSING THEORY

Patricia Tomlinson, RN, PhDAssociate Professor, Mental Health Nursing

Project DirectorParent/Infant Research Project

Joyce Semradek, RN, MSNAssociate Professor, Director

Office of Research Development and UtilizationSchool of Nursing

The Oregon Health Sciences University, Portland

Theory-building collaborative research is proposed asan alternative to the more competitive research effortswhich are part of the traditional academic role. Thispaper describes a collaborative model of programmaticnursing research designed to build upon previous Indi-vidual research and maximize contribution of individualinvestigators through a carefully designed four-year pro-gram; This program involves faculty in different clinical,departments In examining a population and exploringhealth variables that contribute to nursing theory fromperspectives unique to each contributing investigator's/specialty: women's health care, maternity nursing, men-tal health nursing, and community health systems.

The program consists of four concurrent longitudinalstudies, with a total sample size of over 600 urbanfamilies. forming the nUcleus. The' focus is on theinteraction between family relationships, social change,health variables, and health outcomes in the beginningfamily. Although the studies, use separate cohorts, the'patterns of Investigation include two dimensions: longi-ttxEllyna , marital_ _and_family relationships throughoutpregnancy and early 'parenthood are examined with anexploration of the effect, of the parent on the child andthe effect of the child on the family; cross-sectionally,the effects of social, psychological, and physiologicalstressors at various stages of the perinatal period are

157 148

explored. The investigations are designed specifically tocontribute to the theoretical constructs of attachmentsand interactions within the family, social support, and."

stress. The program is also designed to make method-ological contributions through instrument developmentand validity testing of existing measures.

Four issues were identified whose management maybe fundamental to any successful collaboratiVe researchprogram among multiple investigators. The issues Iltre:(1) differences in individuals' value systems; (2) conflicts-between individual and group goals; (3) variations in

commitment to the program, vis-a-vis otherprofessional demands; and (4) conflicts over control andownership. The importance of recognizing and dealingwith these issues is accentuated by the interdependenceof the projects in programmatic research, but the issuesare likely to surface in other types of collaborativearrangements d's well.

Biweekly work sessions provided opportunities for,discussing and resolving problems that arose, in imple-menting the program. In these sessions the followingstrategies for dealing with the collaboration issues wereused and 'found to-be effective:

1. Early organizational -planning to define grOupgoals and to identify potential conflicts betweenindividual and group goals.

2. Acknowledgment of potential conflicts and varia-tions in individual expectations and commitments.

3. Development of a " plan whereby the teammembers help each other achieve individualgoals and permits flexibility in level of commit-ment without jeopardizing the quality of theteam effort.

4. Open discussion of power issues and sharing oleadership.

14.9 138

PARENTAL PERCEPTIONS OF INFANT,BEHAVIOR AND RHYTHMIC PATTERNS

OF THE PARENTS. AND THE INFANT

Sharon A. Wilkerson, RN, PhD,Director of Nursing Research

St. Joseph Mercy HospitalPontiac, Michigan \

Adjunct Assistant ProfessOr,Wayne State University College of Nursing

Detroit, 'Michigan

Parental perceptions of the infant can affect parent-child interactions as well as -the development of the child(Blank, 1964; Broussard and Hartner, 1971; Snyltar,-Eyres,ac Barnard, 1979). Parent-infant interactions are com-plex in nature (Sander, Stechler, Julia, ac Burns, 1976);therefore, it is important for nursing to consider thevarious factors, that may influence parental perceptionsand interactions. This study identified the individualphenomena of personal tempo as potentially influencing-parental perceptions of their infant.

The purpose of this descriptive correlational study wasto identify rhyttimi-C patternS in, parents and infants thatcontribute to synchrony or dysynchrony between them.'These .rhythmic patterns were -then related to parents'perceptions of their infants. .

Conceptualized from Newman's (1979) framework for.nursing, personal tempo, as measured by rate of fingertapping, and time perteption, "as - measured by productionof a time interval, were considered components of theparents' rhythmic structLire. Braun (1927) found that therelative speed of finger tapping for each subject- wasconsistent over a 3-day period (r = £7 to .94). Produc-=,tion of intervals of clock time ws identified asi,reliable

1. 5 0'

159

instrument for measurement of time perception. Thetest-retest reliability of production was .62 to .92(Clausen, 1950). Sleep-wake activity, as recorded byparents on the Nursin3 Child Assessment Sleep/ActivityRecord_(Barnard, 1979), was used to measure the infant'srhythmic structure. A dysynchrony score was producedfrom these parent and infant tempo scores.

The parents' perceptions of their infant were mea-sured by the Perception of Baby Temperament (PBT)-scale (Pedersen,. Zaslow, Cain, Anderson, & Thomas,1977). This card-sort procedure, which has a split-halfreliability of .68, measures perception of the baby'smood, rhythmicity, activity, approach, and adaptability.A second instrument was developed for the study, theParent Report of Infant Behavior (DRIB), which mea-sured the parents' perception of their infant as comparedto their idea of an average baby. The items includedperceptions of amount of sleep, activity, crying, caretime, difficulty in feeding, and regularity of schedule.Experts in the field of infant behavior evaluated itemsfor content validity.

A self-selected sample of volunteers from a con-.venience sample was chosen. The sample included 100parents of 50 healthy 6-month-olds. Data were collectedby self-administered questionnaires; tempo me,,bure-ments were taken by data collectors during home visits.All variables were examined by descriptive analysis aswell as by Pearson product moment correlations. Mul-tiple regression was used to analyze the relationshipbetween discrepancies of parental perceptions and per-sonal tempo. The dependent t-test for matched pairswas used to analyze differences between mothers' andfathers"Tespqt. onses2for all variables.

Support was found for the relationship between ac-tivity dysynchrony and perception ,r_screpancies for themothers (r = .37, 2 and r = .24, p. <.01). Fathers'perceptions of infants varied from the mothers' and were

151160

less influenced by their rhythmic structures. Parentalage, experience with children, and the birth order of theinfant were found to correlate with perceptions. Fatherssaw their babies as crying more thin the mothers did (t

p <.01), but the mothers saw the babies as needingmore. care than the fathers did (t = 6.05, p < .01).Additional statistics compared working and non-workingmothers. The employed- mothers had a higher personaltempo (t(48) = 2.49, p < .05), and they perceived theirbabies as crying less and needing less care than did theunemployed mothers. The families were grouped ac-cording to synchrony patterns (4 types) and analysis ofvariance was made. Fathers and mothers in synchrony,but with the baby not, accounted for the most variancein the perception discrepancy scores.

4This study demonstrated that the evaluation of fam-

ily synchrony and parental perceptions is useful to nursesin the promotion of optimum infant development. Ap-propriate-interventions based on findings can be plannedto improve parent-child interactions during well-childconferences.. Family behavior patterns may be alteredto increase synchrony within the family that demon-strates disruptive patterns. The perception instrumentscan also be used by the nurse to increase parent aware-ness of infant behavior. This study reinforced the needfor prenatal classes, especially for new parents who havenot had previous experience with babies

References

Barnard, K. (1979). The nursing child assessment sleepactivity record. University of Washington: NCAT.

Blank, M. (1964). The mother's role in infant development.Journal of American Society of Child Psychiatry,3, 89-105.

Braun, F. (1927). UntersuChungen uber das personlichetempo. Archiv fur Gesamte Psychologie, 60, 317-360.

161 152

Broussard, E. R., & Hartner, M. S. S. (1971). Furtherconsiderations regarding maternal perceptions'ofthe first born. In 3. Hellmuth (Ed.) Exceptional in-fant Vol. Q. Studies in abnormalities. New York:Brunner /hazel.

Clausen, 3. (1950). An evaluation of experimental methodsof time judgement. Journal of Experimental Psy-chology, DJ 756-761.

Newman, M. (1979). Theory development in nursing.Philadelphia: F. A. Davis Co.

Pedersen, F. A., Zaslow, M., Cain, R. L., Anderson,B. 3., & Thomas, M. (1977, March). A methodologyfor assessing parent perception of baby temperament.Paper presented at the meeting of the Society forResearch in Child Development, New Orleans.

Sander, L. W., Stechler, G., 3ulia, H., ac Burns, P. (1976).Primary prevention and some aspects of temporalorganization in early infant- caretaker interaction.In E. N. Rexford, L.W. Sanders, and T. Shapiro (Eds.)Infant psychiatry. New Haven: Yale UniversityPress.

Snyder, C., Lyres, S. 3., & Barnard, K. (1979). Newfindings about mothers' antenatal expectations andtheir relationship to infant development. MaternalChild Nursing,A, 354-357.

153 162

BLOOD PRESSURE AND HEART RATECHANGES DURING CARDIAC CATHETERIZATION

Fran Wimbush, RN, MSInstructor, School of Nursing

University of Maryland at Baltimore

Ellen Sappington, RN, MSMedical Program Coordinator

Merck, Sharp, and DohmeResearch Laboratories

West Point, Pennsylvania

Erika Friedmann, PhDAssistant-Professor

Brooklyn College of the CityUniversity of New York

- Sue Ann Thomas, RN, PhDAssociate Professor, School of Nursing

-lames a. Lynch, PhDProfessor, School of Medicine

University of Maryland at Baltimore

Cardiac catheterization is an invasive diagnostic pro7cedure. During this procedure there is constant moni-toring of the heart's internal pressures, rate, and rhythm.Several studies have suggested that certain phases of theprocedure are more painful and stressful than others.injection of the local anesthetic, insertion of the cathe-ter, table turning, dye injection, and left ventriculogramwere cited as most stressful.

The purpose of this study was to compare the bloodpressure and heart rate changes during' an acute clinicalprocedure with those occurring during verbal commu-_nication. Recent studies (Friedman, et al., 1982; Long,et al., 1982; Lynch, et al., 1980, 1981,1982;& Thomas, etal., 1975, 1982) have shown significant increases in heartrate and blood pressure during verbal communication in

163

15;1

non-stressful situations. In this study the cardiovaSculareffect of the clinical procedure was evaluated.. Inaddition, the effects of verbal communication during theprocedure were examined. Changes in peripheral bloodpressure and heart rate during specific phases of thecardiac catheterization were studied and compared with.the effects of casual conversation on the day aftercardiac catheterization. A convenience sample of 14subjects who were scheduled for cardiac cathaterizationprocedures were recruited to participate in this study.After explaining the study, patients' informed written*"consent was obtained. Blood pressure (BP) and heartrate (HR) were obtained using an automatic °sell-.lometric BP device, the Dinamap 845. BP and HR weremeasured at selected critical intervals throughout thecatheterization.

Subjects were asked to talk for two minutes duringthe catheterizations. This period was preceded andfollowed by two-minute quiet resting periods. On thefollowing day the 'same nurse who was present at thecatheterization visited the patient's room and recordedBP and HR during a six minute series: two minutes quietrest, two minutes talk, two minutes rest.

The highest BP obtained during catheterization oc-curred with the injection of local anesthetic and leftventriculogram. The highest HR obtained during Cath-eterization occurred at the time of the left ven-triculogram. The test selected to examine the dif=ferences in HR and BP was the Newman-Kdels. Bloodpressures taken while talking to the nurse the next daywere significantly lower (p < .05) than those taken duringcatheterization. The heart rates during left ven-triculogram were significantly higher (p<.05) Than duringother stressful events.

These results show that BP and HR obtained duringthe most stressful events of a cardiac catheterization areno higher than those obtained during casual conversation

155 164

with a nurse. Even a casual conversation betWeen _anurse and patient during a procedure such as a cardiaccatheterization could have direct consequences on thepatient's BP. While the presence of the nurse Is sup-portive, the patient should not be encouraged to talkduring an acute clinical procedure.

References

Friedmann E., Thomas, S. A., Kulick-Ciuffo, D., Lynch,3., drSuginohara, M. (1982). The effects of normal

and rapid speech on blood_pressure. PsychosomaticMedicine, 44 545-553.

Long, J. M., Lynch, J. 3., Machiran, N. M., Thomas,S. A. ac Malinow, K. L. (1982). The effect of statuson blood pressure during verbal communication.Journal of Behavioral Medicine, S 165-172.

Lynch, 3. 3., Long, J. M., Thomas, S. A., Malinow, K.,& Katcher, A. (1981). The effecfs of talking onthe blood pressure of hypertensive and normotensiveindividuals. Psychosomatic Medicine, 112, 25-33.

Lynch, J. 3., Thomas, S. A., Long, 3. M., Malinow, K.L., Chicadonz, G., & Katcher, A. (1980). Humanspeech and blood pressure. 3ournal of Nervous andMental Disease. 161, 526-534.

Lynch, J. 3., Thomas, S. A, Long, J. M., Malinow, K.,Friedmann, E., dc Katcher, A. H. (1982). Blood pres-sure changes while talking. Israeli 3ournal of MedicalScience, 18 575-579.

Lynch, J. 3., Thomas, S. A., Paskewitz, D. A., Malinow,K. L., & Long, J. M. (1982). Interpersonal aspectsof blood pressure control. 3ournal of Nervous andMental Disease, 170 143-153.

165 156

Thomas, S..A., Friedmann- , E., Noctor, M., Sappington,E., Gross, H., dc Lynch, 3. 3._(1982). Patient's cardiacresponses to nursing interviews in a CCU. Dimen-sions in Critical Care Nursing, 1, 198-205.

Thomas, S. A., Lynch, 3. 3., do Mills, M. E. (1975).. Psycho=social influences on heart arrhythmia in a coronarycare patient. Heart and Lung, 119 746450-

157166

3

PAPER REVIEW' COMMITTEE

Marion Anema, 'RN, PhDCollege of Nurs1ngTexaslioman's University, Denton

Kay Avant, RN, PhDSchool of. NursingUniversity- of. Texas at Austin

Beverly- Baldwin, 'RN, PhDSchool of Nursing _

University of-Maryland at Baltimore

Barker Bausill, PhDSchool of NursingUniversity -of' Maryland at Baltimore

Rachel Bodth, RN, PhDSchool of NuisingUniversity of- Maryland at Baltimore

Kathy Brown, 'RN, PhDSchool of NursingUniversity of Alabama in Birmirigham

Helen Bush, RN, PhDCollege of NursingTexas Woman's. University, Denton

Barbara Byfield, RN, MSSchool otNars:ngUniversity of Maryland at Baltimore

Janet Cogliano, RN, DNScSchool of NursingUniversity of Maryland at Baltimore

Shirley Damrosch, PhDSchool of NursingUniversity of Maryland at Baltimore

Beverly Henry, RN, PhDSchool of NursingUniversity of Alabama in Birmingham

167 158

/Ada Jacox, RN,, PhDSchool of- NursifitUniversity of /Maryland. at Baltimore

Margie Johnson, RN, PhDCollege of NursingTexas Wornan's University, Denton

Karen Kleeman, RN,ThDSchool of NursingUniversity of Maryland at Baltimore

Helen Kohler, RN, PhDSchool of NursingUniversity of Maryland at Baltimore

Mildred Kreider, RN, PhDSchool of NursingUniversity_of Maryland at Baltimore

Elizabeth Lenz, RN, PhDSchool of NursingUniversity of Maryland at Baltimore

Margaret McEntee, RN, MSSchool of NursingUniversity of Maryland at Baltimore

Wealtha McGurn, RN, PhDSchool of NursingUniversity of Maryland at P-dltimore

Cynthia Northrop, RN, 1D -School of NursingUniversity of Maryland at Baltimore

Peggy Parks, PhDSchool of NursingUniversity of. Maryland at Baltimore

Lesley Perry, RN, PhDSchool of NursingUniversitrof Maryland at Baltimore

159 168

1

Patricia Prescott, RN, PhD.School of NursingUniversity-of Maryland at Bal imore

Mary Rapion, RN, PhDSchool of, Nursing -

Udiversity,of Maryland at altimore

Bonnie Rickelman, RN, DSchool of NursingUniversity otTeicas at Austin

1

Joanne Scungio, RN, PhD.School, of NursingUniversity of Alab ma in Birmingham

Betty Shubkage RN, PhDSchool of NurehgUniversity of aryland at Baltimore

Karen Soe n, PhDSchool of ursingUniversi of Maryland at Baltimore

EdhaSchUn

tilwell, RN, PhD1 of Nursing

ersity of Maryland at Baltimore

ra Strickland, RN, PhDSchool of NursingUniversity of Maryland at Baltimore

Lorraine Walker, RN, EdDSchool of NursingUniversity of 7exasat Austin

Constance Walleck, RN, MSSchool-of NursingUniversity of Maryland at Baltimore

Carolyn Waltz;RN, PhDSchool of NursingUniversity of Maryland at Baltimore

169 160

CONFERENCE PLANNING 'COMMITTEE

Chairperson

Ora L. Strickland, RN, PhD, FAAN, Associate Professor,.University of Maryland at Baltimore School of Nursing

Members

Beverly Baldwin, RN, PhD, Assistant Professor, GraduateFaOlty, University of Maryland at Baltimore Schoolof Nursing

Barbara Byfield, RN, MS, Assistant Dean for ContinuingEducation and Faculty Development, Universityof Maryland at Baltimore School of Nursing

Shirley Darnrosch, PhD, Assistant Professor, Centerfor. Research, University of Maryland at BaltimoreSchool of Nursing

Mary Rapson, RN, PhD, Chairperson, Junior Year Under-graduate Studies, University of Maryland at Balti-.more 'School of Nursing

161171