Research on Nurse Staffing in Hospitals; Report of the Conference, May 1972

DOCUMENT RESUME

ED 101 183 CE 002 952

AUTHOR Levine, Eugene, Ed. TITLE Research on Nurse Staffing in Hospitals; Report of the Conference, May 1972. INSTITUTION National Institutes of Health (DREW), Bethesda, Md. Div. of Nursing. REPORT NO DREW-(NIH)-73-43-1 PUB DATE May 72 NOTE 193p.; Report of the conference (Fredericksburg, Virginia, May 23-25, 1972) AVAILABLE FROMSuperintendent of Documents, U.S. Government Printing Office, Washington, D.C. :n402 (Stock Number 1741-00055, :2.35)

EDRS PRICE MF-80.76 MC-89.51 PLUS POSTAGE DESCRIPTORS Administrative Problems; *Conference Reports; Costs; Evaluation Criteria; Facility Requirements; *Hospital Personnel; Human Services; Information Systems; Medical Services; Nurses; *Nursing; Patients

(Persons); Personnel Needs; *Research; Research . Methodology; Research Needs; Speeches; *Staff Utilization

ABSTRACT The conference brought together 45 persons who have had extensive experience in nurse staffing research. After the leadoff paper by Myrtle K. Aydelotte, which presented an overview of nurse staffing research, 10 papers werepresented on variables considered to be significant in influencing the quantitative and qualitative demand for nurse staffing. These included patients' requirements for nursing services, architectural design of the hospital, administrative and cost factors, andsocial-psychological factors. In addition, papers were presented on the evaluation of quality of nursing care and the impact of computerized information systems on staffing. The papers were followed by presentationsby discussion leaders who addressed themselves to points raised in the papers. These presentaions, the openingremarks of Jessie N. Scott, and a summary and synthesis of the entire conference are also included in the publication. Appended are four task force reports offering recommendations on future directions for research. (Author/MW) 6 BEST COPY AVAILABLE in

U S. DIPARTMINT OP HEALTH, EDUCATION & WILPARS NATIONAL INSTITUTE OP EDUCATION THIS DOCUMENT HAS SEEN REPRO OUCEO ExACTL Y AS RF.CEIVEO PROM THE PERSON OR ORGANIZATION ORIGIN MING IT POINTS OF VIEW OR OPINIONS STATED 00 NOT NECESSARILY REPRE SENT OFFICIAL NATIONAL INSTITUTE OF EDUCATION POSITION OR POLICY

rch on \urse Stactg in Hosottals Report of the Conference i'robibitedTit le of the Civil Rights ,\(-t ol I!H states; "No etson in the!tined states shall. on the ground of um C. Color, or national origin, he excluded how patticipationIII,be denied the Lic.nefits ol, or he scil.ijected to discrimination 1111(h.)' any, progiam or ii(iiity teceiving Federal financial assistan(e." There- tore, the ntirse stalling progiant, like eve, y program or aclivily receiving financial as5istan«e nom the 1)cp11r11tent of I lealth, 1(111- tat ion. and 11'ellarc, must he operated in compliance with this law.

1))1%.11('I) 111).S1111e1.1111(1)11(.111of 11))1Inrnrnt., 41,(,11%rititilr111 Ptin,ing I) (.1.1)1112 Piny. ':' ';`1 Sin(I, \mullet 1711 mu)) Research on Nurse Staffing in Hospitals Report of the Conference MAY 1972

Conducted by DIVISION OF NURSING

Eugene Levine, Ph.D. Scientific Editor

DHEW Publication No. (NIH) 73-434

U.S. DEPARTMENT or HEALTH, EDUCATION, AND WELFARE PUBLIC HEALTH SERVICE NATI.:NAL INSTITUTES OF HEALTH Bureau of Health MON.te"orf Education Division of Nursing Foreword This publication contains a report on the Conference on Research on Nurse Staffing in Hospitals, sponsored by the Division of Nursing and held in Fred- ericksburg, Virginia, May 23 through 25, 1972. I am hopeful that it will prove to be a valuable addition to the literature on nurse staffing. As the opening paper by Dr. Myrtle K. Aydelotte makes clear, the existing literature has many deficiencies, not the least of which are the many gaps in knowledge. The papers presented at the conference contain the latest findings on many important areas of nurse staffing research and will help to close they! gaps. The Division of Nursing's concern with the problems of nurse staffing goes back many years as we have attempted to improve the quality and quantity of nursing care in the Nation. We have supported considerable research in this area, both by our own staff and through grants and contracts to researchers in universities and professional organizations and in hospitals and other health care settings. The many publications issued by the Division in addition to study findings include descriptions of methodology for studying the use of nursing skills, assessing the quality of nursing care, and planning for better matching of nursing resources and needs. The conference and this resulting publication are natural extensions of our efforts to improve nurse staffing. This publication is a companion volume to Dr. Avdelotte's comprehensive critique of nurse staffing methodology. These two publications should lie useful to researchers, educators, administrators, practitioners, and all who are concerned with the improvement of nursing staffing in hospitals.

Juan M. Scorr Assistant Surgeon General Director Division of Nursing

ill

E. CONFERENCE, PLANNING COMMITTEE

Dr. Eugene Levine, Chairman

Dr. Myrtle K. Aydelotte

Dr. Bernard Ferber

Miss Marie R. Kennedy

Miss Delores M. LeHoty

Dr. Mary K. Mullane

Miss Jessie M. Scott

Mrs. Margaret Sheehan

Mr. Stanley E. Siegel

Miss Elinor D. Stanford

iv CONFERENCE PARTICIPANTS

Dr. Myrtle K. Aydelotte Dr. Charles D. Flag le Director, Nursing Director of Operations Research University Hospitals and Clinics School of Hygiene and Public Health Professor, College of Nursing Johns Hopkins University Hospital University of Iowa Baltimore, Maryland Iowa City, Iowa Mr. Richard G. Gardiner Dr. Alexander Barr Operations Research Specialist Chief Records Officer and Statistician Hospital Association of New York State Oxford Regional Hospital Board A1L.iny, New York Headington, Oxford England Mrs. Phyllis Giovannetti Project Director Mr. Jose Blanco, Jr. Nursing Research Controller and Assistant Administrator Hospital Systems Study Washington Hospital Center L niversity of Saskatchewan Washington, D.C. Saskatoon, Saskatchewan Dr. Doris Bloch Canada Executive Secretary Mr. John Griffith Nursing Research ant! Education Professor and Director Advisory Committee Program in Hospital Administration Division of Nursing School of Public Health National Institutes of Health University of Michigan Bethesda, Maryland Ann Arbor, Michigan Dr. Mario F. Bognanno Assistant Professor Miss Verna Grovert University of Minnesota Stall Assistant to the Director Industrial Relations Center Health Systems Research and Minneapolis, Minnesota Development Service Veterans Administration Central Office Dr. Virginia S. Cleland Washington, D.C. Prof essoa. of Nursing Wayne State University Dr. Elizabeth Hagen Detroit. Michigan Professor of Psychology Mrs. Margaret Ellsworth Teachers College 763 St. Charles Street Columbia University Elgin, Illinois New York, New York Dr. Bernard Ferber Dr. Richard Howlancl Director, Bureau of Research Services Westinghouse Elect sic Corporation American Hospital Association Rescarch and Development Chicago, Illinois Pittsburgh, Pennsylvania

v 7 Conference Participants (continued)

Dr. E. Gartly Jaco Dr. Wanda E. McDowell Professor, Department of Sociology Associate Dean for Research University of California University of Michigan Riverside, California School of Nursing Dr. Stanley Jacobs Ann Arbor, Michigan American Hospital Association Dr. Mary K. Mullane Chicago, Illinois Professor, College of Nursing University of Illinois Dr. Richard C. Jelinek Chicago, Illinois Vice President for Research and Development Medicus Systems Corporation Mr. George J. Mullen Chicago, Illinois Senior Editor Engineering Services and Publications Miss Angie Kammeraad Gaithersburg, Maryland Regional Nursing Consultant Division of Nursing Dr. Maria Phaneuf Professor of Nursing National Institutes of Health College of Nursing DHEW Region IV Wayne State, University Atlanta, Georgia Detroit, Michigan Miss Marie R. Kennedy Dr. Lillian M. Pierce Nurse Consultant Professor, School of Nursing Nursing Education Branch The Ohio State University Hospital Division of Nursing Columbus, Ohio National Institutes of Health Bethesda, Maryland Miss Pamela Poole Nursing Consultant Miss Dolores M. LeHoty Department of National Health and Welfare Program Analyst Ottawa 3, Ontario Manpower Evaluation and Planning Branch Canada Division of Nursing Dr. Muriel Poulin National Institutes of Health Bethesda, Maryland Boston University School of Nursing Boston, Massachusetts Dr. Eugene Levine Dr. Reginald W. Revans Chief, Manpower Evaluation and Planning Branch Senior Research Fellow an -1 Scientific Adviser Division of Nursing Fondation Industrie-University National Institutes of Health 1030 Brussels, Belgium Bethesda, Maryland Miss Jessie M. Scott Miss Ruby M. Martin Assistant Surgeon General Director of Nursing Services Director, Division of Nursing The Ohio State University Hospital National Institutes of Health Columbus, Ohio Bethesda, Maryland Miss Marylou McAthie. Mrs. Margaret F. Sheehan Regional Nursing Consultant Nurse Consultant Division of Nursing Nursing Education Branch National Institutes of Health Division of Nursing DHEW ".egion IX National Institutes of Health San Fran4 isco, California Bethesda, Maryland

8 Conference Participants (continued)

Mr. Stanley E. Siegel Miss Elinor D. Stanford Assistant Chief, Manpower Evaluation Director, Nursing Research Field Center and Planning Branch Division of Nursing Division of Nursing National Institutes of Health National Institutes of Health San Francisco, California Bethesda, Maryland Miss Bernice Szukalla Dr. Donald W. Simborg Regional Nursing Consultant Chief, Clinical Systems Development Division of Nursing Johns Hopkins University Hospital National Institutes of Health Baltimore, Maryland DHEW Region VIII Miss Marjorie Simpson Denver, Colorado Nursing Officer in Research Department of Health and Social Security Miss Deane B. Taylor Associate Professor Alexander Flemming House State Universir' of New York at Buffalo London, England School of Nursing Miss Marlene R. Slawson Buffalo, New York -Research Associate Operations Research Unit Dr. Mabel A. Wandelt Division of Community Psychiatry Professor of Nursing School of Medicine Wayne State University State University of New York at Buffalo Detroit, Michigan Buffalo, New York Dr. Gerrit Wolf Dr. Frank A. Sloan Assistant Professor of Administrative Del art ment of 'Zconomics Sciences and Psychology Uri; versity of Florida Yale Universit.) Gainesville, Florida New Haven, Connecticut Dr. Harvey Wolfe Associate Professor School of Engineering University of Pittsburgh Pittsburgh, Pennsylvania

vii Contents

Page Foreword Conference Planning Committee iv Conference Participants Figures Tables xi Introduction 1 Opening Remarks 3 Miss Jessie M. Scott State of Knowledge: Nurse Staffing Methodology 7 Myrtle K. Avdelotte, Ph.D. Factors Related to Nurse Staffing and Problems of TheirMeasurement__ 25 Charles D. Flagle, Dr.Eng. Discussion of Dr. Flagle's Paper 35 Measurement of Patients' Requirements for NursingServices 41 Mrs. Phyllis Giovannetti Discussion of Mrs. Giovatinetti's Paper 57 Nurse Staffing Patterns and Hospital Unit Design:An Experimental Analysis 59 Dr. E. Gartly jaco Discussion of Dr. Jaco's Paper 77 Impact of Administrative and Cost Factors on NurseStaffing 79 Mr. John R. Grimth Discussion of Mr. Griffith's Paper 98 Nursing Directors as Participants in Hospital PeerReview Boards 101 Dr. Gerrit Wolf Social Psychological Factors Relating to the Employmentof Nurses 111 Dr. Virginia S. Cleland Discussion of Dr. Wolf's and Dr. Cleland's Papers 122 Evaluation of Quality of Nursing Care 125 Dr. Mabel A. Wandei: Discussion of Dr. Wandelt's Paper 133 Psychosocial Factors in Hospitals and Nurse Staffing 137 Dr. Reginald W. Revans Impact of Computerized Information Systems on N .:seStaffing in Hospitals 151 Donald W. Simborg, M.D.

10 Contents (continued)

Title Page

Discussion of Dr. Simborg's Paper ._ 164 Concept Modeling in Hospital Staffing 167 Dr, Lillian M. Pierce

Discussion of Dr. Pierce's Paper. _ _ _ _ - _ 174

Summary, Synth's's, and Future Directions _ 177 Dr. Mary K. Mullane

Appendix: Task Force Reports . 181 1

Figures

Title Page Scheme of nursing care delivery system 9 Diagram of evolvement of methodologies for the study ofnurse staffing: application of logic and increases it abstractions 15 Calculation of nursing staff needs 28 Interrelationship of staffing and patient care Patient classification tool____ 45 Graph of patient load index 49 Graph of daily workload index 50 Nursing: An interdependent variable in the health cattlystem 52

Radial and angular hospital units. 61 Marginal utility of nursing care 82 High and low region medium AM &.S nursing hours per patient day 86 Prototype outcomes quality assessment in .trument, patient observation section, diabetes mellitus 89

Flow chat of weight order_ _ _ 153 Flow chart of vital sign order 154 Flow chart of intravenous fluids order_ 155 Physicians' orders study 156 Functional requirements of a new system to carry out physicians' orders__ 156 General flow chart of new system 158 System output documents 158 Team action sheet 159 Patient care model 170 Tables

Title Page Comparison of staffing ,methodologies on selected features _ - _ 16 Ratios of direct nursing care time received per shift and patient category ______Daily patient load index 47 Conversion of index to staff required _ 47 Major type of nursing care in an acute, general hospital 65 Major type of nursing care, by six nurse staffing patterns ______. 66 Major type of mirsing (are, by nurse staffing patterns and by radial and angular units, percentages only 66 Medical, physical, and social types of direct care, by number of observations and percentages, total study, by shape of unit _ 69 Type of direct patient care, by nurse staffing patterns _ 69 Type of direct patient care by nurse staffing patterns, by radial and

...... 711. angular units, percentages o n l y...... Ratios of patient involved to nonpatient involved care, by totalstaff and by nurse staffing patterns 70 Ratios of patient involved to nonpatient involved care, by nurse staffing patterns and radial and angular units _..______71 Types of trips to patients' mums by nursing staff in radial and angular 71 units, by number of observations and percentages __ _ _ Average daily number of trips to patients' rooms by nurse_ staffing patterns, by total trips and by types of trips 72 Average, daily number of trips to patients' rooms by types of trips, by nurse staffing patterns, and by radial and angular units 72 Utilization of nursing unit by total nursing staff in radial and angular

units, by number of observations and percentages ___ _ _ 72 Utilization of nursing unit by nurse staffing patterns, by number of observations and percentages 73 Utilization of nursing unit by nurse staffing patterns and radialand angular units, by percentage of observations only 74 Location of MST's employed in hospitals in relation tohospital beds: 1968 84 Medan nursing hours per adult medical surgical patient day, HAS subscriber hospitals, 6 months ending 12/31/71 85 Percentiles of nursing hours per medical surgical patient day in HAS subscriber hospitals, 3 months ending 10/31/71 87 Interrater reliabilities for all strategy items 107 Mean contribution and influence rating according to role (N 15) _ _ 107 161 Nursing activity study_ _ 161 Summary of repeat studies on experimental floor _ _ 161 Sample nursing order: "NG Tube Care" __. _ Total number of nursing tasks on one nursing unit onselected days_ 161

xi Introduction

The Conference on Research on Nurse Staffing in Hospitals brought together .45 persons from a variety of disciplines who have had extensive experiencein nurse' staffing research. The conference was held in Fiedericksburg, Virginia, May through May 25, 1972. After the leadoff paper by Dr. Myrtle K. Aydelotte, which presented an overview of rtrse staffing research, 10 papers were presented on variables considered to he significant in influencing the quantitative and qualitativedemand for nurse staffing. These included patients' requirements fornursing services, architectural design of the hospital, administrative and cost factors, and social 1.sychologkal factors. In addition, papers were presented on theevaluation of quality of nursing care and the impact of computerized information systems oo staffing. Ti,e papers were followed by presentations by discussion leaderswho addressed themselves to points raised in the papers. Thesepresentations, which were planned to stimulate thediscussion during the period that followed each paper, zre also contained in this publication. Also included are the opening remarks of Miss Jessie M. Scott,Director of the Division of Nursing, and a summary and synthesis of the entireconference by Dr. Mary K. Mean:- In addition to p rtici pa ting in the general discussion, each person attending the conference was assigned co a task force. The purposes of the task forces were to delineate areas of needed research in nursestaffing in hospitals, determine research priorities, and explore means of coordinating the work of researchers in this field. Reports of the four task forces are contained in the appendix. Taken as a whole, the material presented in this publication serves several purposes. First, it brings together the latest findings onimportant areas of research into factors related to nurse staffing in hospitals. Second, it points up some of the ;;aps in existingknowledge and suggests how these may be closed. Finally, i; constitutes an important step forward in correcting the meager,diffuse condition of the existing research in hospitals.

Eugene Levine, Ph.D. Chairman, Planning Committee Conference on Research on Nurse Staffing in Hospitals

I 13 Opening Remarks

Miss Jessie M. Scott

Assistant Surgeon General Director, Division of Nursing Bureau of Health Manpower Education National Institutes of Health

Greetings and Welcome

I am very pleased to welcome all of you to the in the field of nurse staffing orbecause of other Division of Nursing Conference on Research on activitiesthat have advanced the knowledge or Nurse Staffing in Hospitals. This conference is the understanding of thefield. We indeed feel we culmination of 6 months of planning by a com- have a truly distinguished group ofparticipants. mittee made up of our Division staff, whose names You represent different professional backgrounds are listed in your programs:Dr. Bernard Ferber; and different geographic settings. Dr, Myrtle Aydelotte, who will serve as conference We are pleased to note the presenceof partici- chairman; and Dr. Mary Kelly Mullane, who will pants from England andCanada. Among you are be summarizer. The committee tells me their work nurse administrators, nurseeducators, and nurse was made much easier bythe excellent cooperation researchers, economists, industrial engineers, sys- they received from each of you in accepting as- tems analysts, statisticians,psychologists, sociolo- signments to prepare papers or to serve as discus- gists, computer scientists, and hospitaladministra- sion leaders and in making your travel arrange- tors. You have writtenabout and have studied ments to the conference center. the problems of nurse staffing deeplyand exten- Each of you was selected as a participant in the sively. We hope you will share with us during the conference because of the research you have done course of the conference youractivities and ideas,

Background of the Conference

The Division of Nursing, which is part of the thrust of our activities is in the areaof nursing Bureau of Health Manpower Education in the education, we have been concerned over the many National Institutes of Health, has as its primary years with the improvementof nursing services. mission the improvement of the quantity and qual- To catalog the many activities that theDivision the ityof nursing manpower. Although the major has been engaged in over the past 20 years in

14 4 RESEARCH ON NURSE STAFFING IN HOSITAIS area of nurse stalling, both inn amtirally and extra- dude wot kload measurement, work simplification, murally. would require much more time thanI task analysis, mathematical modeling, and quality have allotted. butlet me citea few examples. control. In the early 1950's, concerned with the severe In recent years the attention of our Division shortage of nursing personnelinhospitals and staff has turned toward a broadened look at the the widespread desire of hospitals and niksing nurse staffing problem in hospitals. Our statistical administrators to have numerical guidelines for studies have shown that quantitative levels of nurse staffing, our Division initiated a study that became staffinginhospitals areinfluenced by a great known as "the green book." This was a design variety of factors, including the needs of the pa- to test the relationship between nursing care and tients, tin. administrative organization and admini- patient welfare. The central hypothesis ofthis strativestyle,the characteristics of the nursing design was that if nursing care were varied quanti- staff, the physical plant, the economic structure tatively and qualitatively the effects of these changes within and outside the hospital, and a host of could he measured. psychological and sociologicalfactors. A recent This design became the basis for several major study we supported, known as The Illinois Nurse studies. One was directed by our chairman, Dr. Utilization Study, attempted to measure the effects Aydelotte. Another was conducted in Kansas City of many of these factors on nurse staffing in a by the Community Studies Organization. A third sample of 31 hospitals. The project director, Mrs. was conducted by members of our own Division, Margaret Ellsworth, is here with us and I am sure which attempted to establish a relationship be- you will be hearing about the methodology and tween nursing hours availabletopatients, and results of this study in the course of the con- patient and personnel satisfaction with care. ference. In the area of industrial engineering applications Also among our current concerns and activities to nurse staffing, the Division of Nursing has sup- is the matter of the expanded role of the nurse. ported several important projects that have been carried out extramurally. One of these, conducted With the health system undergoing rapid changes at Johns Hopkins Hospital under the direction and the rol.' of the nurse in this system widening, of Dr. Charles Flagle, has inspired considerable itis important for us to anticipate these changes application of industrial engineering methods to and to he prepared to provide the educational the problems of nurse staffing. These methods in- base necessary for such an expanded role.

Objectives of the Conference

We could say the main theme of this conference We are limiting ourselves to nurse staffing in is an examination of the important factors that hospitals withthefullrealizationthatout-of- affect nurse staffing. We are going to hear about hospital health care settings are becoming increas- some of the methodology for measuring the impact ingly important in the delivery of health services. of these factors and we shall examine the findings We are focusing on hospitals because, frankly, most of studies directed toward specificfactors such of the research has been done there. This is so as patient requirements for nursing services, ar- becausehospitalsarethelargest employers of chitectural design, economic costs, administrative nursing personnel, employing over 60 percent of techniques,andpsychologicalandsociological the 748,000 registered nurses actively employed in factors. We shall hear about the problems of meas nursing and over 90 percent of the more than one urement in nurse staffing research, including meas- millionlicensedpracticalnursesandnursing urement of the quality of nursing care. Finally, aides, orderlies and attendants. we shall hear about the work being clone in apply- Ifeelthat much of the researchinhospital ing computerization to nurse staffing and the im- nurse staffing can have useful application in other pact this could have on the quantity and quality health care settings. Specifically,Iseeloin'im- of staffing. mediate and long range objectives of our con- 5 OPENING REMARKS ference.First, Isee thiso oulerence as asupple. for the future for the kinds of studies that should ment to Dr. yole lotte's study entitledN'nr.se Ie undertaken, including suggestions asto how fug i1fellsmi1ogy: A Review And Critiglw of Se- they might he initiated and with what kindsof lected Literature.' In this study, done under con. support. tract with our Division, Dr. .\vtlelotte has identi- This support does not necessarily have to come fied a number of important areas of research on only from the Federal Government.I am often the factors we have cited as related to nurse staff- pleasantly surprised by how intic'.1 useful research ing. Her study has identified the researchers in has been accomplished on limited resources.Al. These areaS, and it was on the basis of her report though the problems of nurse stalling are large. that we were able to identify participants for this useful solutions can sometimes he found through conference and to get a perspective on the work small, modestinvestigations,including doctoral they have been doing.Infact,the research of dissertations and even master's theses, Let us then that we many of those present here todayis discussed in keep in mind as the conference proceeds Dr. Aydelotte's report. So, in a sense this conference' would like to delineate areas for furtherresearch, brings her report to life. The proceedings of this and we invite suggestions about the most prac- conference isa companion volume to Dr.Ayde. tical,feasible, and productive ways of pursuing lode's study. this needed research. Our small group sessions on A second objective is to bring together a group Wednesday afternoon will be specificallydirected of of knowledgeable researchers, management engi- toward this subject and will provide for each neers, administrators, consultants, andeducators you an opportunity to share yourideas. this to exchangeinformation andideas aboutthe Finally, 1 would hope that the proceedings of problems of muse staffing. We shall be particularly conference could provide educationalmaterials concerned with the factors that influence nurse for nursing service administratorsand other practi- staffing, either positively or negatively, as well as tioners as they face their day-to-dayproblems of deal quantitatively orqualitatively.Itis my feeling runtthig their health care instittitions. A.great staffing that a great deal of research and methodological has been written about research on nurse development are going on in various places but in hospitals, but much of itis of a technical and that there is not enough communication among somewhat specialized nature. As Isaidearlier, the different persons engaged in these activities. a major mission of ourDivision is to help advance and I hope One of Dr. Aydelotte cconclusions isthat re- the education of nursing personnel, in writing up their work. only infre- the proceedings of this conferencewill give ad- their in- quently cite previous work. To me this indicates ministrators some assistance in managing a lack of informationexchange among those work- stitutions. in ing in the field. Hopefully, this conference will he I hope you will be moialed to participate the starting point for opening up channels of com- the discussions as much as possible.Each presenta- munication that will continue beyond the con- tion will he followed by a 45-minutediscussion partici- ference. period.I cannot urge you too strongly to A third and most important purpose of the con- pate to the fullest extentpossible during these been selected ferenceis to setforth a progrun of research on discussion periods. Lad) of you has nurse staffing for the future. Weshall learn from because of your extensive backgroundand interest worth- our presentations anddiscussions that some signif- in this field, and I am sure you have many icant research has been conducted. We shallhear while ideas and much information toshare with us. about the methodologies and the findings of these In conclusion, let me say thatI anticipate this projects and from this should be able toidentify conference will be a rewardingand fruitful ex- the important gaps in 'enowledge and methodology. perience for all of us and that it willprovide guide- From our deliberations we can then chart a plan lines for future conferences of thiskind and lay the groundwork for continuingdialogue. Now I U.S.DepartmentofHealth, Education, and Welfare. would like to introduce your conferencechairman, Public Health Service, National Institutes of Health. DHEW Aydelotte is Director l'.5. Dr. Myrtle Ay0elotte. Dr. PublicationNo.(N111) 73-133,Washington,D.C.: and Clinics, and Government Printing Office. January 1973. of Nursing, lJniversity Hospitals

16 6. RESEARCH ON NURSE STAFFING IN HOSPITAIS a professor at the College of Nursing, University and patient welfare conducted during the late of Iowa in Iowa City. As 1 ILive already mentioned, 1950's and recently has completed a comprehensive she was the project director ofa very important critique of nurse staffing methodology. 1 place this study of the relationship betweennurse staffing conference under her most capable direction. State of Knowledge: Nurse Staffing Methodology

Myrtle K. Aydelotte, Ph.D Director of Nursing, University Hospitals and Clinics and Professor, College of Nursing University of Iowa

Introduction

Many individuals working within the delivery 1970 and was completed in December 1971. The and education systems of the health field recognize lengthy,frustrating search of the literature re- the need for a well defined body of knowledge sulted in the identification of over a thousand that can serve as a basis for nurse staffing meth- pieces related to the subject. After screening, 182 odologies. Researchers from several disciplines have of those that seemed most important and represen- attempted to study the effect of one or more vari- tative were selected, reviewed, and critiqued. ables upon staffing, and methodologists have tried The primary aim was to "screen, review, and to develop specificapproachesfor determining critique relevant research, studies, and approaches nurse staffing needs. Often, however, these efforts focused on measuring nursing care needs in the have been carried out in isolation and have been hospitals," Development of the research protocol inadequately reported to the profession. Many have involved preparing statements of objectives, specify- also been fragmented in approach and have not ing criteria for the estimation of relevance and dealt with all the relevant problems. Frequently, value, formulating definitions of terms for research such studies have been prompted by the need for and classification and constructing a glossary of the economic use of personnel and have ignored terminology. Procedures for review and critique variables other than cost, Unfortunately, no person were outlined and followed. or group has assembled and assessed the literature Each research paper was examined to determine on the subject of staffing. its value as a research study with regard to the In May of 1970, I was approached by the Division following: of Nursing with the request that I undertake this the significance of the problem to the field task. Work on the project' began late in June of nursing, to nurse staffing methodology,

'Contract N u m her NI H-70-4193. the theoretical framework for investigation,

18 R RFNEARcu ON NuRsE svm; F !NG nosprrAt s the aspects 1' rtinent to the measurement of above points. Many of them didnot, in any case. nursingare requii villeins and prediction of Import to he scholarly undertakings. nurse staffing needs, These items were read for whatever assistance they couldpro- the relevance of the research designto general vide in chit ifying and identifyingideas that might proiaents of nurse staffing, be used itt approachesto the question of nurse the use of specific innovations,including adap- staffing. tations of other approaches, The variation intypes of literature reviewed the variables treated in thestudy that were makes a general summationof the analysis dif- significant tothe methodology of staffing, (alt. The materialswere diverse in many re- the tools and instrumentsmost easily trans- spects. They varied in the purposes forwhich they lated into present dayuse, were written; in publication outlet; in thedetail of documentation of the feasihility of applyingany portion of the content; and 'above all,in research to staffing methodology. scholarly skill as measured byOrganization of content, quality of source materials, logical This review dealt primarily rigor, style with books, includ- of writing, and presentationof new knowledge. ing monographs, researchreports, and theses. Of There is no dearth of literature -literatureinthis on the general assification, 77pieces were subject of nurse staffing, but theunniher of tightly major researchreports. Certain other materials, conceived, well structuredexplorations of the which did tatcome under any of these headings. central issuesis sadly limited. Thereare only a few were also examined. Unfortunately,most of these comprehensive,welldesignedresearchstudies did not provide enoughdetail to make possible directed toward examining the application of the it specific staffing model criteria developed for the and its impact on patientcare and cost.

General Characteristics ofthe Literature

Most of the literature reviewedwas poor in made available to the nursing quality. Though unit, physical layout this judgmentisofnecessity of the unit, andcompetency of the nursing staff. porely subjective.itis at least based on an ap- The scheme of total nursingcare delivery also in- praisal of each study withrespectto the eight eludes forces affecting nursing unit points mentioned at the production from outset. Many authors fail the environments external andinternalto the to give a thoughtfui, carefully stateddescription hospital including the of a problem and its many systems other than analysis. For the mostpart nursing (see figure 1). they do not provide reviewsof the literatureor The authors of the various pieces references to related research of literature studies, even when vary in their style of writing and qualityof report- these are needed tosupport the presentation. Theo- ing. The sophistication of analysis reticalframework, or in different atleastsome embryonic studies appears to be relates.to the background structure for a rationale, is absent inmany studies of the author, thepurpose of the writing, and and reports. Tile lack ofacknowledgments, even the nature of the problem. In the when similar points research reports, are made, suggests that some only a few writersstate the assumptions underlying topics were explored concurrently by writers work- the selection of the statisticsused in the analysis ing Alependently.Itis difficultto determine and present evidence thatthese assumptions are the impact of specifit worksor the links between met, In a feu studies, however, that could individual efforts. serve as models, the writing is clear,w al edited. and logi- The number of variablestreatedindifferent cally organized. The studies ranges from use of language seems to re- a few to hundreds. The service flect the author's understandingof the audience offeredinanursing unit depends upon many to which he is,firecting his ideas and, insome factors other than the requirementsof the patients. cases,his understanding of the editorialpolicy It depends upon thepurpose of the unit, medical of the journal in which and nsing programs of the article appears. What care, supporting services is disappointing in much of theliterature is the

19 STATE OF KNOWLEDGE: NURSE S l'AFFING METHODOt.OGY 9

Figure 1. Scheme of nursing care delivery system.

Community Environment

Expectation of services Clientele:Knowledge of medical-hospital system Health knowledge and utilization Financial resources of community Physician supply and expectation Nursing manpower supply 1 Hospital Environment

organization Physician Eind4 uldual differences attitudes and expectations Patient utilization of servicesoccupancy rates andadmission, discharge patterns Specialization effort: complexity, intensity, and vuriety Floor layout: types and size of units Budget and budget control organization Hospital administration management style leadership Assignment of supporting services:accessibility, condition, adequacy, and individuality of items Hospital policies, cat* programs Organization tension and authority structure Intent and purpose, philosophy, objectives

Nursing Administration

Status of nursing and nurses in the total structure Leadership and management style Perception of purpose: philosophy, objectives Policies and programs: development and execution Authority structure--decentralization, decision making,delegation Communication Coordination Budget and budget control

Nursing Unit

Nursing supervision: clinical knowledge and skill management and training Amount, type, and kind of nursing staff and their combination in work Employment status of staff Absenteeism, turnover, propensity of staff to leave, illness Personnel policies: salary Inservice and staff development programs Allocation and assignment of staff Education and experience of staff Role clarity, role definition, and expectations of staff Knowledge and skill of staff Affective states of staff: satisfaction, morale, overload, tension, sense of achievement

20 10 RESEARCH ON NURSE STAFFING IN HOSPITALS lack of precise wording. the ambiguity of thestate- as related to staffing. It also indicates thatwe must ments, and the absence of a logical development urgently. proceed with studies of this important of ideas. This criticism, however, should besome- subject. It i., aim, highly desirable that theresearch what softened in view of the fact thatsome of results be cumulative. To hell) toward thisgoal, the literature was not written for publication.A individuals studying staffing should attacksimilar number of pieces were clearly designed merely problems,asksimilarquestions,and keepin as exhibits to be used in an oral presentation. touch with one another. It is imperative thatthe Study of the literature leaves the reader withthe investigations be bases, upon carefully_ designed impression that the amount writtenon the subject theoretical structures. isvast, highly uneven in quality, and available The generalizations I am goingto make are through many outlets. The quality of thereport- ing itself often gives grounds forconcern. The organized around the elements ofa staffing pro- value of whatiswrittenisoften questionable, gram: measurement of what nurses do and what since the writing lacks sufficient detailto provide patients require;measurement of quality of pro- effective support for the argument. duct produced; factors which influence estework This literature review makes possible several of nurses and, consequently, numbers andkind statements about what we do and do not know of staff that must be predicted; andmethodologies about nurse staffing. It pointsup the gaps in our currently used in hospitals to determine staffing knowledge. about many variables generally regarded needs.

Measurement of Nursing Activity

In much of theliterature reviewed, nursing First, the conceptualization of nursing practice work basic to staffing predictionswas measured derived from these workmeasurement techniques by the application of four standard techniquesof is limited both in scope and character. Theiruse work measurement developed by the engineering implies that nursing work is distinctlyprocedural profession: in character: task oriented with specific beginnings time study and task frequency count, and endings. Such a measurement doesnot reflect work sampling of nurse activity, the sweep of the effort. Nor does it properlyconvey continuous observation of nurse performing a sense of the great complexity of knowledge and activities, skill necessary in nursing practice. The application of these techniques self-report of nurse activity. assumes as an optimal goal the type of nursing that even through ordinary practices All four techniques involve the estimate of time may involve considerable interruption and dis- required for performance. Differences in techniques continuity. reside in the methods by which dataare collected, Alihough the nursing units selected bythe in- how the data are categorized, howamounts of time vestigator may be those whose staffsare regarded are estimated, and how minutely nursing work is as providing the best present nursing care in a described. hospital, one can raise questionsas to whether There is no question about the feasibility of the nursing practice observed is themost effective thesetechniques.Furthermore,theyare sound that can be provided. Innone of the literature and well tested and can produce reliable data reviewed were there models of the optimumnurs- if properly applied. Nevertheless, thereare two ing practice that could be achieved tinder thecon- major objections to theuse of these techniques straints of the situation. Descriptions of nursing in the measurement of nursing. The first relates activities obtained by these techniques tendto re- to the logic underlying their use in the study of flect mainly the statusquo of nursing practice, nursing practice. The second is the degreeof pre- including the ritualistic andunnecessary, ineffec- cision with which the needed datacan be or have tive acts. Omissions ofcare are not identified. been collected. To design a model of nursing meeting desired STATE OFKNOWLEDGE: NURSE STAFFING METHODOLOGY 11 yet feasible criteria of quality is the major problem seined are objective, reliable, and accurate. Only in nurse staffing research. Unfortunately, nursing a few of the studies describe the pretest of forms practitioners and nursing admintrators have not and data collection worksheets. Although several given attention to this basic problem and are often mention orientation and training of nurse stall unable to identifyrealistic and feasiblecriteria and observers for data collection, few have made of patient ewe. Furthermore, since nurses are un- tests of the agreement of observers or the reliability able to articulate their concerns about the lack of raters on specificfeatures. The problems of of models, they are unable to give guidance to in- error in recording or reporting data are seriously dividuals sincerely interested in helping with staff- ignored in most studies. Problems of the accuracy ing methodology. In none of the literaturere- of reporting are seldom discussed. Often, sample viewed was the seriousness of the lack of models size and sampling procedures are not given. Con- identified. sequently, the data presented in nursing activity Second, in many reports using these techniques, measurement are open to the suspicion of being evidence is markedly lacking that the data pre- biased and incomplete.

Patient Classification Schemes

Forty-one pieces of the literature referred to a (f)patient's geographic placement or status in scheme to be used inclassifying patients. Con the hospital system. ceptually speaking, nursing practice anditsde- The capabilities of the patient to meet his own livery evolve from the patient population that the needs are defined by statements that attest to his staffserves. The typeof nursingpractice,the ability to ambulate, to feed lihnself, to bathe him- amount, and the time and sequence of its delivery self, and to care for his own elimination. are derived from the requirements of patients. The Not all variables are used inallclassification introduction of the categorization of patientsis schemes. However, the schemes areusually dean attempt to quantify the nursing care workload veloped along a continuum of requirements for created by patient care demands. nursing assistance, moving from little to great need. The number of classes in the patient classifica- Items picturing the behavior of the patient and tion schemes described ranges from three, the most his special care requirements are ordered along usual, to nine. The most frequently used categories the continuum. At specific points, the patientis are selfcare, partial care, and complete care. Some judged to he in a different category. Most authors studies refer to these groups as categories I, II, and or investigators describe a scheme and provide 111, indicating that I I 1requires the most atten- a form and instructions for its use in theclassifica- tion. The names given by still other authors and tion of patients. investigators are related to acuteness of illness and The .analysis of the literature describing charact- care requirements, such as minimal, partial, moder- eristics of patient classification leaves four impres- ate, and intensive. The expression "variations in sions. First, the schemes are almost exclusively de- physical dependency" is applicable to terms used in veloped along the physiological dimensions of care. one study (1) as the title of a set of classes. Few items in the scalesrefertothepatient's In the literature, there seems to be general agree- sociopsychological behavior or requirements. Since ment about the variables entering into the patient many of the schemes grow out of acute care set- classification systems: tings, the emphasis on these characteristics is not (a) capabilities of the patient to care for himself, surprising. Second, littleis reported in the litera- (h) special characteristics of the patient, related ture about the precision with which patients can to sensory deprivation, be classified by the methods proposed. There is a (c) acuteness of illness, general failure of the authors to assess the degree (d) requirements for specific nursing activities, of reliability obtained in the classifications of pa- (e) skill level of personnel required in his care, tients. The problems of selecting observers and 12 REtEARCH ON NURSE STAFFING IN HOSPITALS training them to classify patients are almost totally classification schemes, despite their limitations and ignored. Third, only one study reported anat- imperfections, are widely used in staffing metho- tempt totestforvalidity(2).Fourth,patient dologies.

Quality Control

In the-182 pieces of literature reviewed there was Instead, it leads one to speculate that perhaps the noadequate,operationallydefined,modelof mode! ofqualityispoorly conceptualized,the nursing itself. Furthermore there were no descrip- variables selected are invalid, and the instrument tions- of the program of evaluating nursing care used is lacking in sensitivity, reliability, and ob- used by the departments of nursing service in the jectivity. hospitals where the studies were being conducted. These ideas are in part derived from the progress Yet the most critical staffing problem element is report of the research study by Wolfe and Breslin definition and identification of the quality of the (12) in their survey of eight nursing units. Four product (nursing care) and itseffect upon the ranked high and four low, based on the rankings patient. Often. staffing is viewed as an end in itin terms of the quality of nursing care made by self, rather than the means to an end. Measures 34 head nurses, supervisors, and instructors in a to determine the effect of changes in staffing upon medical center. The rankings of these nurses as to the quality of nursing production are essential in placement of the units on the high-low continuum staffing programs. of quality were highly consistent. These investiga- A most complicated and confounding are?.for tors developed a tool incorporating some of the study is assessment of quality. Though much at- items of CASH quality control with other items tention has been given to this question, never has and tested the nurses' rankings. Although the find- a completely acceptable model been devised. Re- ings of Wolfe md Breslin are inconclusive, the search directed toward tools for assessment of nurs- report does raise questions about the ability of the ing care is vital. Without valid and reliable tools, items drawn from the CASH material to discrimi- progress in measurement of care requirements, staff- late between nursing units providing low quality ing programs, and staff utilization, lags. of care and those providing high quality. The authors varied in the number of measures The materials describing quality measurement used to assess quality. The most comprehensive were foithe most part disappointing. Evidence works, involving a large number of variables relat- supporting validity and reliability of tools was ing to quality measurement or the criterion prob- missing in tnost studies. Rationales for the selec- lem, are those of Abdellah and Levine (3), Ayde- tion of specific items for inclusion in a measuring lotteand Tener(4),theIlliroisStudy(5), device were not presented. The question of the Ingmire and Taylor (6), Jaco (7), Lewis (8), Jelinek, significance of the content of an item, in terms et al (9), Alfano and Levine (10), and Wolfe (11). of changes in the state of a patient and his re- These studies, in particular Wolfe's, illustrate the covery, was enerally ignored. It seems clear that, difficulty in measurement and the complexity of the in spite of t problem of quality determination. few heroic attempts at research on Reports of hospitals using the CASH (Com- this question,relativelylittleis known about mission for Administrative Services in Hospitals) quality. Continued attention must be given to the program suggest that something is not right with matter of describing and resolving problems of present methods of evaluation of quality, but this quality measurement. Of prime importance is the does not necessarily entail the conclusion that the identification of what is to be delivered as the basic concept of continuing appraisalis wrong. nursing product in the first place.

23 STATE OF KNOWLEDGE: NURSE STAFFING METHODOLOGY 13

RelatedVariables

Some studies dealt with other variables. These administration less staff will be required than under did not deal with patient classification (nursing another. workload), measurement of nursingactivity,or quality measurement. Rather they dealt with variables related only indirectly to staffing and reflect- Architectural Variables ing, instead. probable use of time, effectiveness of The literature search for writings concerned with service, or organization of effect. architectual variables was not ehhaustive. Six major These related variables have been grouped into reports were found that examined the effect of five categories: design, travel time required of nursing staff, or (a). architectual variables, such as design of units, both. implying travel distance, room placement, The results of these si. studies, four of which visibility of patients, and communication sys- are major ones in the field (13,14,15,16), indicate tems: that nurse travel time does differ in types of nurs- (b) organizationalvariables,including charac- ing units. Three studies indicate that in the cir- teristics of the hospital, the introduction of cular or radial units nurses spent more time in new roles, redesign of supporting systems, and direct care. The general impression, although one realignment of authority and responsibility cannot generalize because of the different swings in nursing; and the limitations of the studies, is that circular (c) variables relating to nursing staff background, units are favored by staff over the other two types preparation in service programs, turnover, of units. and internal states such as satisfaction and No study addresses itself to the question of c'if- morale; ferences in staffing requirements. Yet a distinct (d) hours of nursing care; impression emergesthattypesofunitsdiffer (e) miscellaneous variables. greatly in travel distanCe and time. Whether or not Many writings discussed a combination of vari- the distance and time required are sufficient to ables. Each work was read to learn whether it require more staff in single corridor or rectangu- included specific findings regarding the relation- lar units is not known. ship between these variables and nurse staffing prediction. Nursing Staff Characteristics Organizational Changes A number of different variables relating to the nursing staff were found in the studies reviewed. Twelve pieces of literature were specifically con- One group of studies was concerned with nursing cerned with the introduction of the service unit staff satisfaction, turnover, and propensity of nurses manager system or other changes in supporting to leave their positions. Nursing staff perceptions systems. The review of these writings leaves the of adequacy of services were considered in a num- reader with the impression that there is limited ber of studies. These studies are few and, even knowledge about the impact of these highly im- though they include two fine research reports (3, portant variables. 17), are too limited to permit generalizations about One cannot, with confidence, say that an increase the relationships of nurse satisfaction to other vari- in supporting services will affect the amount of ables. They point to relevant questions forre- staff required; one can probably state that the search. amount of direct care to patients and standby time of nursing personnel will be increased.It can also be said that administrative variables pro- lnservice Educational Programs bably have an effect on utilization and amount of staff required. Unfortunately, there is still no ade- The impact of inservice nursing education upon quate basis for stating that under one style of staffing requirements was not examined in most of

24 14 RESEARCH ON NURSE. STAFFING IN HOSPITALS tits studies. The rile( Iof inservice educational pro- Age and Care Requirements grams cartstall performance, either on the ability to achieve gveater Workload or on theproduction Three studies examined the effect of age upon of higher quality G.re to patients, is still obscure. nursing care hour requirements. These three sup- There is some evidence the programs eflectiely port each other's findings that the over 65 years increased patient contacts. However, the question, of age patient group exceeds the under 65 years Does ail inserice educational program result in of age group in hour requirements. Jacobs' study, greate:liciency and economy of staff?istitian- (18) the most comiirehensive of the three, reported met ed. thatthe 65-74 year old group requires 21-31 minutes per day more and the 75 year and older Nursing Hours per Patient Day group needs 55-79 minutes per day more. He found there was little difference between the time Only a limited number of research vudics pro- requirements of adult medical and adult surgical vide information on actual nursing hours per pa- patients, andthatspecializationeffortsof the tient day (N FWD) and the validity and reliability hospitals were not factors in the amount of care of the data used in those studies are in most in- required. stances questionable. Techniques are loosely de- scibed, samples arc small or ill der.ned, instrument toting is deficient, and rationales are weak. In most Control of Costs instances, information about the quality of care is derived solely from the judgments of the nurse Most of the methodological studies of staffing involved. The studies vary greatly in scope and were highly concerned about the control of costs. quality. Most apply only to specific hospitals. Ade- However, none of the studies reviewed was suc- quacy of staff utilization was found to varygreatly cessful in showing the relationship between the among hospitals, and the hospitals variedgreatly variables of cost effectiveness and nursing staffing. in the average number of daily hours nurses were The problems of defining cost, obtaining reliable available. data, and isolating the dependent variable, quality, The confounding question of patient care qual- were recognized by .nany otherauthors. Two ity arises in the examination of these regtarch studies presented highly interesting and relevant studies. This kind of assessment was ignored, taken material about the shortage of nurses, using dif- for granted, or poorly handled. It is obvious that ferent economic models for analysis (19, 20). Both NHPD is less titan satisfactory in staffing determine tion. The literature suggests that the use of guides authors provided provocative pertinent data about and procedures applied to a specific hospital situa- the dilemma confronting employers and especially tion is a more logical approach. the public.

Staffing Methodologies

Staffing methodologies in the literature fell into and the use of abstraction its resolution (see four major classes: figure 2). (a) descriptive, using a numoer of variables, sur- To clarify the following disculsion, the terms vey methods, and the subjective judgments 'of "methodology" and "approach" and these four individuals; classes of methodologies will be defined. The term (b) industrial engineering; "methodology" is used to describe. a "logically and (c) management engineering (2'1); procedurally organ:zed arrpagement of steps. When (d) operations reseat ch. formally documented utilized, these steps may These methodologies also represent a gradual in- constitute a science of method for whatever dis- crease in the application of logic to the problem cipline uses them (22).

25 STATE OF KNOWLEDGE: NURSE STAFFING METHODOLOGY 15

Figure 2.Diagram of evolvement of methodologies for the study of nurse staffing; application of logic and increases in abstractions.

Descriptive Industrial Management Operations Engineering Engineering Research

The first term, "methodology," is regarded as a The operations research methodology is directed much larger concept than "approach." The lattlz tnward enabling one to make decisions for real termis more limitedinthatitrepresents an life situations. Mathematical models are built to orientation or a preliminary step and does not en- represent reallifeproblems. The essentialele- compass a full methodology.. ments are abstracted from the situation so that a The descriptive methodology makes use of a search can be made for a relevant solution to the number of data gathering devices about a large problem. The structures of solutions are explored number of variables,she relationships of which and procedures are established for obtaining them. are uot clear. The final decision abo _4t amount of The optimal solution and set of procedures are staff resides in the judgments of Individuals who then made available. have a background of experience. The approach Since authors of the pieces reviewed often uti- may use simple ratios, formulas, and suggested pro- lized the same techniques, such as counting, patient portions between types of personnel.However, classification, work measurement, and included there are no consistent strategies in the methodo- many of the same variables, a set of arbitrary rules logy. was drawn up for classifying methodologies. The The industrial engineering methodology is di- rules established exclusive categories. The descrip- rected at the study of nursing work on a nursing tive methodology is one in which thereis no unit. It uses the techniques of work measurement, quantification of nursing work and no use of patient work distribution, task or function analysis, pro- classification system. The methodology drawn from cedure analysis, and the like. It is primarily directed operations research requires the construction of atreorganization,reassignment, and redistribu- a mathematical model. tion of work on specific nursing units. The industrial management methodology em- The managementengineeringmethodology, ploys methods from industrial engineering and which claims to be based upon common sense, restricts it to a study of the nursing unit. It does uses tools and techniques from industrial engineer- not include the full range of strategies seen in ing and from systems analysis, and builds upon the management engineering methodology and does findings from operationsresearchstudies. This not interface the work on the unit with systems of methodology must give evidence that the majority the hospital. It does not include p? ',nit classifica- of the component strategies are present, although tion. The management engineering methodology completeness of each component may vary. The must give evidence that the majority of component component strategies include: strategies are present, although completeness of each a statement of performance objectives, component may vary. It may use r thematical tools an analysis of components and functions, and may draw upon the research conducted by per- distribution of functions, sons using operations research methodology. These four types of methodology for the deter- scheduling, mination of nursing staff required in a hospital training individuals for the use and testing of are described in many of the research reports. the system, manuals, pamphlet, and guides reviewed in the installation of system, literature. Linkarbetween the work of individ- quality control (23) ual arc discernis le, although the times at which

26 16 RESEARCH ON NURSE STAFFING IN HOSPITALS

Table 1.Comparison of staffing methodologies on selected features

Feasibility of approach Title of methodology Reproducibility Reliability of data Validity of data of forms Setting requirements

Descriptive No evidence Face validity Some excellent No specific require. examples in literature ments (C7, Gunn) (C14, Paetznick) (D29, Lud wig and Humphrey)

Industrial engineering Depends on accuracy Face validity Excellent examples in Industrial engineering of reporting, training literature (A47, Mer consultants or nursing programs, and clarity gen)(D4, Bartscht) staff knowledgeable of forms. Checks pro. (D20, Hansen) about concepts and vided in some Leptis tools

Management Depends on accuracy Face validity; relies on Examples in literature Consultant help; no engineering of reportingroutine Connor's scheme and (CI9, Scottish Health specific hospital setting checks are not built in, its modification. Service, No. 9)(C13, training of personnel Operations Research for data collection, and Unit)(C4, CASH, necessary clarity of Staff Utilization) forms. (Cl2, Mass. Hospital Association)

Operations research Evidence presented. Face validity; relies on Examples in literature Computer; consultant Depends on accuracy Connor's classification. (A72, Wolfe) (D46, help of reporting. Wolfe, and Young, Part II)

Descriptive Subjective judgment Survey methods: Many. Examples: Broadly applicable, about value of counting, question. census data, hospital but relies upon ex variables naires statistics, financial re periences and knowl sources, training, edge of the director, nursing objectives, the administrator, and policies, clinical serv the staff ices, etc.

Industrial engineering Studies and reports Industrial engineering Personnel statistics, Various types of work from industry. As. nursing care activity, measu:ement used. samption made that and time require. Nursing staff can make nursing is a composite ments, work layout, application. of tasks or activities census statistics, work distribution, and work. load factors.

27 STATE OF KNOWLEDGE: NURSE STAFFING METHODOLOGY 17

Table 1Continued

Title of methodology Basis of rationale General approach used Variables considered Comments

Management Research studies of the Industrial engineering (a) nursing workload Programs are pur- engineering Johns Hopkins Opera- methods and systems by task frequency portedly designed with tions Group. Studies analysis and level these components: and reports from busi- (h)standard time for (a) identification of ness and industry tasks or categories performance ob. using industrial en. of patients jectives gineering and manage. number of patients (b) function analysis ment methodology. in classes (c) scheduling of per. Assumptions made: (d) supporting services sound that nursing provided (e)I. innel levels (d) prediction of staff represents quality de. skill allocation tables sired, that patient (f) administration (e) training of per. classification is valid. ,licy and proce sonnet that nursing sell ice dures (f) quality control demands are stochas- (g) attributes of program tically distributed, and quality (g) management con. that standard times of (h) personnel data trol information procedures are valid and statistics reports and reliable. (i) age of patient (j)cost figures

Operations research Studies and reports Operations research: (I) classes of patients Application not widely from industry. Opera. mathematical alloca (2) 16 task complexes observed in the litera dons research by Con. tion and situation (3) cost values tare. Quality measure. nor, Wolfe. and Flag le. !maid, Queueing ment recognized as not Same assumption as Theory. Industrial present. above. engineering and sys terns analysis they occurred are not known. Many of the writings all four methodologies. Relatively few variables are are undated and the investigator did not attempt, considered. yielding narrow basic conceptual mod- by personalcontact withindividuals,to learn els, and the impact of architectural variables, or- precisely when programs were initiated. ganizational variables, and specialization efforts is The majority of the manuals, pamphlets, and almost totally ignored. No attempt was made to guides are poorly written. Little is included in them predict the number of persons required for major about rationaleor assumptions underlyingthe position categories inthe nursing care delivery methodology being described. Protocol and pro- system. Validity and reliability of the data are cedures are likewise sketchily outlined. The use of questionable. the manuals would not be possible without con- Except for a few reports applying the metho- sultant help. Some reports of application of the dologies, tests for accuracy in reporting the data methods were obtained that clarified the metho- and the training of observers are not proposed. dology consilerably. Validity of the data resides in the judgment of A comp anon of thefour. methodologies was nurses selected to work with the methodologists. made with respect to a Tecific set of character- All four methodologies are applicable to a variety iitics that included rationale, approach, variables, of healthsettings. Two types of methodologies and feasibility (refer to table1). In all four me- require consultants in their use. Forms employed thodologies,subjectivejudgments arepresent, in recording tasks are incomplete, and standard since nursing units selected for workload measure- times reported for task performance are not documents are made either by individuals or by group mented. The classification schemes used in two discussions. Measurements of care quality and per- of the methodologies are drawn from Connor's formance standards are weak or totally lacking in (24) basic research. 18 RESEARCH ON NURSESTAFFING IN HOSPITALS Summary of Literature Search

The results of this literature search have been nursing personnel and their ability to mount a sobering. We can say that we know little. However, workload arising from a particular mix of cate- there are a few generalizations we can make in gories of patients is little. a summary. First, nursing activity can be measured Third, a large number of variables appears to by standard methods drawn from the discipline of influence the work of a nursing staff. Again, we engineering. The techniques are feasible and, if know little about the effect of the interaction of appropriately applied, provide reliable data about these. The impression one derives from the litera- work that reflects nursing as it occurs in current ture is that the admission and discharge process of practice. However, the tasks, activities, and cate- patients, the environmental structure of the unit, gories appearing in the literature do not, in my the supporting systems, the capabilities of the nurs- opinion, reflect the nature and full character of ing staff, and the clinical nursing leadership are nursing practice, but only a portion. major factors to be considered. Few pieces of the Second, most patient classification schemes are literature reviewed referred to the impact of the modifications of that proposed by Connor who in philosophy and practice of medical staff, their num- his dissertation provided an admirable rationale of ber, their presence, or absence. Yet I predict that classification. However, most of these classification the physician and his behavior will prove to be a schemes have not been tested for reliability of the prime variable in the creation of nursing work. data they produce. The evidence of validity pre- Fourth, four types of staffing methodologies were sented is weak. Although the variables used for identified. These all are feasible and are widely classifying patients are fairly universal, the number used. The concept of variable staffing is adopted and kind are limited and some requirements of in many institutions. care are either overlooked or ignored. Further, Fifth, quality measurement is poorly handled in questions can be raised about the significance of nearly all the studies. In large part this is due to some of the items included in the scales. Do they the failure to describe in operational terms the reflect states of recovery and the objectives for which the patient is in the health care delivery product to be produced. system? The relationship between patient classi- Sixth, there exist in the literature no adequate fication schemes and nursing workload has been descriptions of models of the nursing care delivery best handled in the research by Connor (24) and system and nursing practice. The majority of stud- Wolfe (//),especiallythelatter.however, the ies give little theoretical structure as a basis for knowledge we have about the combinations of investigation.

Problems and Issues

This review of the literature leads to the con- but also force our attention on the central prob- clusion that our knowledge about nurse staffing lems and issues relating the nurse staffing metho- isvery limited. The number of variables with dology. Nowhere in the literature was found a which one mull deal is large. Furthermore, many well developed model of the nursing care delivery of them are almost incomprehensible and, for the system for use in the development and examina- most part, appear' to defy description and quanti- tion of staffing. The term "model" is used, in this fication. There is little agreement in the literature sense, to refer to an analogy of the pattern of on the nature and characteristics of some obviously relationships observed in the nursing care delivery importantvariables:nursingpractice,patient system. needs, quality of care .control, complexity of care, The problems of model construction are exceed- intensity of care, and levels of practice. ingly troublesome. First, there is the question of Review and assessment of theliterature not setting the boundaries of the model. Will the bound- only point up how little we know of the subject aries be those of the nursing unit, as they are in

49 STATE OF KNOWLEDGE: NURSE STAFFING METHODOLOGY 19

Jelinek's (25) and Wolfe's (II) writings,orwill provided for the patient is sufficient. Therefore, they include parameters beyond the hospital, as the present classification systems have examined implied in the Illinois Commission Study (26,27) physical elements of care or activities performed and Levine's (28) study of Federal and non-Fed- and have accepted them as requirements for care, eral hospitals? Second, how many variables can be without giving adequate thought to the question, adequately explained and managed in the model? What does the patient truly require? There is no The problem of identifying relevant variables and empirical evidence to support the assumption that eliminating irrelevant ones is extremely complex. what the patient is getting is all that he requires Knowledge about the variables and their relation- or is necessary for his recovery. In fact, there is ships is vague and measurement of some of the some evidence that what he is getting is given as variables is almost impossible. a matter of ritual. Nurses who have been within the real work The majority of patient classification schemes world would undoubtedly identify three groupings reflect specific nursing tasks, a number of which of variables: arise because of medical order and acuteness of (a) selected factors external to the hospital in- illness. The schemes do not reflect emotional needs, fluencing nursing delivery, such as nurse- orientation of the patient, instruction needs, and power pool, financial resources of the clien- comfort, other than through the process of provid- tele, and the health status of the personnel; ing physical care. Few have even attempted to in- (b) components of the hospital system, such as clude these items. No scheme, for example, has patients, medical staff, management, financial been built upon the nursing problems of the resources, supporting services, traffic patterns, patient, although several investigators began with nursing leadership, qualification of staff, staff this intent. Yet, experienced practitioners in nurs productivity, and staff stability; ing can set a priority ranking of nursing problems and combinations of problems that would (c) independent factors that can be used to de- reflect both amount of time and professional knowl- termine the effectiveness of the system, such edge base and skill required for performing the aspatientsatisfaction,patient behavioral care to solve the problem. Only one study reported changes, and efficiency of services. the reordering of priorities of nursing care as a Some good ideas are found in some writings. How- result of the admission of a critical patient. ever, thoughtfully conceived and carefully written If one assumed presentpatientclassification descriptive models are greatly needed. schemes valid, another problem exists. Except for The second major problem confronting staffing one study, patient classification schemes have not methodology development is the difficulty of iden- been rigidly tested and have not been used in a tifying in advance what the patient or group of variety of different settings in a way that would patientsisentitledto receiveastheproduct, indicate how reliable they are. Connor's scheme nursing rare,for which they are paying. The fun- was designed for a large medical center and only damental reason for assembling a staff is to take one formal test of it was found in the literature. care of the patient. Is it equally applicable to medical-surgical patients Identifying the components of care a patient in a small community hospital? What is the re- must receive is essential to any staffing methodology. liability between raters? The arrival and discharge times, the hourly and ?roperly used, the various methods for study- daily variation in the patient's requirements, the ing nursing work are feasible, reliable, and simple. medical programs and scope of medical practice, The problems are those of sampling, of selecting the training programs, the problems of measure- model units for survey, of establishing categories ment of quality, and the lack of an operational that represent major nursing functions, and of the definition of care itself make the problem of measurement formidable. None of these items were ade- limitations inherent in the use of the method. quately treated in the literature, and some were What is most questionable is the assumption that totally ignored. what is being observed is desirable. Further, the The assumption made by the present methodol- categories used often do not reflect the planning ogists is that, at the time of the survey, what is and thinking time required if nursing is perceived

30 20 RESEARCH ON NURSE STAFFING IN HOSPITALS as having an intellectual base. The selection of a a patient, the skill with which a particular pro- model unit for study to make staffing predictions cedure is performed, the number of times a patient iscritical. The management engineering reports receivesaprocedure,the number of times the indicate that the selection of the unit rests upon patient has contact withthe nursing stall,the the judgment made by an advisory group composed flow of communications, the results of a nursing of nurses and hospital administrators. It appears, audit, or even patient satisfaction are inadequate however, that this judgment is usually not based criteria for the evaluation of the quality of nursing on hard dam, reflecting the achievement of a spe- care the patient has received until therelationship cific quality of care, evaluated by well defined cri- between these factors and the progress of condi-. teria. tion of the patient has been established.It has Another probleminnursing measurementis generally been assumed that a positive relation- grouping tasks or categories into complexes for ship exists between certain accepted nursing prac- further research on staffing methodology. It is ex- tices and the patient's condition. The evidence to tremely difficult for sonic nurses to perceive that support most oftheseassumptionsismeager, personal services to patients may not be necessary. particularly applied to a total nursing unit feasible, or economical. The use of expensive pro- rather than a select group of patients. fessionaltimeforsimpletasks,repetitivebut The major reasontherelationship between personal for the patient,is overuse of skill and nursing activities and patient condition has not talent. Ilia this statement is unacceptable to many teen generally determined may be the problems nurses. involved in the development of patient condition These nurses state that use of auxiliaries de- criteria;i.e., identification of what one must ob- personaliies nursing care, that these personal tasks serve in, on, or around the patient to determine may serve as vehicles for nursing action, and that whether the nursing care hr receiving is bene- the elimination of personal care from registered nurses' repertoire of duties is unwise. The crux ficialto him for his present and future status. of the argument is not whether itis personal or Itistime the questionof the construction of not, but what is the knowledge and skill requite- criteria be squarely faced. ment for its performance and how does this affect The measurement of nursing care quality and patient recovery? It is uneconomical for registered the establishment of performance standards can- nurses to spend time on giving nontherapeutic not be accomplished until nursing practitioners care to patients. We should ask what registered and leaders in nursing practice give attention to nurses are doing for patients that no one else can the development of such criteria. The primacy of provide and what they are not providing. the problem is apparent. The questions to be asked The ultimate test of a staffing program, or any are these: Where are the practitioners who can nursing activity, is the progress of the patient, not develop stated criteria? How can the criteria be the quality of performance of the nursing staff, the appearance of nursing units, or the complete- built into an ongoing program of quality measure- ness of records.'- The amount of time spent with ment? What modifications are required of the ongoing program to build scales or indices which ' Acknowledgment is given Dr. Marjorie Moore, Associate Director. the Universit of Iowa Hospitals and Clinics. for can be used at specific points in time for research the following paragraphs. purposes?

Future Research Direction

Although. the problems and issues reviewed are acquainted with the real work world of nursing almost overwhelming, I inn ccmfident progress can care delivery systems should be sought, encouraged, be made on their tesolution. First of all, nurses and supported in the development of conceptual knowledgeable of nursing practice and intimately models of the nursing care delivery system. This is STATE OF KNOWLEDGE: NURSE STAFFING METHODOLOGY 21 noeasytask.Furtliermore,theseindividuals Therefore, different sets of criteria may be needed should be encouraged to seek out and collaborate since the patients in the hospital vary. The patient with individuals from other disciplines. variation is dependent upon t"purpose for his Second, carefully designed research studies test- being in the hospital, the pigress he is making, ing present classification schemes should be con- and the course of his recovery. Development of ducted. Since limited evidence appears in the lit- these criteria will require the involvement of ex- erature that these schemes are providing reliable pert nurse clinicians. datacarefully controlledtests should be made Sixth, guidelines for use of the four specific of the schemes in a variety of hospitals. The ex- methodologies should be developed. These guide- ternal factors that influence the character of the lines should state the rationale upon which each patient population should be identified as fully as is based, the variables studied, the limitations of possible. each the appropriate use of the methodology, the Third, patient classification schemes built around component strategies, and the procedures. It should nurse perceptions of care should be built and tested. include examples of well developed forms, reports, These schemes should be based on reasonable care the guides for use of allocation tables, if included, requirementsasnurses seethem. The present and the like. The value of the methodology as schemes do not reflect omissions of care that may well as its limitations should be stated. The im- be serious; on the other hand, omissions may not portance of concurrently studying supporting sys- be important. However, the recommendationis tems, or even preceding the analysis of nursing that proposed schemes for the classification of pa- with a study of these, should be emphasized. Proper tients indicating nursing loads be developed by utilizationof supportiveservices and expecta- knowledgeable practitioner nurses. These schemes tions of adequate service from them should be should be established around nursing problems made clear. and should reflect nursing priorities, knowledge, Seventh,guidelinesfornewmethodologies and skill level. The scheme should lend itself to should be explored, greatly enlarging the variables computerization. It should be based on data used considered. The methodology should be based up- by the nurses in assessing patient tare requirements. on a carefully described conceptual model of nurs- Foul th, since the effect of improved supportive ing care delivery. It should provide guidelines that systems upon nursing staff requirements has nt.t will take into account providing for omissions of been well documented, controlled studies of the care as well as eliminating repetitions of care, the effect of improved delivery and retrieval systems organization of the work of nursir personnel, and placement of patient care items, equipment, and the sequence of its delivery. andsuppliesshould beconducted.Holding These seven recommendations are specific pro- qual:ty constant, what is the effect of improved posals for research. My final comment is a general availability of work materials and tools for pa- one but one which I am convinced is of major tient care upon the size and kind of nursing staff importance. Research on these seven must be mul- required? What is the effect upon cost? tidisciplinarythe discipline of nursing must be Fifth, a concentrated attack on measurement of involved. Mechanisms for sharing researchfind- quality must be made. From the review, the im- ings must be set up and knowledge systematically pressionisthat atiempts are too global. Except for a few instruments, hot much is available to built. To achieve this end, we must address atten- assistthe director of nursing and the hospital tion to the structuring of theoretical frameworks administrator in making estimates of quality. The for the research on staffing of nurse delivery sys- evaluation must relate to the effect on the patient. tems.

, 22 RFSEARCH ON NURSE STAFFING IN HOSPITALS

References

(1) Scottish Health Service Studies, No. 9.Nurs- logg Foundation, Battle Creek, Michigan, ing Workload per Patient as a Basis for 1971. Staffing.Scottish Home and Health Depart- (10) Alfano, GenroseJ.,and Levine,Herbert ment, 1969. S. Progress Report.Longterm Effects on an (2) Mainguy,J.,etal.The Reliability and Experimental Nursing Process.(NU 00308- Validity Testing of a Subjective Clussifica- 40).MontefioreHospitalandMedical tionSystem.Report of Canadian Public Center, New York, New York, 1971. Research Grant No. 609-7-197, Vancouver (11) Wolfe, H. "A ;Multiple Assignment Model General Hospital, Vancouver, British Co- for Staffing Nursing Units." (GN-5537, HM- lumbia, June 1970. 00279, and FR-0004), Unpublished Doctoral (3) Abdellah, Faye G., and Levine,Eugene. Dissertation,JohnsHopkinsUniversity, Effectof Nurse Staffiing on Satisfactions Baltimore, Maryland, 1964. with Nursing Care.Hospital Monograph (12) Wolfe,H..and Breslin,Patricia.Factors Series No. 4. American Hospital Associa- Affecting Quality of Nursing Care.Progress tion, Chicago, Illinois, 1958. Report, (NU 00192-0101), School of En- (4) Aydelotte,Myrtle, and Tener, Marie.An gineering, School of Nursing, University of InvestigationoftheRelationBetween Pittsburgh, Pittsburgh, Pennsylvania, 1968. Nursing Activity and Patient Welfare.Final (13) Sturdavant, Madelyne.Comparisons of In- Research Report (GN 4786 andN 8570), tensive Nursing Service and a Rectangular University of Iowa, Iowa City, Iowa, 1960. Unit.Hospital Monograph Series No. 8, (5) IllinoisStudyCommissiononNursing. AmericanHospitalAssociation,Chicago, "Nurse Utilizationin Thirty-oneIllinois Illinois, 1960. HospitalsMethodologicalReport."Labo- (14) Trites, David K.,etal.Final Report on ratory data presented to staff members of Research Project EntitledInfluence of Nurs- the Division of Nursing, National Institutes ing Unit Design on the Activity and Sub- ofHealth, Washington,D.C.,December jective Feeling of Nursing Personnel.Re- 1968. search Office, Rochester Methodist lIospital, (6) Ingmire. AliceE., and Taylor, Carl.The Rochester, Minnesota, (no date). Effectiveness ofaLeadership Program in (15) Huseby, Robert. "Impact of Nursing Design Nursing.WesternInterstateCommission on Patients' Opinion." Unpublished Master's for Higher Education, Boulder, Colorado, Thesis,University of Minnesota,Minne- (NU 0078), 1967. apolis, Minnesota, 1969. (7) Jaco, E. G.Evaluation of Nursing and Pa- (16) Lyons, Thomas F.Nursing Attitudes and tient Care in a Circular and Rectangular Turnover: The Relation of Social-Psycholog- Hospital Unit.Final Report, prepared for icalVariables to Turnover, Propensity to theHillFamily Foundation,St.Luke's Leave, and Absenteeism Among Hospital Hospital, St.Paul, Minnesota, June 1967. Staff Nurses.Industrial Relations Center, (8) Lewis,Charles.A Privileged Communica- Iowa State University, Ames, Iowa,1965. tionThe Dynamics of Nursing in Ambula- (17) Jacobs,Stanley.etal. American Hospital twv Patient Care.Final Research Report, Association Nursing Activity Study Project U.S.PublicHealthServiceGrant (NU Report. Public Health Service Contract No. 00145), (no date). PH 110-235, U.S. Department of Health, (9) Je linek. R. C.: Munson, Fred; and Smith, Education, and Welfare, Washington, D.C. Robert. SUM(Semler Unit Management): (no date). An Organizational Approach to Improved (18) Yett, Donald E. "The Chronic 'Shortage' Patient Care.A Study Report for the Kel- ofNurses: A PublicPolicyPile,ama."

33 STATE OF KNOWLEDGE: NURSE STAFFING METHODOLOGY 23

EmpiricalStudiesinHealthEconomics, (23) Connor, R.J. "A Hospital Inpatient Classi- Johns Hopkins University Pre.,s, Baltimore, ficationSystem."UnpublishedI) Ictoral Maryland, 1970,357-97. Dissertation, Industrial Engineering Depart- (19) Bognanno,MarioFrank."AnEconomic ment, Johns Hopkins University, Baltimore, Study of the Hours of Labor Offered by the Maryland. 1960. Registered Nurse." Unpublished Doctoral (24) felinek, R. C. "Nursing: The Development Thesis, University of Iowa, Iowa City, Iowa, of An Activity Model." Unpublished Doc- June 1969. 'oral Dissertation, University of Michigan, (20) This categoryis used sinceitis a general Ann Arbor, Michigan, 1964. term applied to the methodology described (25) Illinois Study Commission on Nursing. Nurse in Dollar and Sense, a study report prepared Utilization: A Study in Thirty-one Hospi- tals.Vol. The for forthe W.K. KelloggFoundationby III. IllinoisLeague Patric I.Ludwig, October 1971. Individuals Nursing, Illinois Ntrses' As.ociation, Chi- arc referred to this monograph for informa- cago, Illinois, 1966-70. tion about the extent to which this metho- (26) Levine, Eugene. "A Comparative Analysis dology has been used. of the Administration of Nursing Services in a Federaland Non-Federal General (21) Optner, StanfordL.Systems Analysisfor Hospital." Unpublished Doctoral Disserta- Business and Industrial Problem Solving. tion, NmericanUniversity,Washington, Englewood Cliffs, New Jersey: Prentice-Hall D.C., 1960. Inc.. 1965. (27) Haussmann, R. K. "Waiting Time as an (22) Hi Frederick S., and Lieberman, Jeral 1. Index of Quality of Nursing Care." Health troductionto Operations Research. S n Services Research, vol.5,no. 2 (Summer Francisco: Holden-Day, Inc., 1967. 1970).

34 Factors Related to Nurse Staffing and Problems of Their Measurement

Charles D. Flagle, Dr. Eng. Professor of Operations Research Department of Public Health Administration School of Hygiene and Public Health The Johns Hopkins University

Introduction

Looked at in any way, the problem of staffing trol in the health care delivery system than the hospital nursing units is profoundly important and resources used in a day of care. profoundly difficult. In economic terms alone, we In social terms, we have lived through a fore- are dealing with several billion dollars in wages shortened transition from a vocation expecting each year in the United States, something on the much personal self-sacrifice to one bearing con- order of AO percent of all medical costs, and a siderable resemblance to the way of life and af- fourth of hospital costs.l Little wonder then that fluence of other skilled purveyors of human serv- nursing care is a focus of attention, not only in ices.Changesinattitudesandimageshave its formulation and distribution of tasks, but in accomim lied this transitionthe secularization, if the criteria for admission and discharge to inpa- you will, of patient care. The tangible correlates tient care. Throughout our discussion we should of these changes have been such things as shortened keep in mind that the total cost of impatient and more formalized working hours for nurses, nursing care to a community is the product of higher wages, unionization, occasional strikes, and the cost of nurse hours per patient day times in the auxiliary positions an undeniable demand the total number of patient days. The total number for ethnic equality and upward mobility in jobs. of patient days of care may be more subject tr, con- Interwoven with allthe economic and social M. Y. Pennell and David B. Hoover, in Health Manpower changes affecting the organization for patient care, Source Book Section 21, "Allied Health Manpower 1950-80, there has been an unending series of technological Public Health Service," (Publication No. 263, 1970) project developmentsphysical technology and medical. present total nursing and related sources from two million Intensive care units and life islands, for example, employeeiat.presentto three million by1980, roughly have imposed demands on nurses for increasing onefourth being registered. The Commission of Health Manpower,1987. estimated 60 percent of the totalare specialization. employed in shortterm general hospitals. In the face of all this change, heaped upon the

25 35 26 RESEARCH ON NURSE STAFFING IN HOSPITALS inherent difficulty of caring continuously and si- nurses, physical setting, and supporting services. multaneously for 30 or so sick people in a nursing The goals of the nursing unit reflect the goals unit,itislittlewonder thata greatdealof of the larger institution of whichitisapart. administrative and research attention should be Care ofitspatients,training ofitsphysicians, centered on the staffing problem. Determining who nurses, and otherpersonnel, perhaps some re- should be where, when, and doing what, isa prob- searchall these are the stated objectives of the lem deserving all the attention it has had. Judging unit. To understand the behavior of the actors from our accomplishments todate, one might. in the hospital inpatient unit, one must also be suggest that even more or broader effort should aware that individuals may be guided by profes- have been expended through the years. sional values quite apart from the immediate de- But I am concerned that, for all its importance, mands of the work setting or the formal goals of afocus on the mechanisms and decision rules the institution. - for assigning personnel of various skills togroups To gain insight into the complex processes of of nursing tasks will obscure larger issues. Such patient care, various investigators have constructed a focus will certainly obscure the real emphasis models. Daniel Howland,for example, has proof much of the research carriedout during the posed the concept of the nurse-patient-physician 1960's. IfI may speak reminiscently of our own triad as a subsystem whose objective is the mainte- experience at Johns Hopkins, I can only capture nance of homeostasis in the patient, and whose the whole story by saying that from the outsetwe illnessis seen as the inabilityof his own sub- were hopeful of developing a rational system of systems to maintain homeostasis. How well the inpatient care, considering allresources, human physician and nurse performthe quality of their and physical, as potential elements of thatsys- careis evidenced by the promptness of their de- tem. If nurse staffing received special emphasis it cisions and actions in restoring physiological norms. was because of the dominant aspect of nursing Given the diverse responsibilities of the physi- shortages and personnel turnover. cian and nurse in the simultaneous care of many In those days there was an abundance of pa- patients, traditional roles have evolved for each. tients, doctors, and beds, a growing army of ad- Though time has changed their roles somewhat, ministrators and ancillary specialists, a stream of one can stillidentify formal functional relation- new equipment and labor saving devices, and a ships. The nurse has continuous custody of the progressive shortage of nurses.It was to correct inpatient and, through the physical design of the or compensate for that shortage that hospital ad- nursing unit, has the capacity for continuous ob- ministrations sought innovations. However, in the servation or awareness of the state of the patients. process, new strains were imposed on nursing per- The physician sees patients intermittently but is sonnel,foralthoughsubstantialassistancedid responsible for signaling actions to be taken on the come from the centralization of such support serv- patient's behalf. Thus the activities of the nurse ices as sterile supply away from the nursing units, are responsive to two sets of signals, one from new problems of communication and coordination her own communication and observation of the were created as a consequence. patient's conditions, the other from theset of Seeing the nurse burdened with a residue of physicians' orders, which she may have had some nonnursing tasks and simultaneously obliged to influence in shaping. play a large role in ensuring the logistical support An examination of the whole nursing unit as a of her unit, research took a new direction. Seeking set of triads of patients-doctors-nurses devoted to on one hand to introduce more new technology, maintaining homeostasis in individual patients does more external support, researchers attempted to not reveal anything corresponding to homeostasis restore and improve upon the cybern'etic integrity in the nursing unititself. There isinfact no of the nursing unit, which had existed in the clays constancy, no convenient norms to be maintained. of internal self-sufficiency by creating new mechan- Instead there is a constantly and widely changing isms of communication and coordination. Think of a nursing unit in those terms. 'See the series of papers by D. Howland and W. McDowell Its in Nursing Research, 1963 and 1964, and American Journal elements are its patients, set of doctors, teams of of Nursing, 1966. FACTORS RELATED TO NURSE STAFFING AND PROBLEMS OF THEIR MEASUREMENT 27 pattern of aggregate patient needs.l'he cybernetic expected shortage or overage of personnel on the system problem is to formalize the process of ob- units.4 It is usual for the workload index to vary servation and respond in such a way that resources from 15 to 45 nurse hours on 30-bed units, while and skills can he called forth over the course of the total nurse hours available range from 35 to time in amounts appropriate to the level of need. 37. In quantitative terms, this means matching two What causes these variations? To what extent erratic and not very controllable variables:the are they controllable and whose decisions and. ac- aggregate hoursofdirectpatient careelicited tions are required to exert such control? How are (the direct rare workload), and the total number the vagaries of patient needs and available per- of nurse hours available to render care. The nature sonnel interrelated to arrive at nurse staffing prac- of the problom can be seen in figure1, which is tices? Consider first the demand factors related to a routine daily forecast of nurse hour requirements patients. for a set of nursing units, a statement of scheduled hours for the day, and a gross estimate of the 'Connor, R. "A Work Sampling Study of Variations in Nursing Workload," Hospitals, May I,1961. It would be 3 See Fla*. C. D. "A Decade of Operations Research in difficult to defend the accuracy of the fixed or variable nurse Health," in New Methods of Thought and Procedure, Zwicky. hours (figure I) transferred to another service, 10 years later. F.. and Wilson, A.C...editors. SpringerVerlag, 1967. The Nevertheless.similar work studies elsewhere confirmthe hospital organisation chart is depicted as a set of phsician- approximate magnitudes, and the estimated gap betty !en patient relationships. with line delegation to nursing and the projected needs and resources is a useful signal for necessary remainder of the hospital as staff support. action.

Patient Characteristics as a Factor in Staffing

On any given day, direct care activities in a dicting nursing needs. Dependencies have been nursing unit are dependent on a number of aspects relatively easy to define and observe objectively of the patient population that day. These may and such measures, or the patient classification as- include the following: sociated with them, do correlate empirically with (a) the total number of patients in the unit, patient care as rendered, The weaknesses in using (b) the physical and psychological dependency dependency classification alone for staffing deter- of patients on nursing support, minations are several. In particular they do not (c) the content and sequencing of physician or- indicate time of day of required actions, and they ders, force generalizations of past experiences to the (d) day of stay of patient, present and future. Their virtue is that they are (e) legal requirements and hospital doctrine. an easily obtained signal of abnormally light or heavy demand ahead. Total Number of Patients in the Unit The Content and Sequencing of Physicians' Although mere presence as a patient is not a Orders major determinant of workload, it does require a set of activities, a flow of visitors and inquiries, Much of what nurses will do on any given day the maintenance of records, communications, and may be inferred from the set of physicians' orders preservation of the nurse-patient relationship. in effect for that day. The paper by Dr. Simborg (1)to be presented atthis conference demonstrates the usefulness of orders, not only in estimat- Dependence on Nursing Support ing total workload but in phasing nurse activities throughout the day. The measurement problem in Much of the practical outcome of the nursing using physicians' orders as a guide to staffing lies research under review assumes that measures of in the transformation of an order into a set of tasks patient dependency are adequate bases for pre- for which there are nurse times associated.

37 Figure 1.Calculation of nursing staff needs. Date 8/2,9/742 Hours Building A.)0Nt1141. Degree of care required scheduled Tk 2-c' a) 1 " 0cgffs . e';'In=...... S :LI 0 I ..._ (1)(., ..2 a 01c) 0. =-er fa--c' 01 -is--- c'tu._ 0 REMARKS ..- '1=H _ 73 I 4-- X 44- Pro Aux Total 2 .1-1"0 .. %.-..2-E Cri 1311.a- i: E g c..) CD 7- o a_re rcs .cn (7) -0 -J"P o -''.: o=ri c O=2, e) in(9 4.1)Z-4,4_ti:' CI VI(7)2)cu z cc 4z 31 24 9I 4 11 32 5Z /5 /8 33 -19 -/ iliev , /8 4382 z.frq 2//8 ooq6.-5 2 // 2/.520 4/.540 95 /5:52/ 3625 --56--15 iLig2 5 43.6-20.5:,#,-40,9 1-1RAl - A44-ivc7-7! 74" /4 26.5 /5 - /3S 4-2.5 1-1 Szti AILg4.53 vo 23 2 7 32 /214 14 4.534 /52/ IV18 3q33 #i 1-3.5 -1-1-5 - ,JHH 1-007, 11/64 104/ 28 /3 b2 t at 2:30 p.m. f 9:00 a.m. - 6:00 p.m. aln2!I I (Panul)uo3) -Pie/nit/9M vneerio ____ - . OL bulPlina

Jo o al smoH k fa aathaa 0 Z3 cN -, .4' __I - paqnbai painpaps AT --a_ 11 0 g-ic -0 ,74: c 9, ft S)IIVVV321 2. , a ft o= =-n 'cog c X A) et I --4- -fr to rs.),-. 0 0 ci. 0 in 1 aid xnV le101 =-=- . 771- a c As x aL a a 0 New in ID no c q 13) ,cya g P: =i; 0_ Su -r rt, a 73- 2 0 A) CD 0 to v) ---, c, to -4 13 (.1) s r-- = 0 SW- (E-ticy)(Y--i-c4147- 5V- ze / s .9 8 hZ 6* It 51 1E El- 1iv-di-1.57 II 0 0 II - , -fin,' ivory 5'ft 50- I -xrpod-oiv,- Su LI 0 Z 51 YA -51Z 51Z s'A li Z-f- - -ow.* /vat (z-evadwai) roi-i- 9--"lc inr eikv 5:0-1- 5/- rg zZ Z 0 OZ 5/ 5i b 1r oi 5- -k '' ria""/ (2°1 5'81- 4, Na' lar v 4 4, ''''Y Ar, aswegilvarrpor, 1 11 1,5 1 bb 51- ."14. 5 ig a b 17i 8/ siz- .1- E ten.

1,1i bl Z S 5/ yg a g/ *4- 5v- 5'£- 711 *I tool, a II gv HHr LOO-I t9/TI - W OVZ "urd + 00:6 e 00:9 d ur 30 RESEARCH ON NURSE STAFFING IN HOSF1 rAts

Day of Stay of Patient Legal Requirements and Hospital Doctrine

Evidence continues to accumulate to quantify Society imputes certaincharacteristicstohos- the difference between fi-st clay 'A stay ina hospital patients vt d to those who serve them. Asa pital and subsequent days. Moores (2)reports a result some mandatory staffing practices have sig- nificantinfluence()Itstaffingratios.Registered halving of direct care after the sEcond day. A. nurse coverage during the night and preparation corollaryisthe suwInent that nurse hours per and administration of drugs by professional nurses patient day vary inversely with a hospital's average (RN) are examples. Assessing the effects Jf such requirements does not pose severe measurement length of stay, but this statement obscures theex- problems. Work measurement studies of the ac- planation of significantly large transaction times tivities and frequency of performance estimates relatal tc admission and discharge days. suffice to show the magnitude of these factors.

Organizational Response as a Factor in Staffing

Given the set of responsibilities resultant of pa- problem of random variation in workload. Moores tient conditions, physicians' orders, legal require- (2) reports a smooth drop in staffing ratio from ments. and nursing doctrine, we can address :.Le over four nurse hours/patient day in units of 10 problem of organizational response asitaffects beds or fewer to 2.5 or less for units of 30 to 37 staffing. A number of measures may be used to beds. characterizestaffing:e.g.,thefamiliarratio of nurse hours (or nurse wages) per patient day and Mechanisms of Communication and Control budgeted or total staff positions per bed. Some of the rational determinants of these num- From an administrativepointofview,the bers are as follows: simplest way to determine staffing requirements size of the nursing unit, for a nursing unit of given size and service is to mechanisms of communication and control ohserve its pattern of patient demand over a pe- tu)it, riod of time and assign staff to meet observed external supports of the nursing unit, peak loads. This is an easy but costly process architecture and technology, leading to large amounts of idle time. Given some organizational factors, mechanism fOr short term prediction of patient administrative policy. needs, say a clay in advance, several arrangements are possible each of which in its own way permits accommodation of peak loads without a large re- Size of the Nursing Unit serve staff. Progressive patient care.By grouping patients In any service system subjected to randomly according to level of intensity of illness or de- variable demands, a reserve of resources must be pendency, the variations in demand within each provided to accommodate periods of peak load. group are relatively less than the variation of a The statistical law of large numbers tells us that nursing unit in which all levels of intensity are for any specified probability of available service randomly mixed (3). the necessary reserve capacity is proportional to the Variable staffing.It has been observed that the square loot of the capacity required for average major source of variation in aggregate patient de- demand. By way of example, the reserve for peak mands on any day is the number of intensive care demand in a I6-bed intensive care unit may be patients in the unit, followed closely by the num- half again as large as the base staff. This argues ber of admissions. Given short term forecasts of for larger nursing units to achieve economies of each of, these numbers,itid possible R. predict scale, although there are other ways to handle the roughly the number of direct care hours needed. It

40 FACTORS RELATED TO NURSE STAFFING AND PROIII.EMS OF THEIR MEASUREMENT has also been observed that workloads in the many functions. It is simple enough to measure the bits nursing care units of a hospital are not correlated, of nursing time released by each elemental task except for seasonal variations or epidemics. Again, supported or removed, butitis very difficult to the law of large numbers suggests that the ability find an influence on the staffing ratio. In part this to share personnel among nursing units can reduce is due to the fragmented nature of the time re- staff requirements. or conversely, can increase the leased. In part it reflects that the aggregate of such probability of local adequacy of staff given a fixed time savings may be outweighed by the admission total hospital staff (1). of one or two intensive care patients. Controlled admissions. It was noted earlier that the first day of patim stay, particularly for non- elective admissions, imposes more than twice the Architecture and Technology demand on nurse staff than do subsequent days. The comments above apply equally wellto Given some choice in placing a new admission plus architecture and labor saving technology on the knowledge of the number of intensive and first nursing unit. Specific elements of time saving can day patients on each unit, there is a potential for be demonstrated, butitis difficult to trace the stabilizing workloadsbyselectiveadmissionto elements to a justifiable reduction in staffing ratio. units. Theoretically, controlled admission can have A primaryeffectof architectural layoutis on the same effect on staffing requirements as pro- transit time of nurses in the course of their duties. gressive care or variable staffing. In practice it has The measurement technique called "link analysis" had some drawbacks: is quite useful for observing the intensity of traffic In teaching hospitals, the criteria for patient between important nodes intheunit.Having placement by service and type of care narrow identified the busy linksitis often possible to the options for localizing any particular pa- shorten them by making the nodal points con- tient. tiguous. Even with controlled admission there are still Various studies have shown transit time reduc- substantial variations in workload, calling for tions to be effected by placing medication rooms variable staffing as well to match aggregate de- and criticalpatient rooms close to the nursing mands to nursing resources. stations. A work corridor has been shown to halve Variably work routine.--In searching for a pool transit times if nursing teams remain in the cor- of nursing time to draw upon for peak patient de- ridor and near their patients. mands, one should keep in mind that direct care Technology, in its accustomed role of lifting the constitutes only about one-third. oftotal nurse burden from man's back, has had little to offer staff time. A priority ranking of activities of the nursing. It is most effective in taking over routine, remaining two-thirds may give clues for combining repetitive tasks, and these are generally scarce in daily forecasts of patient needs with other tasks to human services. The promise of technology, as yet create daily routines in some detail. The detail in unexploited, lies inits potential role as part of volvecl is immense, but it can be aided hopefully the communication and control process. Indeed the by computer based information handling proce- only service areas where technology has made con dures. tributions to productivity comparable to those in industry and agriculture arethose in which a External Support of .,ie Nursing Unit large amount of paperwork is involved. We know that formal communications on a nursing unit There have been numerous studies and admin- account for as much time as patient care itself. istrative moves to transfer some tasks traditionally Serious efforts are under way to recover some of performed bynursestoadministrativedepart- this time through computer based systems, but ments in the hospital or to externally purchased these are in difficult development stages. supplies or services. The same arguments for eron- nnies of scale that support the notion of large Organizational Factors nursing units also point to efficiencies in managing housekeeping, dietary, escorting, and sterile supply In its heyday the high productivity of American

41. 32 RESEARCH ON NURSE STAFFING IN HOSPITALS labor was attributed to techniques of specializa- but they are not powerful. All can be thwarted by tion and division of labor. The notion of dividinga noncompliance on the scene of patientcare or by total task into elements and assigning each element administrative fiat. An administrative decisionto to a well trained well equipped worker underlies base nursing oudget ona rule of thumb, say 3.2 most of our mass production techniques. If tasks nurse hours per bed day, assures that the only can be balanced and machine pace,: on a precise variable having an influenceon staffing ratios is schedule, such a system can be nearly 100 percent occupancyrate.Under such circumstancesthe efficient in utilizing its man days in the shortrun. total cost of nursing to a community would be Itis,however, vulnerable to interruptions and strongly related to the number of hospital beds becomes costly when used to less than its capacity. constructed. In a way, specialization and division of laborare consistent with the guild system, in which minor tasks were handed to apprentices with major tasks Empirical Evidence of Factors Affecting reserved for masters. If the apprentice wearied of Nurse Staffing the menial tasks assigned him, he could comfort himself with the thought that the wryprocess as- fx an-attempt to Astra the influence of many of sured that one day he would be a master himself. the factors discussed here on nurse staffing,a sam- It has been this way with bricklayers and plumbers, ple of 31 Illinois hospitals was studied (5). Means lawyers and doctors, but not with the nursing were devised to measure such variables as hos- staff. In the absence of both upward mobility of pital size, nursing 'mit size, architectural configu- auxiliary workers and the tightly controlled set of ration, degree of external support, management operations on a nursing unit, one wonders whether factors,state of technology, patient satisfaction, a hierarchical distribution of tasks can be efficient staff 4nd administrative characteristics. No efforts or effective in the long run. were made to control these variables, but correla- One alternative to the hierarchical distribution tions were sought between their natural ranging of tasks is to require versatile performance by most and the observed range of 3.78 to 5.50 in the members of the team. Indeed in two experimental dependent variable, nurse hours per patient day. units in the San Francisco Bay area, the Kaiser The strongest relationships were found in those Foundation Hospitals have demonstrated a re variables relatedtophysicalplant. The larger duction in staffing ratio from 6 to 4 nurse hours nursing units and the larger hospitals operated per patient day by assigning full care responsibil- with smaller staffs in terms of nurse hours per ities of groups of eight patients to each staff nurse patient day. Also, those nursing unit designs that assisted by nursing aides. An assessment of the required less nursing travel to perform duties op- hierarchical versus versatile approaches to nursing erated with fewer nurse hours. Patient satisfaction careposesdifficult measurement problemsfor also appeared more related to environment (noise, what must be captured lies in the subtle dynamics ventilation, comfort) than to nursing care. of performance of a sequence of tasks by more Variables of a day-to-day nature, such as mix of than one person. Nevertheless the most significant intensity of patient dependency, even with the at- influences evident on nurse staffing are in the area tempts to accommodate this variable by provision of organization of tasks, where the issue is speciali- of an intensive care unit, did not correlate with zation versus versatility. nurse staffing ratios. Whether there is an inherent weakness in many of the variables we havepos. tulated or whether the administrativeprocess is Administrative Policy inadequate to take advantage of them is notre vealed by the Illinois study, but the questions The variables discussed so far may be rational raised demand further inquiry. FACTORS RELATED TO NURSE STAFFING AND PROBLEMS OF THEIR MEASUREMENT 33

Another View of the Nurse Staffing Problem

One emerges from the foregoing considerations by which one can test the degree to which the overwhelmed by the weight of it all. So much ef- objective is achieved. We have dealt with staffing fort has gone into research related to nurse staff- ratio as a measure of effectiveness in nursing care. ing, so many ideas have been espoused, and so What objectives does it reflect, or more important, little can be prescribed as a result. what objectives does it fail to reflect? If the problem of nurse staffing in hospitals were If it is possible to list a coherent set of objectives not such an old one, we might have begun on and their associated measures of effectiveness, a another tack, asking fundamental questions about next step in systems analysis is to ask what varia- the objectives of the health care system and the bles affect these measures of effectiveness that can possible roles of nursing in achieving those objec- be regarded as outcome variable. tives. We may find an optimal balance among We seek a definitive list of two kinds of input preventive care, ambulatory, and inpatient care variables:controllable anduncontrollable. We that would diminish the total number of inpatient must then estimate the influence of tacit of the days of care. One signal of success in achieving input variables on the outputs. This is essentially such a balance in the large system would likely be what we have been doing in a limited way in this an increase in nurse staffing ratios and per diem paper and in the research it describes.If as a hospital cost, as outpatient care is substituted for result we had, over the years, produced an ade- some hospitalization. quate supply of nurses who were happy in their We have the opportunity to ask in some detail jobs and who, along with physicians, administra- what the beliefs, aspirations, and goals are of vari- tors, and patients themselves, were satisfied with ous segments of society relative to health care and the care rendered, then no more need be said of hospitals. If we wish to confine our considerations objectives and variables other than those we had to hospital care, we can do so after visualizing the been considering explicitly.. But the factis that hospital's goal in a larger system and forecasting many objectives are still to be achieved. We still the characteristics of the inpatient census. Surely are not able to specify an ideal nurse staffing ratio. there will be inpatients, and we can begin by The factors we have considered have not been asking their wants and objectives. What are the shown by themselves to explain or control the objectives of administrators, physicians, and nurses staffing ratios experienced. We must continue our themselves? Are some of the objectives conflicting? inquiries and experiments to bring to bear new If so, how is the conflict to be resolved? knowledge on the old and deeply human problem A well stated objective should imply measures of nursing care.

43 34 RESEARCH ON NURSE STAFFING IN HOSPITALS References

(1) Simborg, D. "Impact of Computerized In- Michigan Graduate School of Business Ad- formation Systems in Hospitals on Nurse ministration, Bureau of Hospitals Adminis- Staffing." tration, Ann Arbor, Michigan, 1964. (2) Moores,Brian."ReallocationofStudent (4) Flag le,C.D. "The Problem of Organiza- Nurses to Match Patient Requirements in tion for Hospital Input Care." Management Different Hospitals on the Assumption that Sciences:Models andTechniques, New a Shortage of Staff Leads to Increased Wast- York: Pergamon Press, 1960. age."Unpublished DoctoralDissertation, (5) IllinoisStudyCommissiononNursing, Johns Hopkins University, Baltimore, Mary- NurseUtilization,Illinois:A Studyin land, 1971. Thirty-one Hospitals. The Illinois League (3) Weeks, L., and Griffith, J. "Progressive Pa- for Nursing and the Illinois Nurses' Associ- tient CareAn Anthology." University of ation, Chicago, Illinois, 1966-70. DISCUSSION OF DR. FLAGLE'S PAPER

DiscussionLeaders Dr. Alexander Barr Dr. Richard C. Jelinek Chief Records Officer and Statistician Vice President for Research and Development Oxford Regional Hospital Board Medicus Systems Corporatim. Oxford, England Chicago, Illinois

Dr. Barr

Good morning, Madam Chairman, Ladies, and Research is going on in each of these areas, but Gentlemen. It is always a great pleasure for me to there isa certain amount of speculation as to listen to Dr. Flag le, and I am happy to comment where we go in the future. Hopefully, these vari- on his paper. If I may use the blackboard, I shall ous elements, research components, will somehow use my 5 minutes this morning to draw you a write them into a special box that we will have to diagram(seefigure 1) illustratingaresearch look on at the moment as a black box of total framework into patient care and nurse staffing in patient care. If we could define this black box, the hospitals. This gives me an opportunity of indi- research framework could be drawn with greater cating some of the things we have been doing in confidence. The real danger is that in our striving Oxford, and I hope it also illustrates that there for efficient organization the real business of pa- isn't a very wide divergence between our thinking tient care will get overlooked or given4i. low, pri- in England and the thinking here today. ority, 1 Our basic prcblem, or the one we think we Let me tell you a story about an American havelet's startthat wayis todefinehospital general, which may help to illustrate the sort of staff complement. As far as we are concerned in relationship I think we have gotten ourselves into. the United Kingdom, we have tended to look at An American general was supposed to have had a thisin two separate ways. First of allthereis problem in Vietnam. As is usual nowadays, he deployment of staff and, in particular, deployment asked a large computer in the States to solve the of nursing staff. This can be divided into two problem for him. We don't do our own thinking separateareasofresearch.Oneoftheseis nowadays; we ask a computer to do it for us. The trainingtraining ofnurses or training of any general cabled the computer and said, "I'm sur- staff, and the other is allocation of staff to the jobs. rounded on the north, south, east and west. What On the other side we must look at the service should I do?" And the computer cabled back and element, the actual running of the hospital and the gave him the very tersereply, "Yes." Nothing delivery of care to the patients. This also has two else, just, "Yes." And the general became a bit branches. One of these is the quality of care that irritated at this and he cabled back to the com- we want to give. The other is the quantity. We puter, "Yes, what?" And the computer returned have already heard something about quantity and the reply, "Yes, Sirs" quality and been given to understand that the How can the problem be defined in a way that bridge between them is a bit flimsy at the mo- it can be answered intelligently? I rather suspect ment. that we have done much of the research because it

45 RESEARCH ON NURSE STAFFING IN HOSPITALS

Figure 1.Interrelationship of staffing and patient care.

Hospital nursing staff complement

Deployment Service -if

Training Al location Quality Quantity of staff of staff of care of care

Total patient care DISCUSSION OF DR. RADIX'S PAPER 37 is relatively easy to do. Topics are selected because patient. Most of us recognize there are limitations we can, in fact, get solutions to them, and it always to the actual method, but in fact it has some very looks better to have some sort of a solution in useful practical elements attached to it. hand. However, asI have already pointed out, On the quality of care,Ithink most people perhaps we should have been asking a different agree this is a very difficult area to work in. As a type of question. What do we want to give the preliminary we have looked at what the staff think patient? What is total patient care? And follow- about the care given. We have asked professional ing this, what does the nurse do about the delivery nursing staff of all grades along with 'doctors and of this care? paramedical staff to classify wards into goodor Sooner or later hospital care will have to be re- poor care, into strict and slack discipline, and into lated to the community services because obviously happy and unhappy wards. This has been very, the sort of care that is given in the hospital has useful because there is a fair degree of conformity its counterpart in the community itself. Early dis- of thought among the staff about wards in general. charge exacerbates the load on the community Dr. Revans probably has a lot of information on healthstaff.Ifpatients are not discharged very this as well. early, you obvichisty live It --ii iitttent pattern of We also extended that to ask patients what they care withinthehospital, perhaps relieving the think about the l-lspital service provided, and community of a lot of its present problems. The again there is a high degree of correlation between point worth emphasising isthat these are inter- what the patients think and what the staff think. dependent and not independent services. Perhaps we haven't made enough of the con- Just before I finish, may I draw your attention sumer reaction in hospital affairs. We should be to one or two specific things we have been doing more attuned to what the patient thinks about the inthisfield? On the training of staff we have service. found some conflict between training and serv- Then, on the black box of total patient care, ice.What the Training Council requiresThe which is probably the most difficult area for all, General Nursing Council in Englandsometimes we have recently been looking at the possibility of is at variance with the actual hospital needs as using videotape records of the actual service pro- far as services are concerned. vided in the ward. This is a very promising tech- If the training model isunbalanced (for ex- nique. It enables us to capture data and to use ample, the time spent by students in a specialty and assess it over and over again. The trouble, of is too long), it has repercussions in the staff com- course, is in reducing this data to some sort of plement. You can, infact, have increased staff manageable indices, but again I don't think that complements just because training is scheduled inthis in itself should deter us. If we did capture this efficiently. On quantity of nursing care, the nurs- kind of data and showed itto the relative pro- ing dependency based on the work Dr. Flagle has fessional staff, we might through experiment and done in Johns Hopkins is very useful for measuring discussion arrive at a stage where we could say the quantity of nursing care actually given to the w:tat total patient care is all about.

Dr. Jelinek

I am probably the wrong person to critique Dr. Flagle made. I have been for 6 years in the aca- Flagle. As you all know, Dr. Flagle is considered demic environment involved in health carere- the "father" of operations research applications in search. For the last 3 years I have been associated the health care field and in a way is like God. To with an organization attemptingto implement critique him would be the same as to ask you to some of the research ideas. I would like to use ex- critique- God and, therefore, I have to agree with periences stemming from both the academic and everything Dr. Flagle said. the operational settings to comment on Dr. Flagle's However, I'd like to make a few points, most of pa per. which directly support some of the statements Dr. The problem, in my opinion, in the area of

. . 47 38 RESEARCH ON NURSE STAFFING IN HOSPITAI.S nurse staffing is that we do not have any objective therefore is often found more in works than in means of relating staffing to the level of service. actually achieving the established objectives. This can best be illustrated by suggesting that in To conclude, I would like to make a number of most hospitalsitisnot possible to observe any observations and raise a number of questions re significant impact on the level of service or quality lated to this area of research. of care resulting from a change in staff size. Most First, the problem that faces all of usisthe certainly sur.h impact is difficult if not impossible need to establish objectives for the nursing pro- to quantify. Patient care will not go up a great deal fession as a whole. What is nursing? For us as if you double the staff, nor will it decrease signif- engineers working with nursing itis hard to do icantly if you reduce the staff by 50 percent. anything until we understand what it is that nurs- Of course, this is a tremendous problem. How ing is trying to achieve. do you staffif you cannot evaluate a situation Second, there exists a problem of explicitly de- such asI have described? This, in fact,is what fining certain terms used by nurses in identifying Dr. Flag le pointed out through some of his illus- themselves. For example, what is clinical nursing? trations. As his presentation has shown, ona par- We keep hearing the difference between task ori- ticular patient unit the staff may he twice as high ented nursing and clinical nursing. We have diffi- on a per patient day basis one week as it is another culty getting a precise definition of what is meant week. The impact of this variance on patient care by clinical nursing, clinical nurse specialists, clin- is not understood, nor is it measured. We also see ical leadership, etc. Many terms have been used to tremendous differences in staff site and in cost of define the nurse and her role. running a patient unit from one geographic area What is the psychosocial element of care? Any of the country to another. It is not clear that qual- engineering study that comes out in terms of ob- ity of care is related to these regional staffing and jective measures k generally criticized in terms of cost differences. Of course, the main probletnis lacking an understanding of the psychosocial asthat we do not have any clearcut measure.; of qual- pect of care. What exactly isit? We know itin ity. Such measures will have to be derived before words. We have difficulty quantifying it. the staffing problems can be resolved. What is meant by continuity of care? We know For purposes of discussion I would like to raise there are advocates of continuity of rare. One way the following question. If in fact there exist such this is expressed is to propose that the same nurse large differences in staff on a per patient day basis provide continuous care to the same patient. But betweeninstitutions without comparablediffer- we also know this is impossible to achieve totally, ences in quality of patient care, why not immedi- because the nurse has to sleep and take her days ately cut staff size in those hospitals that show rela- off. What then is continuity of care? tive overstaffing? Other than major reduction in Next is the question of relating research to op- cost, what would be the impact of such action? I erations. Are there enough ties? We see nurses am just asking the question, not making a recom- doing research in the academic environment. Are mendation. they doing this research for the sake of publishing I would like to make some general comments and being recognized among their peers or are regarding another dimension ofthis area ofre- they doing itto improve operations? 1 believe search. A great deal of effort has been devoted to research can be clone for the purpose of being this subject and many conceptual ideas have been recognized inthe academic community, but the advocated, including patient classification scheLies, question I'm posing is, Do we have enough of a variable staffingprocedures,selectiveadmissions tie between research and operations? procedures, and reporting systems. However,very Even more important is the tie between educa- few of these ideas have been implemented ina tion, and research and operation. I think this is an broad scope. For example, we see many hospitals area where the Division of Nursing is playing an having variable staffing, but when we look closely important role in trying to establish a better tie, at their system it imposes little control over match- but a significant amount of additional work is ing staffsizeto workload. The implementation needed. DISCUSSION OF DR. PLAGUE'S PAPER 39

I have a general comment regarding the alter- not ?hie to establish what the effect of each of native forms of organizing patient care operations. these systems is on quality of care and cost. As Dr. Aydelotte pointed out, we now see many My observations boil down to stressing the tre- organizational systems in operation. For example, mendous need for establishing objectives for nursing and for devising methods for measuring level there are many variations of unit management and quality of care. I believe that if this conference systems. We have many variations of the use of could somehow locus or direct us as to how we clinical nurse specialists. All of these systems are should handle these problems we could make a feasible in that they get the job done, yet we are contribution to the field,

49 Measurement of Patients' Requirements for Nursing Services

Mrs. Phyllis Giovannetti Project Director, Nursing Research Hospital Systems Study Group University of Saskatchewan Saskatoon, Saskatchewan, Canada

IVe dance round in a ring and suppose (DVA) Hospitals. The second investigated the re- But the secret sits in the middle and knows. liability and validity of a completely subjective Robert Frost classificationtoolfor determiningpatients'requirements for nursing care. The third study, somewhat more comprehen- Iam pleased to share with you the results of sive than the others, began with the development three research studies conducted in Canada as well of an objective classification tool that was subse- as many of my own views on the subject of pa- quently quantified to produce an index of work- tients' requirements for nursing care. load. In addition, this study developed and tested a Although none of the research investigations set new concept of ward organization, the Unit As- out principally to measure patients' requirements signment System. All of the studies were supported for nursing services, they did come face to face in whole or in part by Canadian National Health with this problem at an early stage. They focused Research Grants. identificationofpatients'nursing care on the The underlying theme has been predominant in needs, and where attempts were made to quantify the vast majority of related studies of matching these needs, the researchers had to conduct their nursing resources to the nursing care requirements investigations withinthe realm of actual care of inhospital patients. The keynote was effective given. We arestillfaced with the prohlem of evaluating whether in fact the care given is that utilization. which is required. Filially, some comments regarding guidelines for The first study developed an index of care, a future research are presented. A descriptive model documentation of the types of nursing activities for the study of nursing as an interdependent vari- carried outin Department of Veterans'Affairs able in the health care system is discussed.

41 42 RESEARCH ON NURSE STAFFING IN HOSPITALS An Index of Care

The firstinvestigation (1) was carried out at cording to the type of care provided to them had DVA Deer Lodge Hospital, Winnipeg, Manitoba, been developed. It required only that the charge in 1965. It arose out of an interest in modifying nurse complete a Level of Care Assessment Return morbidity and mortality data to show the type of for all patients. Following data processing, total patients being cared for in DVA hospitals across patient census was divided into three levels ofcare. Canada. Prior to the study the statistical data col- Further validation was established in a number lected under the Veterans Treatment Regulations of hospitals through the timing of the procedures provided little information to indicate the type of listed in the Level of Care Assessment Return (2). care the patients were receiving in the hospitals. Both the direct and indirect care times associated A classification of patients according to the type with each procedure were obtained for eachpa- of care provided to them was required. tient level. Through continuous observation for 24 A questionnaire was devised utilizing a selection hours the total amount of direct nursing care re of procedures that appeared to best indicate nurs- ceived by each observed patient was determined. ing efforts. This questionnaire became known as The timing studies demonstrated that the Level of the Level of Care Assessment Return, a checkoff Care Classification of a patient, correlated with the sheet of some 55 items to describe a patient. The amount of care he received and measured in time, 55 items were grouped under the following head- is a valid method of categorizing the severity of the ings:ClinicalMonitoring,TechnicalNursing, case. Nontechnical (basic) Nursing, Physical Medicine, The research resulted in a method of classifying and Organized Psychiatric Unit. patients, independent of observer judgment, into At the time that the Level of Care Assessment levels of care and a system for estimating the Return was being completed on every patient in nursing workload associated with each level. In the hospital, the patients were all classified accord- addition, the analysis contributed a great deal to /mg toseven predefined care categories ranging our knowledge about the variables influencing from intensive to residential. staffing. Analysis of the results of the survey showed that There has been, however, a limited amount of although the amount of direct nursing care ranged application to non-DVA hospitals. I would propose from a large amount to a small amount, only three three major reasons for this. First, the classification levels of care could be identified: Level Iminimal system is based totally on the physiological needs amount of direct nursing care, Level IImoderate and care of the patient. Instructional and emo- amount of direct care, Level IIImaximum amount tional needs, independent of the nursing procedures of direct nursing care. In fact, the residential, in- studied, do not appear to have been considered. Ac- vestigative, and convalescent care categories were cording to Dr. MacDonell, the assessment provided grouped intoLevelI(minimalcare),the ex- little evidence of emotional or psychosocial needs tended treatment and intermediatecare categories being met apart from that accompanying regular grouped into Level II (moderate cal.!), and the treatments. Second, the nursing activities and proce- acute and intensive care categories into Level HI dures identified and their determined times may (maximum care). Analysis of the completed Level be peculiar to the types of hospitals studied. Third, of Care Assessment Returns matched with each the patient mix in terms of kinds of care neces- level revealed the procedures of direct patient care that characterized each care category. sary, coupled with the treatment modalities, may A numerical value was assigned to each of the not be similar in other institutions. 55 items of the Level of Care Assessment Return, Regardless of the limitations, real or perceived, bearing in mind the amount of effort involved in the system does provide a more accurate account the particular procedure. This step made it possi- of patient care than the usual census or diagnos- ble to produce a classification of patients according tic category estimate. For the DVA hospitals across to the levels of care by data processing methods. Canada precise information regarding type of care Thus a method of classification of patientsac- provided to patients at any given time is available

51 MEASUREMENT OF PATIENT'S REQUIREMENTS FOR NURSING SERVICES 43 and allows a more realistic judgment to be exer- (fictions and projections concerning future require- cised concerning immediate needs of the hospitals. ments forfacilities and personnel.Finally,the Trends in the types of patients for whom care is information can be utilizedfor staffing patterns provided maybeidentifiedonthebasiso; indicatingtypes and numbers of nursing staff accumulated returns, making it possible for pre- within the hospitals studied.

A Subjective Classification Analysis

The second investigation (3) was conducted at wastotestfor differencesinthe way various the Vancouver General Hospital, Vancouver, Brit- nursing personnel from different hospitals classify ish Columbia, during 1970. The study was initiated patients. in the interest of testing the reliability and valid- It was found that there were significant differ- ity of a completely subjective patient classification ences in the way different categories of nursing tool.A comprehensive review of classification personnel classified patients, both within a hos- methodologies hadledtotheconclusionthat pital and between hospitals. Thus, using discrete existing tools paid little attention to the patients' analysis, the subjective claisification tool as used requirements for emotional support and instruc- by the nurses could not be considered a reliable tional needs and were too restrictive in terms of method of determining patients' needs. the predetermined criteria selected. In addition it There was a tendency, however, for the nurses was felt that the more commou three and four to become more discrete in their assessment of category classification systems were not sensitive patients' needs as their familiarity with the classi- enough to the diverse needs of patients in acute fica on tool increased. An alternative form of care hospitals. analysis was presented that allowed one degree of A criticalindicator, independent of observer scale difference between nurse classifiers. Using the judgment, to encompass the diverse psychosocial alternative analysi, the tool was found to be re- needs of patients was not apparent. Therefore liable. the notion of an objective checklist was discarded While the nursing personnel were classifying in favor of a completely subjective tool. A classifi- the patients using the designed tool, a nurse re- cationtool was developed withfivecategories searcher obtained additional information on each ranging from minimal to intensive care.It was patient. This information was comparable to the this completely subjective tool that was selected objectiveclassificationtool developedat Johns for the rigid reliability and validity testing. Hopkins Hospital and to the procedures docu- Five acute care hospitals provided the study mented in the CASH studies on staff utilization setting, using one adult medical and/or surgical (4,5). The validity of the five category subjective nursing unit of approximately 35 beds from each. classification tool was tested and confirmed against All patients in the selected units were classified the Johns Hopkins system. The patient character- on five different occasions independently by four istics data fashioned after the CASH studies were categories o. nurses. The head nurse classified all collected primarily for their possible contribution patients. Registered nurses on the unit classified to a measurement of workload related to the five those patients assigned to them. One nurse born categories. These data, however, were not analyzed another nursing unit of the same hospital classified at the conclusion of the report. all the patients as did a nurse from one other The study contributed a great deal to solving hospital. the problems encountered in establishing reliabil- This procedure allowed researchers to test two ity in classification tools. In addition it identified major hypotheses of reliability. The first was to some of the problems to be reckoned with if we test for differences in the way various categories are to use classification tools to provide compara- of personnel in a hospital classify patients:i.e., tive information on more than one hospital. Even head nurse, float nurse, ward nurse. The second in the face of the reliability restrictions the tool

52 44 RESEARCH ON NURSE STAFFING IN HOSPITALS

can represent agEossindit atm of the nursing care times more informative than the presently used needs ofthe patients an indicatorthatisfive criteria census.

A 5 -Year Study of Nursing Problems

I would like now to report on a third study Identification of the characteristic needs of each justrecently completed after 5 years ofinvesti- category of patientin terms of the amount of gation. The study was conductedlrthe Hospital nursing care time they receive, the type of staff Systems Study Group at the t iniversity of Saskatche- who care for them, and the type of functions per- wan in Saskatoon, Canada, with most of the data formed was the second step. This was achieved collected in the University Hospital in Saskatoon. by conducting direct care (inroom) activity stud- Five major reports dealing with different phases ies on both general medical and surgical nursing or aspects of the study were published during the units. period of investigation. The results of the study documented a wide PhaseI began in 1966 with the objective of variety of problems: finding an operational solution to arc problems of There was wide variationinnursing care reorienting hospital organizational structure into time spent with patients of different categories. direct contact with the patient and his needs (6). There were extreme fluctuations in workload It is not surprising that the first step was to identify when one or more intense care patients were the needs of the patients. removed or added to the ward population. A four categoryclassificationsystem was de- There was significant difference in allocation signed to objectively categorize patients accord- of functions among different levels of staff de- ing to their needs for nursing care (see figureI). pending on the care category of patients served. Font components of care wereisolatedas the There was variationin magnitude andin critical indicators of the patients' overall require- character of the workload to which staff must mentsofnursingcare.These components adjust. personal care, feeding, observation, and There was lack of clear role definitions for ambulation--were further broken down intoin- the various levels of nursing personnel on the dividual determiners that represented a patient's ward. degree of dependence upon nursing. Two addi- Staff appeared to enter the toom to perform tional determiners were found to be necessary: a single function and then leave. one concerning incontinence and other regarding Numerous people participated in the patients' surgery. daily bedside care, often for only short periods The classification tool is in the form ofa check. time. Intercommunication and supervision off list comprised of the 13 determiners, each des- and checking of subordinates appeared to in- ignated according to the amount of nursing involve a considerable amount of time, effort and volvement they represent: Aminirnal, 8average, confusion. Cintense. Continuity of assignment was difficult or im- Although the tool has the appearance of an possible. resulting in limited opportunity for objectiveclassificationmethod,thereisinfact nurses to know their patients as individuals some subjectivity involved, Thisis mote clearly and thus to provide personalized patient care. evident in the guidelines developed to assist the Considerable amount of time was spent by nurses in selecting the correct determiners of care. nurses "in flight" between the center of com- The guidelines, presented in appendix II, set down munication (central nursing station), supply the parameters for allof the critical indicators. area (service room), and patient's bedside. Of particular importance is the observationcom- The unpredictable nature of the workload ponent. This component provides for the inclusion made it impossible to staff in a manner that of emotional and instructional needs of thepa- would ensure a consistent level of nursing care tients. matched to patient's need. Periods of signif-

53 4EASUREMENT OF PATIENT'S REqUIREhtENTS FOR NURSING SERVICES 45

Figure 1.Patient classification tool.

PERS)UAL FEED- OBSERVA- AMBULA- WARD CARE ING TION TION _

0.-. I .... ,.., ,-, .-- ...... ,-. .-. .-. c 0 L.1 DATE c., co 4 c c do:, umcum< ...... 4 v ...4 v 0 4,-. vvv vvvv 0 0 V P0 P 0 0 00 ...1 0 0.0 04 4 444 fi4+ 10 pi os o al 4+ 0 v .0 a. 0' c04,n4 0 0 pi 0 V 0 4+ 06+ 0 0 4000 0 * A OPOVI V RI O A Vk 0 P 0-I 00 400 OP0kZ X X X 4 44 *el 0 0.4 X A 0A el X X 4+ 4 0 0 0' 11 V piP '41'0 44 0. 4k* 0 4 00 0 ..4 P 0 0 VA 0 0 0 0 V 0 cm IS 1 .4 A V 0. 4.m,4 P 0 i As A 0 * .c w 0 V Z< IC 1., 04,0012k0 .0 k m o IS RI V b 0 N /0 0Or 1* 0. 1-4 A. u PATIENTS c.) i 03 0 01 04 0 0' 0' 4-4 0.= 0= A B C Comments _ . 1

Brown, Mr.M A.A X X, X 3 0u 1 Minimal care ,

Wacker Mrs. W. X X X X 0 40 2 Avera e care Aove Smith, Mr. G. X X X X 0 13 3 average care

4 Intensive care White, Miss G. X X X X - 0 0 5

... %.'---- ___../---s-----.-"---'--- 46 RESEARCH ON NURSE STAFFING IN HOSPITALS

icant overstaffing and gross understaffing re- nightshift. Thus theratioof the amount of sulted. nursing care received on clay, evening, and night These problems concurred with those expounded shifts was 4:2:1. by the nursing staff during a period characterized By multiplication of the day shift figures by four by high turnover rates and low morale. During and the evening shift figures by two, the result- the course of the research, every effort was made ant numbers become the actual index and reflect to find operational solutions to these problems. not only the relationship among categories on a Since many of the problems were more clearly shift but the relationship among shifts for each expressed as a result of the categorization of pa- category as well. Table 1 shows these relationships. tients, the next step was to validate the classifica- These ratios may in turn be used as weighting tion tool. factors for the corresponding categories on each Validation was carried out by comparing the shift to arrive at an index of patient loadon the results of the classification tool withthe head nursing unit for any shift. nurse's subjective opinion and with the Deer Lodge The procedure is very simple. If, for example, Level of Care Assessment Return (7). The head nurse was instructed to evaluate the patient's need the ward census is composed of 10 minimal, 20 average, 2 above average, and 1 for care on the basis of her knowledge and exper- intense care pa- ience and to assign each patient to one of four tients the workload index for the day shift is de- levels of care ranging from minimal to intense. termined by multiplying the weighting factor for each category for the day shift by the number of A project nurse classified the same patients using patients in each category and summing the prod- the Phase Itool and the Deer Lodge Level of ucts.That is, (10X Care Assessment Return. 4) + (20X 8) + (2 x 24) + (1 x 48) an 296. The same pro- There was no .significant difference among the cedure would he used for the evening and night three systems when the chi-squares test of good- shifts and, assuming the same patient mix, yield ness of fit was Applied to the data. At the same workload indices of 156 and 89 respectively. time, the Phase I tool was found to be easier and less time consuming to use than the Level of Care The impact of the workload index when calcu- Assessment Return. A general description of the lated over a period of time on a nursing unit can be very informative and the relationship be- four categories is presented in Appendix I. . tween census and workload most revealing. Table 2 shows the patient load index for all three shifts Workload Index during 1 month on one surgical nursing unit. It can be seen that the census ranged from a low of As aresult of the extensive activity studies 28 patients on the 21st day to a high of 38 patients conducted, the relationship among the four cate- on the 28th day. However, the day load index gories of care, in terms of the average amount of on the 21st day was quite high at 492, but on the direct nursing care patients received, was identi- 28th, with 10 more patients, the index was much fied.It was seen that on the day shift average lower at 396. Similarly, on the 5th and the 12th patients received twice as much, aboveaverage of the month the census was equal at 35 patients, care patient:, six times as much, and intense care but the day load index was 512 and 272 respec- patients 12 times as much direct nursing care as tively for the 2 days. minimal care patients. The ratio among the cate- gorieson the day shift became identified Table 1.Ratios of direct nursing care time as 1:2:6:12. Subsequently, the ratio of the average received per shift and patient rattgory direct nursing care times on evening shift was found to be1:2:7:14, and on the nightshift, Category 1:2:7:25. Shift Above Further analysis revealed that, on the average, Minimal Average average Intense patients received four times more direct nursing Days 4 8 24 48 care on the day shift than on the night shift and Evenings 2 4 14 28 twice as much care on the evening shift as on the Nights 1 2 7 25 MEASUREMENT OF PATIENT'S R EQC MEM ENTS FOR N"RSING SERVICES 47 BEST COPYAVAILABLE Table 2.Daily patient load index

Patients per category Load index Above Date Minimal Average aierage Intense Census Day Evening Night 116 1 4 19 7 1 31 384 210 110 4 16 7 1 28 360 198 3__. 8 17 5 3 33 432 238 152 166 7 19 3 4 33 444 244 286 206 5 _ _ 8 15 9 3 35 512 110 6 . 6 17 10 - 33 400 220 148 7______. 6 IS 8 2 34 456 252 8_._ _ 6 21 5 - 32 312 166 83 97 9 5 23 3 1 32 324 172 74 10_ 9 20 - 1 30 244 126 156 89 11. 10 20 2 1 33 296 81 12 12 22 1 35 272 140 142 13 14 16 3 3 36 400 218

1 272 140 81 14 . 10 23 - 34 428 236 162 15..... 9 16 3 4 32 106 16. 12 15 2 2 31 312 168 142 82 17 10 20 1 1 32 272 135 18 7 18 6 2 33 412 226 161 19 7 15 7 3 32 460 256 244 166 20 6 16 4 4 30 440 276 215 21 7 13 1 7 28 492 496 276 193 22 . 4 18 4 5 31 NW 21 6 4 33 512 284 186 34 504 280 173 24_. . 4 '440 8 3 464 256 139 25 . _- 2 21 10 1 34 165 26 6 21 6 3 36 480 264 248 146 27. 5 21 7 2 35 452 128 28 ..9 24 3 2 38 396 212 153 29 27 3 3 36 444 240 143 30 6 24 2 3 35 408 220 31

Table 3.Conversion of index to staff required

Derivation of conversion method Example Minutes of direct nursing care required on the Kard INDEX x M 200 x 5 + 1,000 min. INDEX x NI 200 x 5 Hours of direct nursing care required on the ward 16.67 hrs. 60 60

Hours of direct nursing care provided per nurse per fl hour shift . I' x 8 hrs. .42 x S = 3.36 hrs./nurse Total nursing staff needtd to meet all care requirements for the INDEX x 111 16.67 hrs. req'd. 5 8 hour shift 60 3.36 hrs./nurse nurses p 8

Simplifiedexpreision INDEX x 51 200 x 5 200 5 nurses P x 480 .42 x 480 40 "Ntirse" represents ativ nursing staff member, including registered nurses, certified nursing assistants, and orderlies, whose primary role is to providt. d,rcct patient care. Mis the average number of minutes of direct nursing care received by a tnittimal care patient on night shift (e.g., M = 5 minutes). Pk the percentage of nursing staff time spent on direct nursing care per shift (e.g., P = 42% = .42). Indexk the workload index for one shift (e.g., Index = 200). 48 RESEARCH ON NURSE STAFFING iN HOSPITALS

Figure 2 represents a graph by day of the month the total direct and indirect care requirements of of the patient load indices given in table1. The the nursing unit was obtained. Table 3 represents range on the day shift extends from a high load the derivation of the conversion method of work- index of 512 to a low of 244. The daily fluctua- load index to required nursing staff. don in workload is evident through the month Figure 3 graphs the daily workload index per and particularly so when one notes the decrease shift over a 1-month period. The required number in workload from the 7th to the Sth and the sud- of nursing staff as determined by the conversion den increase front the 14th to the 15th. method is also shown. Thus, for a specified work- The fact that census may be most misleading load index, the number of staff required may be when used as the sole basis for judging workload estimated. The broad horizontal bars in this graph is obvious. The problems related to providinga number of nursing positions. Thus, as indicated consistently high standard of patient care in the on the clay shift, the level of activity for the 10 face of dramatic daily fluctuationsinpatients' budgeted positions should not be expected to ex- nursing care needs are also obvious. tend beyond the range of that which 91/2 to 101/2 When several months of graphed information people could manage. It should be noted that the are compared (seefigure2),trendsinthepa- head nurse and assistant head nurse are not con- tient care load over a period of time become evi- sidered part of the budgeted nursing staff, as only dent. They provide an objective basis on which to those staff members whose primary role is to pro- compare patient load from dad to day or month vide direct nursing care are included. to month on any given nursing unit or to compare The utility of such graphs extends far beyond between nursing units and to substantiate, where the realm of academia. Indices that peak far above census cannot, real changes in workload. the bounds or baseline provide documented evidence that additional nursing staff are temporar- Workload Index as a Staffing Guide ily needed. When the indices of more than one nursing unit are peaked above the baseline, the Relating the workload index to the required graphs become useful in aiding administration to number of nursing staff was the next step in the set priorities in the allocation of the nursing staff project and necessitated that the number of hours available. Indices that dip below the baseline may indicate that staff members should be transferred of nursing care that the index represented1.7. known (8). This information was infacteasily temporarily to another nursing unit. identified as a result of the activity stueies. It was Finally, the need for permanent reallocation of observed that, on the average, the minimal care nursing staff becomes evident in the face of con- patients on nights received 5 minutes of care.' Since sistently high peaks above or dips below the base- the ratios of categories to each other and between line. The daily calculation of the workload index shifts are all expressed as multiples of the amount coupled with a flexible staffing policy reduces to a of care a minimal care patient receives on the large extent the great variability in workload to night shift, the number of hours of direct nursing which a given number of nursing staff must adjust. care required could he obtained by multiplying Closely matching the staffing on each shift to the the calculated load index for any shift by 5/60. nursing care needs of patients can do much to The activity studies also show that on the average promote a consistently high standard of patient the nursing staff spent 42 percent of their 8 hour care. day (3.36 hours) giving direct careto patients. By dividing the number of hours of direct care The Unit Assignment System required by the amount of direct care provided by one nursing staff member, an estimate of the The development of a simple objective tool for represent the upper and tower limits for each shift classifying patients according to their requirements that may reasonably be expected by the budgeted for nursing care afforded the opportunity toin- number ofnursingstaffrequiredtoprovide vestigate more clearly the different nursing activi- 'Further studies revealed that 3 minutes of care was more ties required for each patient category. This in- likely to be the average for minimal care patients on nights. vestigation resultedina new concept of ward 49 MEASUREMENT OF PATIENT'S REQUIREMENTS FOR NURSINGSERVICES BEST COPYAVAILABLE

e . . ...

141. r- 4 N ! I ____t._._ ., I te )- -I -4 1---4----t---f f. er ______

i . ..- N

1 ; / N . --*%:- ! I ; I 14% 1 I 1 i

I i ...... _ 44- ---f ,4 all .!. __1.--4,..., -. ------#11-;;._ +- --""1 iii, ... Ni . -----ii . ... SI -,-- t I ,- wo I 0 , I I Imo 4 44: !.1 1 1- 7 , a i I ! , 1 , 4. 1 . ago. 1 , ! aria t ,60. M - 8 4 --- --t----.- 2 */ -,---Ilesrl- i I/

... .

. , . I . 1 . i 1.. N 1111 5 f 1 i 0 _..... Nal __i--, 1 : Mai

. I 1 40. di.al. 00. i -1143.1"- kr... (4. gill . it 1 -1--- --I ow de 1 I iggeggiiii ar!.i."au. I 1 I el ,,§4.-1 1------1--- I 1----j- 1"--114INI--1------i-----.---i--- eel

I 14% 1111111112M ; ibiTilli I

.... - 0 0 8 0 00 g 8 8 kg) N

Or , . . . . . 1111 ' . ' MINNMUM I I Ol IMIMNIEMIMN Ril NOMMEN' CENWORD:'Mil Azwv )24111"MArMINIMir 4E' Mr r 4/4/ 4 .r./ !MI glEINWASOZVIS772r AN.PF7 EMIMIlill1R11111= OW" " Er ' IhAr Anwar Amp'Awl ,wzwipArastwom.APiii'AmpEzr111111119110111 7 PA A' .r Ar." z*E. I .;11.&f rA 11111111141 .F.1' rz40' ' 0' AP mar 4fA ' A . INNINIIIIS 13111111,11 . NI 111111111111111111111111311111011 01NE Via SINEW iniligili Akir :EN'OPAW.08/4EMI MV'W 111111MIENAEIWilltrAiiiiNa INIM i WIEN!151111"IMMEN ANEW J.'a' AN' mmWENNEU .W.A" _Nun. mv.AvAr ,ramor m IIIIIII1 11111 ENUSER. ,Paf,m9A ' MS' 11 ME EM = I Ill as 0[/©SOHO ' .. ... ElMOO MC Ie...... NMI.1111111111. IMO N 11511 I MEASUREMENT OF PATIENT'S REQUIREMENTS FOR NURSING SERVICES 51 organizationan organization that would focus The system entails decentralization of the ward classified theattention ofthe health care team on the structure by dividing it into units of care patients'needs and care.The activitystudies as intense, above average, average,and minimal had reveuled that the central nursing station, tin- care units. A unitis defined as that number of der a modified team nursing method of assign- p;ants that can be effectively caredfor by a ment, was the principal focal point onthe ward. rep, ...erect nurse given adequatenursing assistance In addition, many of the activities of the nursing mid supply services. A standardized portable sup- staff related more to the organization of the ward ply and communication station is centered in each and the method of assignment used than to the unit and is stocked with patients' charts,drugs, needs of the patients. In summary, the patient trays, and a phone intercommunication system. was not the principal focus; theattentions of the Each unit is situated close to a group of patients, health care team were not centered around his making itpossible for the nurse's attention to needs or his care. remain on the patients.

Guidelines for Future Research

Our personal endeavors have frequently cen- complish much beyond theirpresent value of tered around a vision of a system of patient clas- internal information sources.First,classification sification accompanied by some form of workload tools that include the psychosocial needs of the index applicable to the vast majority ofgeneral patients frequently do so at the expense of forcing or acute care hospitals. Wehave been influenced subjective opinions to enter into the assessments. by their potential impact not only on the internal This compounds the problems of reliability among operations of a hospital but also on the entire those using the tools. Even when tools aredesigned health care system. We have envisioned a classifi- to be completely objective assessments, rater re- cation system based on patients' needs accompan- liability should not be considered inherent. Second, ied by an index of the workload required to meet the problems of validating the tools have notbeen those needs that %%odd be applicable toall hos- satisfactorilyreconciled, and the attempts that pitals. have been made are not necessarily transferable It would appear that at the presenttime the tootherinstitutions.Finally, workloadindices developed tools fall short of our aspirations.As have primarily been developed through investi- internal tools, particularly within theinstitutions gations of present operating systems and again where the basic research was conducted, theyhave are not necessarily transferable. a great deal to offerin terms of understanding the patient care process. Our experiencehas been, Study Model as previously mentioned,that patient classification systems alone provide a much morereliable As we approach a path for further research, I account of the nursing careneeds of patients than believe our greatest commitment should beto the usual patient census or diagnostic category approach it within the realm of the health care estimate. system. This singlefact alone should urge us It is my impression, however, that in termsof to identify a model to permit a total systems ap- our present data base,their application as tools proach. Figure 4 is one such model. It identifies by which comparisons between hospitals maybe nursing as an interdependent variable in the health modali- made is limited. Differences in treatment care system. ties,physical structure and design, educational The components are easily identifiedin terms commitments, and policies of out hospitals are system being studied. Patients maybe just a few of the reasons why difficulties are en- ofthe viewed in terms of care given, attitudes, expecta- countered. I would suggest that there are atleast three tions,orposthospitalexperience,etc.Stafi reasons why our toolshave not been able to ac- may be viewed in terms ofvolume, competence or

GO BEST COPY AVAILABLE 52 RESEARCH ON NURSE STAFFINGIN HOSPITALS

Figure 4.Nursing:an interdependent variable in the healthcare system.

:ENO

Patients Staff

Internal External I-- Linkages JILinkages

Adapted -by Shirley M. Stinson, Procevsor, School of Nursingand Division of Health Services Administration, University ofAlberta, Edmonton, Alberta,September 1972. Revised April (mimeographed) from 1972 a Paradigm, "System Properties of EducationalUnits," designed by SLOAN R. WAYLAND, Professor of Sociology and Education, TeachersCollege, Columbia University, York 1966. New

61 MEASUREMENT OF PATIENT'S REQUIREMENTs FOR NURSING SERVICES 53 attitudes, etc. Internal linkage., may refer to a spe- for alterations becomes visible and the model can cific department or nursing station and external act as a manual simulation tool. linkages may refer to a coordinated home care We arc still faced with very serious economic program, a communication system, or a visiting problems. Undoubtedly, much of our future re- student. In the same way, all components may be search will be directed toward achieving an op- described and expanded to relate to the area of erational solutiontotheeffective matching of scale. health care resources to health care needs. We This approach, although not new to health care must not, however, neglect to conduct our research evaluation, provides its with at least three distinct with equal attention to economics: the effective advantages. First, it prevents us from omitting the matching of research resources to research needs. patient as a central focus. Second, it provides the I propose that if we conduct our studies within framework for the difficult task of defining vari- the realm of a model such as presented, we will ables. Third, it assists us not only to recognize but have conducted it in an efficient manner and its to study the relationship of interdependent vari- application will have the potential for implemen- ables. Once the system is defined, the framework tation in a wider variety of settings.

62 54 RESEARCH ON NURSE. STAFFING INnoserrms Appendix I

Definition of Categories

MINIMAL CARE of nursing care, medical support, and use of spe- Patientis physically capable of caring for him- cial equipment, e.g., self but requires mininol support plus treatments after acute phase of CVA (residual paralysis), or monitoring (B P., T.P.R., observation, etc.) by advanced Parkinson's syndrome, radical mastectomy of cholecystectomy, the nurse q411 nless. 2nd day post-op, AVERAGE CARE a debilitated, dependent elderly ,erson. Patient requires an average amount of nursing INTENSE CARE care and medical support, e.g.. past the acute stage of his disease or surgery, Patientrequiresveryfrequentto continuous 3-4 clays post-op cholecystectomy, intensive nursing care and close supervision by diabetic for reassessment, medical personnel with support from technical equipment, e.g., conditionsrequiringextensiveinvestigative procedures. multiple sclerosis on a rocking bed, sevf e burns, ABOVE AVERAGE CARE 1st clay retropubic with continuous irration, comatose patient, Patient requires a greater than average amount severe toxemia of pregnancy.

Appendix II

Guidelines to Assist in Selecting the Correct Determiners in the Classification System

PERSONAL CARE coaching. Patient may only require preparation Complete Bath of the bathtub and assistance in and out of the bathtub. Patientisdependent uponthenursefor his complete bath.It could be that the bed Basin or Tub bath is given by the nurse or that the patient Patient takes bath independently; back may requires continual coaching by the nurse during be washed by nurse. the bath, as in rehabilitation. It could also he a tub bath or sling bathif the patient requires NUTRITION contintring assistance or if more than one per- Fed or N.P.O. sonisneededforassistance;e.g., burn bath. Fedtotal meal fed by nurse and continuous One must consider age, general condition, tubes, supervision during the meal. appliances, etc., which inhibit the patient's abil- N.P.O. -- impliestubefeedings,gastrostomy ity to be independent. feeding, etc. Exclude diapostit procedures. Basin or Tub with Assistance Partial Help Patient requires assistance with washing back Meal tray set up by nurse, followed by partial and legs. Patient needs bath equipment set up, assistance but not continuous supervision or fre- partial assistance and periodic supervision and quent encouragement and teaching.

63 MEASUREMENT OF PATIENT'S REQUIREMENTS FOR NURSING SERVICES 55

OBSERVATiON the above while in bed but is also lifted out of Note: One must recognize the difference between bed and positioned in a chair BID or more frequently; intensive medical therapy and the patient's need e.g.,quadriplegic, chronicallyde- bilitated dependent patient. for intense care when interpreting this component. Thiscomponentshoiddbeinterpretedvery Bed Rest with B.R.P. or up with Assistance. broadly. Patient who requires bed rest but changes It implies observation of the patient's ,,ate of his own position as necessary in bed. This pa- well being and psychological support to patients, tient may require assistance in getting out of relatives, and friends. Include the patient's need bed as well as to and from the B.R. Patients for explanation and teaching; e.g., new mothers, may require assistance with activity because of colostomy, diabetic. Consider procedural activi- the presence of Levine tubes, I.V.'s,catheters, tiesin which nursing observationis the basic etc., or because of their general condition. component; e.g.,vital signs, dressings, I.V.in- Up and About fusions, deep breathing and coughing, medica- Patient who does not require assistance with tions, compresses, intake and output, care of ambulation. He is up as tolerated or independent drainage tubes. with the wheelchair. He has learned to manage Consider diagnostic tests;e.g., gastric analy- nonrestricting tubes such as catheters, T-tubes, sis,lumbar puncture, Hollander test,24-hour orI.V.'s. A rehabilitated paraplegic could be urine collection. considered in this group. It could include a confused patient who is up Note: The two following determiners are above in chair or up walking and requires surveillance a..crage or intense care determiners because they at regular intervals. The frequency of the obser- include psychological support and teaching as well vation will guide you in selecting the appropri- as the procedural activities involved. ate determiner: Incontinence CONSTANT to QIHAbove Average or In- Patient who does not have voluntary control tense. over the excretion of bodily wastes; e.g.,feces, Q2H to Q4HAverage urine fistula, or sinus drainage or patients with Q4H or lessMinimal colostomies, ileostomies or ureterostomies Pre-op ACTIVITY This column is checked the morning of cur- In Bed or Chair with Position and Support goy forallpatients going tothe Operating Bed patient who requires exercise, position. Room. Also,itincludes any diagnostic proce- ing, and support with pillows or sandbags at dures requiring extensive workup, preparation leastQ211;e.g.,ahelplesspatientfollowing ofthepatient, and frequent observation and C.V.N. or an immediatelypost-operativepa- followup; e.g., cardiac catheterization, renal ar- tient. Consider as well the patient who requires teriogram, pneumogram.

64 56 RESEARCH ON NURSE STAFFING IN HOSPITALS

References

(1)MacDonell, J.A.K., and Murray, G.B. "An 1970)6-17:19 (MarchApril1970)151- Index of Care." Medical Services Journal 162: 19 (MayJune 1970) 239-252. (Canada), ReprintXXI:8:499-517 Sep- (10) Blom, David 0. A StudyofaPatient tember 1965. Classification System as a Measure of the (2) MacDonell, J.A.K.,etal.Timing Studies Utilization of Selected Supportive Services. of Nursing Care in Relation to Categories M.H.A. Thesis, University of Minnesota, of Hospital Patients. Deer Lodge Hospital. Vancouver General Hospital, Unpublished Winnipeg, Manitoba, 1969. March 1972. (3)Giovannetti, P., et al. The Reliability and (11) Sjoberg, K. "Unit AssignmentA New Con- ValidityTesting of a Subjective Patient cept." Canadian Nurse, July 1968. ClassificationSystem.Vancouver General (12) Sjoberg, K., andBicknell,P.Nursing Hospital,Vancouver,BritishColumbia, StudyPhase IIA Pilot Study to Imp le- June 1970. ment and Evaluate the Unit Assignment (4) Connor, RobertJ. A Hospital Inpatient System. Hospital Systems Study Group, Un- Classification System. Johns Hopkins Uni- iversity of Saskatchewan, Saskatoon, Saskatch- versity School of Engineering. Unpublished ewan, December 1969. Doctoral Dissertation, 1960. (13) .Nutting StudyPhase 111 The (5)Commission for Administrative Servicesin Assessment of Unit Assignment in a Multi- Hospitals (California). Staff Utilization and WardSetting. HospitalSystemsStudy Control Program. Nursing Service Orienta- Group, University of Saskatchewan, Saska- tion ReportMedical/Surgical 1967, 32 pp. toon, Saskatchewan, August 1971. (6) Holm lund,B.A.Nursing StudyPhaseI. University Hospital, Hospital Systems Study (14) __.The Development of a Research Tool to Evaluate and Compare the Stand- Group, Universityof Saskatchewan,Sas- katoon, Saskatchewan, September 1967. ard of Patient Care. Hospital Systems Study Group, University of Saskatchewan, Saska- (7)Sjoberg, K., and Bicknell, P. Patient Clas- toon,Saskatchewan, sificationStudy.HospitalSystemsStudy UnpublishedApril Group, University of Saskatchewan, Saska- 1971. toon. Saskatchewan, September 1968. (15) Sjoberg, K.; Heiren, E.; and Jackson, M.R. "Unit Assignment, A Patient Centered Sys- (8)Bicknell, P. Staffing by Patient Care Work- load. Hospital Systems Study Group, Uni- tem." Nursing Clinics of North America, versity of Saskatchewan, Saskatoon, Sakatch- 6:2 June 1971. ewan, December 1970. (16) Stimson, R.H., and Stimson, D.H. "Opera- (9) Abdcllah, Faye G. "Overview of Nursing tions Research and the Nurse Staffing Prob- Research1955-1968, PARTS I,II,111." lem."HospitalAdministration,17:61-69 Nursing Research,19 (JanuaryFebruary Winter 1972.

65 DISCUSSION OF MRS. GIOVANNETTI'S PAPER

Discussion Leader Dr. Harvey Wolfe Professor, School of Engineering University of Pittsburgh Pittsburgh, Pennsylvania

Dr. Wolfe

I have been trying to figure out a 1,ay to or- theeffectofclassification on workload. That's ganize my comments and I have come to the con- what we really have to validate, not whether the clusion that thereis no possible way to do it. classifications are valid and reliable or whatever So I'll just make. some general comments and crit- else you like but whether the whole concept of icisms relative to the staffing situation as discussed. attaching classification to workload is valid. First, I disagree that this kind of classification We can assume that nurses are humans and have methodology is not transferable. I was not at all human capacities, and we have all observed that surprised to find that three categories were finally in a nursing unit the ability of the nurses to ac- decided upon. i have seen this happen over and commodate to the workload is terrific. They ant over again. The VA, in investigating its four cate- take shortcuts. They can skip steps. They do what gories, found the last two classifications indiscern- they must to perfurm the minimum tasks, and if ible. I think four classifications were found in one they can do more tasks, fine, but if they can't they situation because many hospitals have intensive at least get the minimum tasks done. care units and we, to some extent, are combining Given the fact that nurses adapt to heavier or progressive patient care and controlled variable lower workloaus, what we need to do is to evaluate staffing to take care of the four categories that whether the minutes of care given are a real vali- probably do exist rather than the three. dation of workload. Real validation requires ex- I am not sure why we are still doing studies on tensive study of how the number of tasks and the classification.It appears that no one wiFties to quality, and by "quality" I mean comprehensive- believe anybody else's results without some kind ness of tasks performed, for the whole nursing of reevaluation for their own institution. I don't unit varies as these workload indices vary. That really think hospitals are as different as we claim is what we have to demonstrate. Patient satisfac- theyare. They have many commonalities and tion and dissatisfaction are not the only measures, many similarities that can be specified if we take but they are certainly measures we can use to de- a very close look. termine whether any change in these factor:, takes I think this lack of acceptance results from the place relative to workload. factthat we nursing people are not willing to My second comment is that the expectation we accept the validity related to patient classification, have of staffing studies is too great. We typically nor have we done any studies to really validate get estimates for times and so forth that are much

57 58 RESEARCH ON NURSE STAFFING IN HOSPITALS more precise than we have any need for. We only If blue dots on the forehead were enough to dis- have the ability to manipulate staff in terms of criminate various patients thenthat'sall we'd whole people or sometimes half peopleinrare need to use, nothing more complex than that. cases, 2 hours' worth of personand the fact that Finally,classificationisonly a guide andit a workload differs by a couple of points within should he used as such. Nothing more should be relatively wide ranges doesn't makeany difference gleaned from it than itis capable of giving. One at all. of the things we haven't discussed is the What we'd like to do is to have use of a classification the case mix, and this is an important considera- scheme thatisreliable within these limits, and that's why I think that tie subjective classification tion in any staffing studycase mix and staff mix. It scheme just discussed is perfectly reliable. If itis doesn't always make senseto be perfectly off by one category, that wouldn't botherme. I efficient, and we only have to be efficient within would considerthereliabilitycriteriontobe relatively wide spectrums. We need aproper bal- whether it would change the number of personnel ance of the specialization and the economics as- you need, not whether itcorrelates withitself. sociated with it. Nursing Staffing Patterns and Hospital Unit Design: An Experimental Analysis

Dr. E. Gartly Jaco P1 of essor, Department of Sociology University of California Riverside, California

Intvoduction

This isa report of part of a larger study (1) the writings of Hall (8), Sommer (9), Barker (10), conducted to evaluate potential differences in pa- and Osmond (11).Conceptualizations of macro- tient care in two basic types of hospital nursing spare deal with the larger physical environment unit design; namely, radial and angular shaped in such terms as "territory," "community," "local- units. Very little research has been conducted, par- ity," "natural area," "ecological area," and have ticularly columned experimental studies, on nurse traditionally been essential aspects of the sullfield of staffing patterns. Espec:aily neglected has been the human ecology. Behavioral scientists have also been effect of physical design and spatial geometry of interested in human ritilizatinn of m.icrospae,-, inhospital nursing units on patient care (2). While cluding perceptions of and reactions to such smaller, several studies have investigated the potential im- more immediatephysicalsurroundingsofin- pact on nursing care of a few differently shaped dividuals, presenting such useful concepts as Hall's nursing units their relationship to specific nurse "proxemics," (12)Osmond's"sociopetalspace" staffing patterns has been ignored (3,1,5). and"sociofugalspace," (11) and Barker's"un- Man's use of space has long been a topic of dermanned" and "overmanned" behavinralset- scientificintezest,particularly in the social and tings (10). behavioral sciences (6,7). These and more recent Thus a rich and varied literature dealing with studies and conceptualizations have provideda many facets oft physical environment, useful theoretical as well as a conceptual foundation upon forabetter unuerstanding of therelat:onof which furthe- and more sophisticated researchon physical design of the hospitallo nursing and the potential linkages between human uehavior patient care, prevails in the social and behavioral .'nd physical settings can be based, particularly sciences. Despite their availability, little use has

59 60 RESEARCH ON NURSE STAFFING IN HOSPITALS

been made of these theoretical and empirical works with 145 running feet. The nurses' station, 70 feet in extant studies of hospital design and patient From the farthest bed, totaled 5.3 pe;cent of total care and particularly with respect to nurse staffing area (see figure1). Both units were of new con- patterns. Indeed,the author's own research on struction, similar in interior decor, air conditioned, these topics did not benefit from these theoretical and in close proximity to each other on the third conceptualizations untilafter such studies were floor of the facility. under way. Patient care at the three levels of intensive, intermediate, and self-care were studied in the total The Research Project project. However, only intermediate care will be reported here, since more variations in nurse staff - Nurse staffing patterns were a major factor in a ing patterns were observed at this level of care large research project on nursing and patient care than at other levels. Also itis the more typical incircular and angular hospital nursing units, level of patient care provided in such hospitals. since the types of nursing personnel could affect An effort was deliberately made to keep the kinds patient care on a nursing unit of a given shape. of patients reasonably similar in both study units The major independent or "causal" variables of during the course of the project. Admission to both the study were the radial and angular shaped units was restricted to general medical and surgical nursing units. The dependent or outcome vari- patients needing intermediate care, who were not ables were major types of nursing care, level and employees or relatives of hospital employees, had amount of nursing care, and nurse utilization of never been admitted to a radial unit, and were not emergencies. the unit.Nursestaffing was a key intervening variable alongvith occupancy levels and certain To control the intake of patient characteristics characteristics of the patients in the unit. These during the study an unusual admitting procedure include primary diagnosis, room accommodation, was set up in the Admitting Office. A pair of dice and socioeconomic and demographic traits. was used to select patients on a random basis. This The overall crudy was conducted in a voluntary, was afterallthe previously mentioned criteria acute. nonp--, 350-bed general hospitalinSt. had been applied. (We then, you might say, had Paul, Minnesota. In the early 1950's this hospital amultitude ofpatientcharacteristics,primary construe:cc! a number of circular or radial nursing diagnosis being one. We were concerned with the units as well as the traditional angular or L-shaped age, sex, and social class of the patient, and we units.It had and still has a progressive patient tried to hold the number of factors reasonably care program, including intensive care, intermedi- constant by randomizing between thci two units ate care, and self-care or minimal care. The pres- during the study session. So we made a deliberate ence of both radial and angular units afforded effort to make as similar as possible 'the patient . anopportunity to do a comparative analysis of populations being given nursing care during the patient care given on the two types or shapes research time.) of units. Both were almost new when the study Twelve study situations were set up in the radial was performed and were on the same floor of the and angular study units. Occupancy levels were hospital. altered from high (90 percent filled or more) to The nursing units for the study comprised 22. low (50 percent filled or less). Nurse staffing pat- beds: two private or singe rooms, six semiprivate terns were varied quantitatively from high through or 2-bed rooms, and two 4-bed wards. The radial average to low. They were variedqualitatively unit had atotal area of 5,180 square feet with by varying the ratios of registered nurses (RN), corridors of 900 square feet with 133 running feet. licensed practical nurses (LPN), and nurse aides Its nurses' station in the center of the unit was (NA) to total staff. Our rationale for the original 18 feet from the farthest patient bed and com- staffing mix was based on consultation with the prised 8.2 pe.rent of the total area The angular Nursing Department of the hospital. unit, resembling an L shape with its nurses' station Each study situation was conducted for 4 con- located at the vortex of corridors. totaled 5,910 secutive weeks during a 5-day workweek (Monday square feet. It had corridors of 1,258 square feet through Friday), excluding holidays. Work sam- Figure 1.Radial and angular hospital units. RADIALSTUDY UNIT BEST COPY AVAILABLE ^) r At ,,c% ulROOM 2W c, w w.....s pt.: cl.use liP imsTASES41) 3300 i') ^I ;,,,...0.24050EnD 04 3217 ROOM2 BED v) ..1% ROOM3 0 .3 .3 42,,, q--- SOON3ri BED SUN DECK 3216 3303 wo.JAN EXAM C'cl 13011 CONF ROOMDAY s ANGULAR STUDY UNIT 4320011 6 elf 3222 "c UTILITY LOBBY 3016 15 4 3522BED RM 1352412 BED if ,3 5 26 4 BED RM 4 BED RM In .ROOM 2 BED3206 -7 ROOm4 BEO 3203..m 30013::: N.1). 3014 PRI MSG 3528 3530 ELEV 3207 35183520PRI Ed 3503-T3504 13=1 3500uTILSTA 3501cola 3540 3538 PRI PRI 3534 PRI 3516 PRI T 05021 T'. 3506 PRIPRI P-35123514 PRI .E. 3508 Saint Lukes Hospital THIRD FLOOR 110..4 SE Ake Station 3500 rn 62 BEST COPYAVAILABLE RESEARCH ON NURSE STAFFING /N HOSPITAUS

piing methods were used to observe the total nurs- made to determine what might be the best or most ing staff for 24 hours each of these days. At ran- appropriate staffing pattern for these hospital units domly assigned subareas within the units, they were or patients since such would comprise an entirely observed at intervals of about 10 to 15 minutes each separate inquiry in. itself. Rather, a major purpose hour by observors trained in work sampling tech was to determine whether the basic physical design niques and in nursing tasks and functions. The of the nursing unit had any effect on nursing care. average number of observations was about 9,000 The effects of different nurse staffing patterns, de- for each study situation. The total observatets liberately varied both quantitatively and quali- for the 12 situations came to nearly 205,000. tatively, were also measured, but other variables Certain variables were recorded by an additional such as occupancy levels and divergent patient observor on the unit during the day shift only. characteristics were held reasonably constant or A 1 week resettling period was set up before each randomized. new study situation to help blur any experimental This project was initiated because the hospital "halo" effects. An effort was made to use identical administration wanted an evaluation of their radial nursing personnel in both types of units forall units. They were quite new. Controlled experi- study situations, so that each nurse would be her own experimental controlforindividual differ- mental studies were,if not limited, not readily ences among the nursing staff. This was not always available. So the study was set up to evaluate what achieved, because of staff turnover. No attempt was was going on in these new units,

The nurse staffing patterns were established as follows:

a.Quantitative Shift RN LPN Nurse AideWard Clerk High nurse staffing 7 3 4 (1 Head N.) 2 2 (above average) 3 11 2 2 2

11-7 1 1 1

Average nurse 7 3 3 (1 Head N.) 1 1 1 staffing 3 11 1 1 2 11-7 1 1

Low nurse staffing 7 3 2 (1 Head N.) 1 1 1 (below average) 3 11 1 2 1/2 11-7 1/2 1/2

b.Qualitative 7 3 5 (1 Head N.) 0 0 1 High RN staffing 3 11 3 11-7 2

High LPN staffing 7 3 1 Head Nurse 5 0 1 3 11 1 Head Nurse 3

11-7 1/2 2

High nurse aide 7 3 1 Head Nurse 0 5 1 staffing 3 11 1 Head Nurse 3 11-7 1/2 2

We feltthat to do the job right we had to in some manner experimentally during our study. control a number of factors that everyone in the So an effort was made in this direction. We did not profession is aware could affect or contaminate the agreeto conductthe studyinthisparticular outcome of the study unless they were controlled facility until everyone agreed to accept any nega- NURSING STAFFING PATTERNS AND HOSPITAL. UNIT DESIGN: AN EXPERIMENTAL ANALYSIS 63 tive evidence that might be generated by the study, what the nurse was doing. (When they were answer- (So many inhouse studies are done to rationalize ing the telephone we wanted to know whether what the administration already wants to do or has they were talking to the doctor or a relative of the a tendencyto be nervous about accepting evi- patient or whom they were talking to, and things dence against.) One of the hospital trustees said, of that sort helped make the eventual classifica- "Well, if we make a mistake in these units, the tion of their activities alittle more, in fact ex- whole field should know about it so they won't ceedingly, reliable.) repeat the same mistake." We also wanted to sustain a nursing staffing The .same went for the nursing staff. This was situation for a period of time should the nurses difficult because we wanted to set up a number put on a show for the observer, (which they can of conditions that the research design required to do, you know). If this occurred,' we would extend be sustained over a suffir!ent period of time to the period of observation aftertheir show was conduct a really valid project from the research over. This was somewhat of a problem when we standpoint. So a second factor essential tothis got into some of the more difficult staffing pattern type of study is a willingness on the part of the situations. staff to conform reasonably well to the dictates Another thing we wanted to check out was the of the research design.This isthe more difficult significance of the activities of the nursing staff problem, particularly when you're experimenting on the night shiftif such were equal to,if not with the number of variables we were trying to more important than, what went on in the day do in our study. shift in terms of the patients' morale, reactions, We began with what was an approximation of and satisfactions for th it nursing care. So much the existing average staffing pattern already going research has been oriented toward the day shift on in the hospital and used that as our baseline when actually other times may be equally signifi to increase it quantitatively or decrease it in just cant. Just because that's when the doctor is usually sheer numbers of nursing personnel and then vary there doesn't mean that other things aren't hap- it qualitatively between the categories of nursing pening at other times of the day. So we wanted personnel. When you talk about understaffing and to check out these possibilities. overstaffing you should say whom youdare under- Then, too, we were quite concerned about the staffing and whom you are overstaffing. Isit the statistical validity of our observations, not only registered nurses, the LPN's, the nurse aides, or in terms of reliability and validity but also to de- who? This turned out to be a very important termine whether we really were getting a good outcome variable, the quantitative variation versus picture of what was going on in these units. So we qualitative, the mix of nursing category personnel. divided the unit into 10 subareas: the nurses' sta- So we proceeded to set up our project in which tion, medication area, patients' rooms, the corridor, we would have observers on the units routinely utility room (they had a tub and shower room to making observations of the activities of the nursing supplement the bathrooms in each of the patients' staff for a 5-day workweek, Monday through Fri- rooms) and allthe different parts of the unit. day, and observe them on all three shifts, 24 hours We random! ted these areas in which the observa- a day, in the unit. We did a good deal of training tions would be conducted by the observers so that of the observers. We used LPN's primarily for the nursing staff hardly knew when they were observers because one thing we found out quickly being observed. Although the observers may be is that RN's don't like to be observed by other in the unit, they may not really be recording at RN's, but they are more tolerant of thelesser that particular place in the unit, and this method trained staff looking at them. Other studies have enabled us to get a random sample. had the same problem. We found some feeling of resentment on the part After the training session, we checked out the of the aides and the LPN's toward the RN's and reliability of coding their observations by a num- the head or charge nurse in the radial unit when ber of coders and observers. We obtained very the latter supervised their activities too closely. high check/recheck reliability of about 95 per- If you were to see this unit you could see why. cent because the recordings were quite detailed on You can be in a patient's room and you can see RESEARCH ON NURSE STAFFING IN HOSPITALS into all the other patients' rooms, too. soif the emergency service, whether or not it has an out - head nurse would be attending a patient in one patientservice, whethe ittakes a broader range room she could stillwatch the other personnel of patients or is confined to a few. In effect, you onthe sameunit. Shecan'tdothatinthe must consider the inflow of patients the nursing angular unit. Such visual contact or visibility of staff must care for. activity on the radial unit did not exist on the We were interested in some of the more tradi- angular unit. and this IHIHPd out10 be a key tional outcome factors such as the potential re- variable in the whole study that was related 10 the lationshipofnursestaffingpatternsandoc- shape of the unit. cupancy levels to length of patient stay. We used Also, because of differential preferences by the Dr. Aydelotte's scales to see whether patient wel- nursing staff for the different units in the hospital, fare would be in any' way affected. We checked we tried to equalize these preferences by choosing the possible effects of basic demographic factors, the nursing personnel equally between those who medical diagnoses, amid a number of basic medical preferred the radial to the angular unit and those behavioral factors on the dependent variables. who preferredthe angular totheradialunit. As itturned out, the length of stay did not So we tried to get an equal mix so that any bias vary significantly among any of the so-called study that might affect how they attend patients and situations. The patient welfare measures and a their activities would be equalized. number of other variables did not affect our results. Also, we made a very strong effort to use the Prior to discharge, each patient was interviewed same nursing personnel all throughout the 2 -year privately off the floor regarding the nursing and project. Instead of trying to get so-called matched medical care he had received, how he liked or groups of nurses, one in one type of unit, another disliked his room, the layout of the unit and so on. in another type of unit, each nurse was her own We soon discovered a bias operating in favor of experimental control. They wound run through the radial unit. The radial unit was rather small, the series of study situations in the radial unit, and we could quickly see a bond develop between then he transferred into the control angular unit the nursing personnel and the patients that we and repeatthestaffingpattern and occupancy did not see at the same intensity on the angular situations. We only had a certain amount of turn- unit. over in the nurses over a 2-year period. About 90 Fortunately theinterviewer, asocialworker, percent of the nursing personnel were the same wasveryskilledatinterviewing and worked throughout the project in our studies, which we through the fact that many patients would simply felt was important. not be critical of the nurses, more on the radial Because we knew the study would lastfora unit than on the angular unit,atthe start of long period of time, we were quite sensitive to the interview. finally, we did get some criticisms the cyclicity of the occupancy level of the hospital. after piercing through the facade of being a "nice Most hospitals do follow peak and trough periods guy who didn't want to embarrass the nurses," and of occupancy. Summer time in this hospital was so forth. its low patient occupancy period and was vacation But this was a reflection of the close relation- time for the staff. We were sensitive to the vacation ship that developed in the circular unit. This was needs of the personnel of the staff so we set up particularly true in those study situations of high our low staffing patterns and our low occupancy occupancy and low staffing. levels when the hospital would be ordinarily at Another factis that the patients on the radial low occupancy and nursing personnel would likely unit could identify the different nurses by the caps be on vacation.Ithink such sensitivityisvery and uniforms of personnel coming in and out of often important to the success or failure of such the unit, which they couldn't do on the angular projects in hospitals of this kind. unit. There were other indices of how patients When you study staffing patterns, you have to empathised with the nurses more on the radial take into account the role and function of the than on the angular unit. It was not the same for facility. the kinds of patients likely to he admitted the doctors, I might add. by the medicaltaff, whether ornotthereis One of our main purposes in the project was to

14,i`if 3 BEST COPYAVAILABLE

NURSING STAFFING PATTERNS AND HosPITAL UNIT nEsioN: AN EXPERIMENTAL.ANALYSIS 65 replicate many factors in a prior study conducted The categories of the direct care activity involv- by Sturdavant and her associates (1) at Rochester, ing treatments, where theactivity was oriented Minnesota. We wanted to see what factors we could toward the medical problem of the patient, were replicate so we used at our facility a number of termed medical nursing care. Care that was geared criteria similar to those in the Rochester study. toward physical comfort like adjusting the pillow, Thisisone thing needed in nursing care re- making the patient more comfortable, attending searchmore replication in different kinds of facili- to physical needs of patients not related to their ties with different personnel to see if we derive illness was called physical (bodily) care. The third the same or divergent results. So we used essentially was social, which involved communicating, nurses the criteria for patient and nursing care that were talking with the patient or observing the patient used at Rochester, with some variations. directly in case there was a need in terms of the For example, we had six basic categories of major social category. types of nursing care essentially defined as direct We had another category termed patient-involved bedside care, which was basically composed of care. That is, such care was definedessent:ally as treatments, supportivecare, ambulation and so a combination of direct care, indirect care,and on and communication directly with patients. Then general assistanceall of which had directly to do we had indirect care, generalassistance, standby, with involving the patientas opposedto such time out, and travelfor patientpurposes.The factors as standby and time out. So we had ratios criteria used were similar to the Rochester meas- of time observed in these three activities compared ures to permit replication of 'these factors. to other activities, and this varied enormously. .We took these criteria and categorized them into These arc just some of the measures we developed what we call medical, social, and physical care. in the course of the project.

Major Results

Type of Nursing Care

Since a major purpose of the overall original more indiiect care, general assistance,and took study was to replicate many facets of Sturdavant's more time out, andeririiidedlessdirectcare, Rochester study (13), similar criteria and measure- standby, and traveron the radial units than on the ments of many variables were used in the present angular unit. inquiry. Six major categories of nursing care were An analysis of the six specific nurse staffing pat- defined: direct (bedside) care, indirect care, gen- terns designed for this study presented a somewhat eral assistance, standby, time out, and travel. For different picture when shape of the unit and oc- the total of 12 previously mentioned study situa- cupancy levels were pooled, as presented in table2. tionsit was found that major types of nursing For the quantitativestaffingpatterns,the high exhibited more time spentin caredifferedsignificantly 1 tweenthenursing staffingpattern units, as shown in table 1. (Percentage differences greater than 1 percent were statisticallysignificant, Table 1.Major type of nursing care in an acute, because of the large size of the samples involved.) general hospital It was found also that the highest percentage Numberof of nursing care observed was in direct, bedside TypeofCare observationsPercent Rank order by standby, care (30 percent), followed in Directcare 61,338 30 1st indirectcare,generalassistance, time out, and Indirect care 46,370 23 3rd lastly intravel, when nurse staffing pattern, oc- General assistance 19,567 9 4th cupancy level and other characteristics wereheld Standby 46,601 24 2nd constant statistically. On 204,793 observations on Time out 17,405 8 5th 11,512 6 6th both nursing units for somewhat over a year each, Travel Total 204,793 100 approximatelyidenticalnursingstaffsprovided 74 66 BEST COPY AVAILABLE RESEARCH ON NURSE STAFFING IN HOSPITALS

Table 2.-Major type of nursing care, by six nurse staffing patterns

Nurse staffiing pattern Type of care High Average low High RN High LPN High NA NumberPer. Number Per- Number ,"er NumberPer- Number Per. Number Per. rent rent rent cent cent cent Direct care 12,261 26 10,821 32 14.456 32 7,988 35 10,632 31 11,180 28 Indirect care 9,849 21 7.832 23 t.161 23 6,170 27 9,240 27 7,118 18 General auistance 4,743 10 2,801 9 2,638 10 2,631 11 3,627 10 3.127 8 Standby 13,088 27 7.075 21 5,832 22 3.686 16 6,424 19 12,496 32 Time out 4,764 10 2,980 9 1,999 8 1,582 7 2,770 8 3,310 8 Travel __ _ _ 2,824 6 2,114 6 1,335 5 961 4 1,879 5 2,399 6

Total _ . 47,529 100 33,623 100 26,421 100 23,018 100 34,572 100 39,630 100

Table 3.-Major type of nursing care, by nurse staffing patterns and by radial and angular units, percentages only

Major type of care (percent of observations only) Staffing Direct Indirect Gen. Asst. Standby Time out Travel pattern RadialAngu RadialAnguRadialAnguRadialAnguRadialAngu RadialAngu lar lar lar lar lar lar

High 26 26 22 20 13 7 24 31 10 10 5 6 Average 31 34 25 21 9 8 20 22 9 9 6 6 Low 33 32 24 22 10 10 19 25 8 7 6 4 High RN . 35 34 28 25 13 10 14 18 7 7 3 6 High LPN______29 33 27 27 11 10 20 17 8 8 5 5 High NA 27 29 19 17 8 8 31 32 9 8 6 6 standby followed in order by direct care, indirect staffing patterns, although type of care was affected care, equal amounts of general assistance and time by all of the staffing patterns differently when oc- out, and last in travel. There was an increase in cupancy level and shape of the unit were pooled standby and time out and a decrease in direct or held constant. and indirect care compared to the percentages of Were these major types of care for each of the time observed giving these types of care for the patterns affected by the shape of the unit? Table total staff. 3 shows that type of care provided on the two Average staffing showed an increase indirect units did indeed differ for the six nurse staffing care and a drop in standby compared to the total patterns.Forthequantitativepatterns,high staff. Low staffing also exhibited an increasein staffing increased time spent on general assistance direct care and less standby than for the average and decreased standby on the radial unit, but the staffing pattern. For qualitative patterns, the high opposite occurred on the angular unit. Average RN pattern showed an increase in direct care, instaffing increased time on indirect care on the cir- direct care, and general assistance but a decrease cular unit and decreased such care while increas- in standby and travel. The high LPN pattern ing direct care on the angular unit. Low staffing exhibited an increase in indirect care and a de- showed opposite results primarily in standby, drop- crease in standby, and the high NA pattern in ping on the radial unit and increasing on the an- dicated an increase in standby and a decrease in gular unit. direct and indirectcare compared to thetotal For the qualitative patterns, increased time oc- staff percentages for such care. curred for general assistance with a drop in standby Thus, major types of care seemed to vary more for the high RN pattern on the circular unit, for the qualitative than for the quantitative nurse while indirect care decreased and standby and NURSING STAFFING PATTERNS AND HOSPITAL. UNIT DESIGN: AN EXPERIMENTAL. ANALYSIS 67 travel increased on the angular unit. Direct care the high levels of quality of nursing care (RN) and dropped on the radial unit, but increased while the least influence on the lowest level of care (NA). standby decreased on the angular unitfor the This was further confirmed when the major types high LPNpattern.Apparently,the high NA of care for specific nursing personnel were analyzed. staffing pattern was not significantly affected by Ilead and charge nurses on the radial unit pro- shape of unit since type of care on both resembled videdmoreindirect care and general assistance closely that of this staffing pattern totally, although andlessstandby than on the angular unit. RN's small differences occurred for type of care between gave more general assistance andlessdirect care the two units. and travel on the radial than on the angular unit. Thus we find that both nurse staffing patterns LPN's on the radial unit gave more indirect care and the physical design of the hospital unit af- and general assistance andlessstandby than on fected the major type of nursing care provided to the angular unit. NA's traveledmoreand gave general medical and surgical patients. Adding ad- lessdirect care on the radial than on the angular ditional nurses to the staffing pattern when shape unit. of the unit was ignored led tomorestandby and In studying the effect of nurse staffing patterns time out andlessdirect patient care. and design of the nursing units, the need to con- This resembles similar findings by Aytlelotte and trol for differences in occupancy levels and stages her associates (3) when compared to the average of illnessisobvious. A separate analysis of the staffing pattern. The low quantitative staffing pat- results was made in terms of high and low levels tern devoted similar percentages of time to direct of occupancy. of the study units. This revealed sig- and indirect care as for the average staffing level. nificant differences in type of care provided for But the qualitative staffing patterns indicated more the staffing patterns and shape of the units,as effective differences in type of patient care, particu- differential demands forpatient care wereaf- larly for the high RN pattern where significant fected by these patient levels. Consequently, our increases in direct and indirect care and general efforts to control for such divergences when analyz- assistance occurred compared to the lesser quality ing the effects of staffing and shape of the unit were staffing ratios. empirically justified. The physical shape of the nursing unit seem- The type of care most affected was direct care. ingly affected the types of care given for most of For example, by and large there was more direct these staffing patterns. Increasing the quantity of care given in the angular unit than in the radial nursing staff had differing results on the two units. unit, althoughit was "easier"to givc itin the For the radial unit an increase in indirect care radial unit. Also to our surprise, there was less and general assistance occurred more than for the social care in the radial than in the angular unit; angular shaped unit, with standby differences dif- thatis,lessverbal communication between pa-

ficult to assess. Varying unit shape led to opposing tients and nurses on this type of unit. . changes intype of care when changes did take So we had to come up with some answers as to place for identical staffing patterns. Indirect care, why this happened. Did the shape of the unit general assistance, and standby were most affected have anything to do with this or what was going by the radial shape, but direct and indirect care, on between the staff and patients or what? The general assistance, and standby were most affected circle is architecturally the most rigid, inflexible by the angular unit although in oppofite ways. shape, and the only way to expand a circular Even so, the radial unit seemingly 1. ad more ef- unit is to add floors vertically like stacking pan- fect on type of nursing care provided by qualita- cakes. It cannot expand laterally as can an angular tivelydifferentthatby quantitatively divergent unit or other shapes. Yet, on most nurse staffing staffing patterns. The high RN pattern gave more patterns we found far more flexibilityin what indirect care and general assistance with less travel careis given by whom in the circular thanin and standby on the circular than on the angular the angular unit.Ithink thatis why the reg- unit. The high LPN staffing pattern gave more istered nurses preferred it, because they were able direct care on the angular than the radial unit. to exercise more options in giving care to patients Shape of the unit, therefore, had more effect on by the vet), fact that they could see the patient 68 RESEARCH ON NURSE STAFFING IN HOSPITALS from the station. They could sit in the station (sit time spent in the patient's room per nurse visit down because the counter is low), and they could was far less on the radial than on the angular unit. look into every room and see whether the patient Why? What happens? You can do a tempera- needed care or not. You can't do that in an angular ture check and you pick up the tray and maybe unit because you don't have that visibility. In the you do some comfort things in one trip in the angular unit, when a call light came on in the angular unit, and those tasks can accumulate into station, the head nurse might delegate an aide, to several functions per visit. A lot of them involve "Go see what Mrs. Jones wants."Butinthe talking and socializing withthepatient, whiCh radial unit she could look up and see whether you don'tsee inthe radialunit because the she was needed, She could exercise options in the nurses don't feel there is as much need for it in delivery of nursing care that she could not do in the radial unit as in the angular unit. This is part a differently shaped unit. of the relationship between patient and nurse. So I think, because of this, there was a pref- We could almost predict the day patients in the erenceforthe radialunit by the RN's. They circular unit would be discharged. One of the couldexercise more professionalnursing judg- things that happens when patients get out of in- ment in providing care. There was a lot of non- tensive care in this type of unit is that as they ait verbal communication going on between patients convalescing,feeling betterafterthecrisishas and nursingstaff waving and smilingateach passed, they start closing the cubicle curtain be- other hut they didn't go into patients' rooms and cause they don't want ,to be bothered. They want visit as much. So there was a kind of vicarious privacy and want to read or watch TV. They care going on between the staff and the patients leave their room, go to the lounge or the coffee in the circular unit. shop, and use other parts of the hospital more Now, we got into some very difficult study situa- while convalescing on the radial unit than they tions where we really overloaded the staff. We do the angular unit. The "curtain pulling syn- were almost risking the well-being of the p;W:.nts, drome" was directly related to convalescence, and but we would terminate the study situation before we recorded the halfway pull, all the way pull, actually jeopardizing the welfare of any patients. and finally when itstarted to be closed allthe During these overloaded situations, many patients, time we knew the doctor was going to discharge the when we interviewed them, said they never touched patient because he was ready to leave. This oc- the call light in a 2- or 3-week stay, and we asked curred only on the circular shaped unit. them why. "Well," the patients would say, "the A number of other events occurred that you nurses were too busy.Ididn't want to bother don'tordinarilytakeforgranted. What goes them." They contr.lsee the nurses in the unit on in the nighttime in different units? This is running around helter skelter sometimes. and they where we had some problems with the radial unit didn't want to bother them. They felt that, "If I because of the architecture. One patient, because need htlp, they'll come," so they had this confidence the drape next to the outside window wouldn't in the nursing staff in the radial unit. quite close, complained about being bothered by We didn't find that in the angular unit. They noises. We found there was street light by her would sometimes get lonely and just want to see if window, and the slightly opened drapes allowed anyone were still out there because they hadn't a sliver of light to come into he;' room all night seen them in a while. So they pushed the call light. long, When we examined the factor of nurse initiated We found these things out when we interviewed visits versus patient initiated visits, we found that the patients.Itis amazing what you can learn about 82 percent of the trips to the patient's room about hospitals from patients, about a lot of little in the radial units was initiated by the nurses things. The night light in the nurses' station both- compared to about 79 percent on the angular ered one room butitdidn't bother any of the unit. Well, you may think this doesn't make sense other rooms on the radial unit because of the since sometimes they may not need to make a shape and room location. Also, the drug cabinet trip to the rooms because they can seein. But was a problem in the radial unit because access a related finding was that the average length of to the radial unit nurses' station was easier. BEST COPY AVAILABLE NURSING STAFFING PATTERNS AND HOSPITAL. UNIT DESIGN: AN EXPERIMENTAL. ANAI.YSIS 69

Now one other thing about our discovery was Medical, Physical, and Social Types of that a number of physicians, in fact every physician Direct Patient Care whom we interviewed, had patients to whom the radial unit was contraindicated, This was not as Direct care was examined interms of three much for medical as for personality reasons, but subcategories. These were medical, measured by none of the physicians had any contraindications the frequency of tests, measurements, IV therapy, against putting patients into the angular units. preoperativecare,medication andprescription This meant that if this hospital went strictly to administering (treatments asa major category); radial units it would lose about 20 percent of its physical, measured by frequency of providing bath- occupancy because the medical staff would admit ing and grooming, comfort, meals, and courtesy that many patients to a facility with other shaped services, elimination and ambulation services to pa- units-notfor intensive care,thatis,but only tients (general supportive care); and social, indi- for intermediate care. cated by direct bedside communicationandob- I should mention too that we were further sur- servation of patients.It was foundthat more prised when we did this same kind of study for physical care was provided, followed by medical self-care patients. The radial unit was designed care and last social care, as presented in table 4. for intensive care so that nurses could constantly Between the two units, no differences occurred in observe patients who need this type of monitoring, medical care, but more physical andless social care not electronically but by eyeball, with minimum occurred in the radial than in the angular unit. effort. When we dropped to the intermediate level For the nurse staffing patterns, the total pattern we had problems. When we dropped to the min- of such care did not differ for the high staffing imal self-care needs we were surprised. The self- pattern. Less social care occurred for the average care unit was farbetter from thestandpoint pattern and less medical and more physical care staff and patientsinthe circular than in the occurred for the low staffingpattern (table 5). For existing angular self-care facility. thequalitative patterns, more medicaland less

Table 4.-Medical, physical, and social types of direct care, by number oE, observationsand percentages, total study, by shape of unit

Type of care Total Radial Angular N,.mber Percent Number Percent Number Percent 11,858 38 Medical _ 23,057 38 11,199 38 14,496 46 Physical . 29.631 48 15,135 51 Social 8,650 14 3,446 11 5,204 16 100 31,558 100 Total - 61,338 100 29,780

Table 5.-Type of direct patient care, by nurse staffing patterns

Nurse staffing prttern Type of care High Average Low RN LPN NA Num Per. Nun, Per Num Per Num- Per. Num Per. Num- Per- ber cent ber cent ber cent ber cent ber cent ber cent Medical 4.594 38 4,232 39 3,000 35 3,238 41 4,220 40 3,773 34 5,014 47 5,465 49 Physical _ 5.628 48 5,264 49 4,363 52 3,896 49 Social - ---. -- 1,739 15 1,324 12 1,093 13 854 10 1,398 13 1,942 17 Total 11,961 100 10,820 100 3,456 100 7,988 100 10,632 100 11,180 100

78 70 BEST COPY AVAILABLE RESEARCH ON NURSE. STAFFING IN HOSPITALS

Table 6. Type of direct patientcare by nurse staffing patterns, by radial and angular units, percentages only

Staffing pattern Type of direct care (percent only) NIethcal Physical Social Radial Angular Radial Angular Radial Angular High 10 37 51 43 9 20 Average 39 39 50 47 11 14 Low :14 37 55 48 11 15 High RN 39 42 51 46 '10 12 High I.PN 39 40 51 44 10 16 High NA 36 32 50 48 14 20 social care occurred for the high RN's,more medi- and general assistance to nonpatient involved ac- cal care for the high LPN's, andmore social care tivity (standby and time out),all derived from and less direct care took place for the high NA the preceding data. As presented in table 7, the pattern when compared to the total staff results ratio of such types of 'care for the total staffwas (table 5). nearly 2 to 1 (1.93) for patient involved to non- When such staffing patterns were analyzed by patient involved care. For the specificnurse staff- the two shaped units, significant divergences again ing patterns, the highest ratio was found for the occurred as shown in table 6. The quantitative high RN pattern(3.2),followed by the high staffing patterns showed increases in medical and LPN (2.6), average staffing(2.1), high and low physicalcare and reduced socialcareathigh staffing (1.5 each), and lastly the high NA pat- staffing on the radial unit andan opposite result tern (1.3), (table 7). on the angular unit. The angular unit caused more Physical design of the nursing unit had a signif- divergences at average staffing levelon the angular icant effect on these ratios of patient involved to than on the circular unit, increasing social and nonpatient involved care for all six staffing pat- reducing physical care. Low quantitative staffing terns as shown in table 8. Such care increased for had opposite resultsfor these forms of care on all three quantitative patterns and. the high NA the two units. Qualitative differences instaffing also had differing resultson the two units. High pattern and decreased for the high RN and high RN staffing exhibited more physical and less medi- I,PN patterns on the radial shaped unit. These cal care on t;:e radial unit andmore social and ratios decreased on the angular unit for high and less physical care on the angular unit. average staffing and high RN and increased for Opposing results also took place for the high the low staffing and high LPN nd high NA pat- LPN pattern on the radial and angularunits. terns on this latter unit. There was more physical and less socialcare on the former than on the latter unit. The high NA Table 7.Ratios of patient involved to nonpatient pattern indicated an increase in medical anda de- involved care, by total staff and by nurse staffing crease in social care on the radial unit and an patterns opposite result on the angular unit(table6). Again, the physical shape of the nursing unit A. Total staff:1)C, IC, GA / Standby, T.O. = 127,275/ 66,006 affected the types of direct patient 1.93 care provided ft by the various nurse staffing Staffing pattern / Nottpt. patterns. Pt. invol. invol. Ratio High = 26,853 / 17,852 = 1.5 Average = 21.454 / 10.055 = 2.1 Patient. Involved Care low = 17.255 / 11,831 = 1.5 Iligh RN = 16.789 /5,268 = 3.2 Patient involved care is measured by the ratio High 1,PN = 23,499 / 9,194 = 2.6 Fl igh NA of direct nedside care combined with indirectcare 21,425 / 15,806 = 1.3

79 BEST COPY AVAILABLE NURSING STAFFING PATTERNS AND HOSPITAL UNIT DESIGN: AN EXPERIMENTALANALYSIS 71

Table 8.Ratios of patient involved to nonpatient that the high NA staffing pattern had the highest involved care, by nurse staffing patterns and average number of total trips perday. This was radial and angular units followed in order by high staffing, RN, average staffing, high LPN, andlastly low staffing, in- Shape of nursingunit dicating a mixed pattern of trip activity (table Staffing pattern Radial Angular 10). However, the ranking shifted somewhat when High 1.8 1.3 the average per diem number of nurse initiated Average 2.2 2.0 tripswas examined. High staffingshowed the I.ow 2.4 2.0 highest average number of such trips followed in High RN ..... 3.6 2.8 order by high NA, average, high LPN, and finally High LPN 2.4 2.8 High NA 1.4 1.4 low staffing. The average number of patient initiated trips was far lower than nurseinitiated trips. Average Other differences in this ratio of care took place staffing had the highest daily average followed in between the two units for these staffing patterns. order by high staffing and high RN (tied), high The radial unit exhibited more patient involved NA, high LPN, and low staffing patterns. The low care than the angular unitforall quantitative staffing pattern was the lowest in such trips, and staffing patterns. As quality of staffing dropped, pa- high NA, high staffing, and average staffing pat- tient involved care also decreased on the ra,lial terns were !iighestin total average trips, nurse unit. initiated trips, and patient initiated trips in that order. The average number of total trips to patient Amount of Nursing Care rooms was more affecod thannurseinitiated trips anpatient initiated trips, as presented in The amount rit nursing care is measured quanti- table 11. Only the high NA pattern had a higher tatively by the number of trips by the nursing arc Age total of trips on theradial unit, and a staff to the patients' rooms in the two study units decrease was noted for the remaining patterns,. and more specifically when such trips are initiated except for average staffing when no change was by the nurse and by the patients' calls. This factor found. On the angular unit the opposite result was observed only during the dayshift when occurred with an increase in the average number such activity is most likely to occur. As illustrated of total trips for high and low staffing and for in table 9, the nursing staff initiated by far the high RN and high LPN. A reduction occurred most trips (79 percent during the total project, for high NA staffing and no difference was found while staff-to-staff trips were the least (1 percent), for the average staffing pattern, which was true with patient initiated trips very low (3 percent). When the average number of total tripsper also on the circular unit for this pattern. day and those that were nurse initiated were com- For nurse initiated trips,fewer but sufficient pared to those initiated by patients in the units differences occurred for the nurse staffing patterns. by the six nurse staffing patterns,it was found There was a decrease in such trips on the radial

Table 9.Types of trips to patients' rooms by nursing staff in radialand angular units, by number of observations and percentages

Types of trips Total Radial unit Angular unit Number Percent Number Percent Number Percent Nurse initiated 118,444 79 60.772 82 57,672 78 2,135 3 Patient initiated _ 4,235 3 2,100 3 Tripsp,. absent 11,843 8 5.619 7 8,224 8 Tripsno pts. involved 13,121 9 4,495 8 8,826 11 1,158 2 Staff to staff 2,409 1 1,251 2 Total __.... 150,052 100 74,237 100 73,813 100

so BEST COPY AVAILABLE 72 RESEARCH ON NURSE STAFFING IN HOSPITALS

Table 10.Average daily number of trips to pa- average ntuW)er of such nips on the radial unit, tints' rooms by nurse staffing patterns, by total and an increase occurredforthishigh staffing trips and by types of trips pattern and a decrease for the high RN staffing pattern on the angular unit. Staffing "fetal Patient pattern trips Untiattil initiated Thus both staffing pattern and shape of unit exhibiteddifferentialeffectsinthe amount of High _ 732 688 Average 665 523 22 nursing care provided in this study.

Low . 589 485 High RN 675 496 21 High LPN 626 497 16 UtilizationoftheUnitby NursingStaff High NA 797 638 Is Since differencesinnursing care were found unit for high and average staffing and for high for all of the preceding variables in terms of nurse RN staffing, auincrease for the high NA pattern, staffing patterns and physical design of the units, and no change for the low and high LPN staffing. itis of special interest to analyze how the nursing On the angulai unit, decreases in such trips oc- personnel utilized these units. Table 12 shows that curredforthe high, average, and low staffing the muses were observed spending the greatest patterns and alsofor the high N pattern. No amount of their time in the nurses' station (31 change took place for the high I.PN pattern. For percent). Following in order were thepatients' patient initiated trips, only the high staffing pat- rooms when the patient was present, nurses' con- tern showed a change. There was a decreasein ference rooms located within the units,offthe

Table 11.Average daily number of trips to patients' rooms by types of trips, bynurse staffing patterns, and by radial and angular units

Type of trip to rooms (daily average) Staffing pat tan Total trip,. Nurse initiated Patient initiated Radial Angular Radial Angular Radial Angular

High . 705 7 i9 617 586 15 14 Average 661 665 5.11 502 20 22 Low 557 621 488 462 15 14

High RN . _ 646 705 487 505 22 19 High LPN 593 659 500 495 17 16 High NA 853 742 692 585 18 19

Table 12.Utilization of nursing unit by total nursing staff in radial and angular units, by numberof observations and percentages

1:nit subarea Total Radial Angular Number Pert, nt Number Percent Number Percent Nurses' station 73,399 34 39.985 37 33,414 31 Mulication arit 0,313 4 4,850 4 4,463 4 Corridor in unit 17,634 8 8,393 8 9.241 9 Patients. rooms 46,608 22 22.002 21 24,606 23

Utility room 2,750 I HIS 1,015 2

Tub and shower room 1.064 657 1 407 Nurses' conterence room 34,574 16 14.010 13 20.564 19 Off unit . 24,380 11 13,068 12 11,312 10 Other areas in unit 5.377 3 3,009 2,368 2 Total 215,099 100 106,789 100 108,310 .....100 I,ess thanIpercent.

81 BEST COPY AVAILABLE

NURSING STAFFING PATTERNS AND nosPrrAt. UNIT DESIGN: AN EXPERIMENTAL. ANALYSIS 73 entireunit, corridors within theunits,medica- For each nurse staffing pattern observed on the tion arca, miscellaneous areasintheunit,pa- two units some inter esting differences were found tients' rooms when the patient was absent, utility in utilization of these subareas in both units, as room, a 10 lastly, in the tub and shower room in shown in table 14. High quantitative staffing led the unit. to an increase in use of the nurses' station and Under conditions of differing nurse staffing pat- in being off the entire floor and a decreaseitu terns, the nursing personnel were found to utilize time spent inpatients' rooms while the patient the two units somewhat difieremly than for the was present and in the nurses' conference room totalstudy, as illustratedintable13. For the on the radial unit. Use of the conference room quantitative staffing patterns, high staffing person- increased on the angular unit and use of the nel used the nurses' conference room and were nurses' station decreased. Average staffing brought off the unit more and spentless time inthe about an increase in use of the nurses' station nurses' station and in patients' rooms when pa- and a drop in time spent in patients' rooms, patients were present Average staffing indicated an tients present, and use of the conference room increase in time in patients' rooms when patients on the cir:ular unit. The opposite occurred on were present and a reduction in time in the con- the angul. r unit. Low staffing resulted in an inference room. Low staffing increased time in pa- crease in time off the floor and a drop in use of tients' rooms only when patient..., were present. the conference room. For the qualitative patterns, high RN staffing In terms of the qualitative staffing patterns, high led to an increase in use of the muses' station, RN staffing led to an increase in time in the medication area, and patients' rooms withpa- nurses' status and patients' rooms with patients tiznts present, and a decrease in use of the con- present and a decline in use of the corridors and ference room and in being off the unit. High nurses' conference room on the radial unit. There LPN staffing led to more time in the nurses' sta- was an increase in.useof the conference room tion and to less use of the conference room. The and a decrease in time spent in the nurses' station high NA staffing pattern led to increased use of and patients' rooms with patients present, on the the conference room and decreased time spent in Angular unit. High LPN staffing on the radial the nurses' station and medication area. unit showed a drop in use of patients' rooms with

Table 13.-Utilization of cursing unit by nurse staffing patterns, by number of observations and percentages

Nurse staffing pattern Unit subarea High Average Low RN LPN NA

Num- Per. Aft1111 Per- A'um- Per. Num- Per Num- Per Num- Per. ber cent ber cet ber cent ber cent ber cent ber cent Nurses' station 16,471 32 12,218 35 9,679 35 8.699 36 14.265 40 12,742 30 Medication area 2,065 4 1.574 4 1,239 4 1.450 6 1,829 5 1,046 2 8 1.571 7 2,926 8 3,563 9 Corridors . _ 4,646 9 3,152 9 2.200 Patients' rooms, patients in 8,534 17 7378 23 5.951 22 5,558 23 7,273 20 8,378 20 Patients' rooms, patients out 1.035 2 346 420 I 482 2 366 1 632 2 Utility room 780 511 348 1 360 1 363 1 629 2 Tub and Shower room _ -- 258 181 76 22 199 399 1 Nurses' confer cute room 9.861 19 4.903 14 4,171 15 3,264 14 4,044 11 8,607 21 Off unit 6,955 13 3,325 10 2,741 10 2,296 9 4,387 12 4,814 11 Other areas 1,026 2 882 2 829 3 470 2 927 2 1,055 2 Total 51,631 100 34,008 100 27,654 100 24,172 100 36,579 100 41,165 100

Less than Ipercent. BEST COPY AVAILABLE 74 RESEARCH ON NURSE STAFFING IN HOSPITAI.s

Table 14. Utilization of nursing unit by nurse staffing patterns and radial andangular units, by percentage of observations only

Nurse staffing pattern and unit (percent only) Iligh A%erage low RN I.PN NA subarea ..%ngillarRadial AngularRadialAngularRadial AngularRadial AngularRadial Angular

Nurses' station 37 27 39 :11 :15 :15 :19 33 '49 39 36 25 Medication ;tin 4 4 5 4 5 4 6 6 5 5 3 2

Corridors . 9 9 8 10 9 7 5 8 8 8 8 9 Patients' rooms. 22 patients in 15 18 21 21 25 21 18 92 19 21 Patients' rooms.

patients out 2 2 I I 1 2 2 2 1 1 2 0 Utility room I 2 I 2 1 2 1 I I 2

Tub and shover room 1

Conference room 24 12 16 13 17 11 16 12 10 16 25

Off unit ;5 12 10 9 12 8 9 I0 12 12 12 11 Other ar-as 2 2 3 2 3 3 2 2 S 2 3 2

'Less thanipercent.

patients present and the opposite result on the Thus, it . staffing patterns, directly oppos- angular unit. The high NA staffing pattern ex- ing results . tilization of various subareas of hibited an increase in use of the nurses' station and a decrease in time spent in the nurses' con- the nursin ;unittook place betweenthe two ference room on theradial shapedunit, with physically shaped units. This again indicated some the opposite result on the angular unit. effects of such design on nurse staffing patterns.

Conclusions and Discussion

Space permits only the mention of some major ences in care provided by the different staffing conclusions fromnut-findings. As indicated by patterns. Also, expanding the divergency in phys- the measurements employed, nursing and patient icaldesignsbeyondmerelythecircularand care differed both in terms of nurse staffing pat- Lshaped angular types included inthis study tern and the physical design of two divergently may have revealed even more differences in im- shaped nursing unitsinan acute care general pact on patient:care. The same could well hold hospital. The results were not always uniform or for expanding the different types of nurse staffing predictable.Bothquantitativeandqualitative patterns along both qualitative and quantitative nurse staffing patterndifferentials were affected lines for additional divergences in results related differently by the two diveigently shaped nursing to patent care. units tor different variables related to patient care Only the most basic types of patient care were and utilization of the unit by nursing personnel. reported here, and such variables would be easily Undoubtedly,partof the reasonfor varying extended into many other phases of such care, results was the rather gross measurements ern. even with the same level of gross measurement, plcyed in this study. Most of the data were of We should perhaps also suggest that our overall the enumerative type with simple percentages of findings might have differed with other nursing the major measurement data. It seems likely that personnel in other hospitals, to mention only a more precise measures of nursing and patient few necessary qualifications of the results of this care than were available for this study may well st tidy. have produced more detailed and precise differ- Our major conclusion nevertheless remains that

83 NURSING STAFFING PA1TFRNS AND HOSPITAL. UNIT DESIGN: AN EXPERIMENTAL.ANALYSIS 75 hwpital physical design affected differentially the and so forth. We need to do the type of basic ,atient care provided by six different nurse staffing research whereby we candelineatespecificdi- patterns for general medical and surgical patients mensions of the caring relationship that can be atthe intermediate level of nursing care. From controlled or enhanced when specified for patient a strictly research standpoint, then,investigations care. of nurse staffing patterns should take the physical Sonic of the medical staff of the hospital where setting into account as a potential factor affecting our study was done would like ideallyfor the the outcome of such research. Omitting these con- hospital admitting office to pick out that room siderations and potential effects will likely bias or and that1,, : that would maximize the personal, distort the results of such research on nurse staffing physical, and medical needs of that patient, We patterns. dreamed about this, for we found that there were 0..ir study suggesteda number of theoretical a lot of personality Factors involved inboth posi- problems. First of all, to return to Dr. Aydelotte's tive and negative reactions to the unit, to certain suggestion that we need theory,I think we have nurses, to certain physicians, the way they were to start to do some pure Imsic research on just taken care of, and so on. Such reactions had a what wt' mean by currnot care in the nursing lot to do with the room and the whole setting sense or medical sense but care ingeneral. Ire- in which all of this kind of care, the total medical gard care as ba8ically a social phenomenon. The care, was given to patients. "caring" relationshipisdefined by society and culture. We can break the societal definition down These are some of the things that we touched into "hospital care" and then into categories of on in our study and some conclusions we reached "patioittcare,""nursing care,""medical care," as a result of the-study. 76 RESEARCH ON NURSE STAFFING IN HOSPITAI S

References

(I)Jao, Gart ly."Ecological Aspects of Pa- (12) Hall,EdwardT."Proxemics."Current ticot Care and Hospital Organisation." Anthropology, vol.9:83-108, AprilJune Organization Research on Health Institu- 1968. tions, Basil S. Georgoponlos, (ed.) institute (13) Alihan, M. A. Social Ecology. New York: forSocialResearch, University of Mich- Columbia University Press, 1938. igan, Ann Arbor, Michigan: 1972, chapter ;14) Lippert, S. "Travel as a Function of Nurs- 10, 418 pp. ing Unit Layout," Tufts University,. Human (2) Price, Elmina NI. Staffing for Patient Care. Engineering Information & Analysis Serv- New York: Springer Publishing Co., 1970. ice Report No. 113, March 12, 1969. 178 pp. (15) McLaughlin, H. P. "What ShapeisBest (3)Aydelotte, Myrtle K.,etal. An Investiga- for Nursing Units?" Modern Hospital, vol. tion of the Relation between Nursing Ac- 103:84-89, December 1964. tivityandPatientWelfare.IowaCity, (16) Pelletier,R.J.,and Thompson,J.D. Iowa: State University of Iowa,1960. 412 "Yale Index Measures DesignEfficiency." pp. Modern Hospital, vol. 95, November 1960. (4) Sturdavl.nt.Madelyne,etal.Comparisons (17) Rosengren, WilliamR., and De Vault,S. of Intensive Nursing Service in a Cirrular "The Sociology of Time and Space in an and a Rectangular Unit. Chicago: Amer- Obstetrical Hospital." in The Hospital in kan Hospital Association, Monograph Se- Modern Society, Eliot Freidson, (ed.) New ries No. 8, 1960. 220 pp. York: The Free Press, 1963, chapter 9. (5)Trites. David K.,etal."Radial Nursing (18) Sommer,RobertandDewar,A."The UnitsProveBestinControlledStudy." Physical Environment ofthe Ward."in Modern Hospital, vol:109:94-99, April The HospitalinModern Society,Eliot 1969. Freidson, (ed.) New York: The Free Press, (6) "Social Interaction as the Definition of the 1963, .hapter 11. Group in Time and Space." Translated ;19) Sorokin, .Peterim A. Sociocultural Causality, from Georg Simmel, Soziologie, by Albion Space,Time. Durham: Duke University W. Small, in Robert E. Park and Ercie-st W. Press, 1943. Burgess,IntroductiontotheScienceof (20) Thompson, John D. "Efficiency and Design Sociology Chicago: University of Chicago in the Hospital Inpatient Unit." Poceed- Press, 1921. ingsof the13thInternational Hospital (7)Hawley, Amos H .uman Ecology. New Congress,(Paris,1963). London: Interna- York: Ronald Press, 1950. tionalHospital Federation, 1963. pp.- 67- (8)Hall, Edward T. The Hidden Dimension, 75. New York: Doubleday & Co.,Inc.,1969. (21) Trites,DavidK.:Galbraith,F.D.Jr.; 217 pp. Sturdavant, M.; and Leckwart, J. F. "Final (9)Sommer,Robert.PersonalSifre.Engle- Reprt of Research Project Entitled:In- wood Cliffs, NJ.: Prentice-HalL Inc., 1969. fluence of Nursing Unit Design on the Ac- 177 pp. tivities and Subjective Feelings of Nursing (10) Barker, R. G. Ecological Psychology. Personnel." Rochester, Minnesota, Re- Stanford: Stanford University Press,1968. search Office, Rochester methodist Hospi- (11) Osmbnd,H."FunctionastheBasisof tal, 1969. Psychiatric Ward Design." Mental Hospi- (22) Wheeler,E. Todd. Hospi:al Design and tals, April 1957 (architectural supplement), Function. New York: McGraw-Hill Book pp. 23-29. Co., 1964. 296 pp. DISCUSSION OF DR. JACO'SPAPER

Discussion Leader Dr. Stanley Jacobs American Hospital Association Chicago, Illkois

Dr. Jacobs

Perhaps I was most impressed by theexperimental de- stimulated by the Kellogg Fouadation. social scientists sign of Dr. jaco's study. The numberof variables the entry of a substantial number of he was able to manipulate in acontrolled situ- could be similarly stimulated by theU.S. Depart- of Health, Education, and Welfare' ationisnot often encounteredinhospitalre'. ment search. The study was more like aninvestigation (DHEW). The American PsychologicalAssoeik- carried mu in a laboratory than oneperformed tion (APA)islocatedin the Washington area: in an operating organization. He is tobe com- As astarter; DHEW mightmake arrangements convention to point mended for "selling"the hospital management for a symposium at the APA and nursing problems on cooperating in the research. out some of the hospital research op- Dr. Jaco, as most of you know, is asociologist. and to inform psychologists of the I, by training, am an industrialpsychologist. Very poi tuni t ies. sufficient body few nursing or personnel research studiesin hos- In some cases there is already a predicttheout- arebeing designed and directedby -in- ofpsychological knowledge to pitals today dividuals with degrees in the social sciences.Most comes of staffing decisions. Earlier such studies reported in the literature havebeen Dr. jelinek pointed out that nurses arereluctant conducted by industrial engineers. They have made to float when they areidentified primarily with wilting many significant contributions.I do riot believe, units in which morale is high, but they are however, that most of them would conceiveof to float from units in whichmorale is low. Most experimental designs as sophisticated asthat de- psychologists would have predictedthis finding. My basic appealinthese commentsisthat, scribed by Dr. Taro. he able Although I have been out of industrial psychol- through research, social scientists should the solution ogy for some time,Istillscan the journals of to make a significant contribution to this discipline. Very seldom do 1see articles de- of many of the problems we havebeen discussing could cer- scribing research in a hospital setting by psychol- in this conference, In addition, they making per- ogists, other than clinicalpsychologists. Infact tainly assist hospital management in the only one that comes to mind isa study at sonnel decisions. Iowa State in which size of staff andpersonal I had hoped when Ifirst saw the title of Dr. relations training for nurses were varied inorder jaco's paper that he was going totellits which patient reactionto was "best," radial orconventional units. He did to determine impact upon he told us of a wealth care received. not do that. However, The rapid growth in the number ofindustrial of other findings discovered during the courseof engineers working in the hospital field wasgreatly the study.

77 78 RESEARCH ON NURSE STAFFING IN HOSPITALS

In regard to hospital architecture, thereare two Finally, I strongly concur in Dr. Jaco's comment reports recently prepared for the Department of aboutreplicationofresearches.Inpsychology, Defense concerning a "new generation ofmili- replication of researches(orcross validationof tary hospitals."Iprocured them becauseI was personnel tests)is most important before one ac- interestedin what they had to say about com- cepts findings with no reservations. We seldom puterization. The two reports consist of nineanti do thisinhospitalresearch. No matte.: how seven volumes, respectively. One was rrepared by sophisticated our statistical tests, significi:nt studies 1Vestinghouse and the other by Arthur D. Little. should be replicatedbeforehndingsare ac- They have a good deal tosay concerning archi- cepted, as doctrine and decisions are basedupon tecttire for hospitals and nursing units therein. them. Impact of Administrativeand Cost Factors onNurse Staffing

Mr. John R. Griffith PTOYMOT and Director Program and Bureau of Hospital Administration The University of Michigan Ann Arbor, Michigan

To explore the impact of administrativeand While this model of the establishmentof nurse cost factors on nurse staffing. onemight use the staffing policies is simplistic, it suggeststhat the model developed by ',con, which implies that the term "administrative and costfactors" would be policies and operating procedures of ahospital all those demands and desiderata ofthe outside at any given time represent the sumof two in- parties. Such a listis constantly changing and is potentially quite lengthy. If the majorgoals of fluences. These are the influences generatedtyy is possible forces from the outside environment and those the outside bargainers are postulated, it to predict something of the natureof the outcome generated from within the organization by proof the bargaining process. This then canbe a fessionalandemployeeattitudesandknowl- guide to understanding the current evidence on edge (1).In the case of nursing, the parties to the actual state of nurse staffing and todrawing the negotiation bring quite divergentbackgrounds some tentative conclusions onthe accuracy of the and therefore have, as a result, quite different sets modelandpostulate.Finally,the model can of cognitive information, attitudes, and values. frame %peculation upon the impact of the major The bargaining process would incorporatethe changes in the demands of the outsideparties usual searches for mutually acceptablesolutions, that currently appear possible. compromises, and innovationsthat meet orat Obviously, there are some weaknesses inthis of in- leastpartiallysatisfymultiple goals simultane- approach. Any assumption about the goals ously. The principal parties from the outside en- dividuals or groups is subject to question.The vironment would be third party financing agen- lists of both outside and inside parties havebeen cies,trustees. and hospital administrators. From arbitrarily curtailed, and because of the pressure within the organization, they would be the proof time, the limited power of ouranalytic tools, fessional and employee groups of nurses, doctors, and the shortage of empirical evidence, wewill and nonprofessional nursing employees. be limited to some general conclusions.

79 80 RESEARCH ON NURSE STAFFING IN HOSPITALS

Problem Definition

The goal of nurse staffingisdelivery of an ated. That does not preclude the use of substantial appropriate quantity and skilllevelof nursing factual material. In fact, one might infer thatif service to each individual patient. The issues that the bargaining were proceeding normally,a wide must be resolved are the definitions of appropri- variety of both evidence and argument would be ate quantity and appropriate skilllevel and the introduced. The linesof argument willfollow mechanisms of delivery, including the tolerances recognizable patterns, however, of which twoare to which the final nurse staffing system isto be perceptible. These can be labeled the engineering held. Theseissuesarerelatedto,but can be argument and the economic argument. separated from, a varietyoforganizational parameters that influence the outcome. The organizational parameters include the training and Engineering Approach recognition of capability of each of the possible skilllevels,the physicalfacilities of the wards, The more commonly encountered definition of the appropriate quantity of nursing service could the nature of supervisory andmanagement structure under which patient care is given, and the be labeled as An engineering approach, both be- uganivatiunal structure of ate hospice iNN v.;tit cause enginefrs frequently erl t and because the nursing staff performs. it can be traced to Taylor's (11) analysis of job shop time standards. This approachassumes there For example, one might assume that theup- is a fixed, describable quantity of activities that grading of registered nurse skillstothe bacca- must be performed for a given patient in a given laureate level would influence the relativequan- time period. Further, given a set of organizational tities of registered nurses and practicalnurses. parameters,the incrementaltimesfor Similarly, decisionsthatregisterednurses may eachof these activities c.r the total time for the full set draw blood specimens or that theyare responsi- can be measured by standard work measurement blefor admitting histories would influencethe techniques. This approach implies that appropriate quantities of RN skilllevels. The atleast a major portion of nursing care can be described architecture of the patientcare units might aflect in terms of discrete, identifiable, measurable thestaffing (2,3),and so might decisionsthat components, and that for any given patient there is change the management style of the headnurses reasonable consensus as to which of these and higher levels of supervision (4), andso would components is appropriate. organizational decisions thatcreate patient unit management departments. The approach has a reasonably successful his- tory in industrial job shops. The underlying as- The goal definition suggests thatoncethese sumptions, however, have not been testedina organizational parameters have been fixedthere nursing environment. While existing work shows exists a unique quantity of nursingresource that that time estimates can be made with satisfactory would be judged appropriate to the needs ofany variancesFor both large and small components given patient for a given time period, suchas a of activity, the author is not aware of any work day or a shift.. This seems to he consistentwith that shows consensus on the components desirable most thought on the problem. Such commentas for specific patients. Neither has it been demon- there is on the appropriate level ofcare is given in hours per patient days (5). Many of the existing strated that the components are trvtly additive. Another difficulty has been oveooked. That is empirical imestigations tend totreat this as hav- thatin the industrial job shopitisthe usual ingfacevalidityasthe dependent variable (6). Further, there appears to be aconsensus that the practice for managemtlit to specify exactly what appropriate quantity and skilllevelfor an in- should be done, butinthe ho ')italitisthe dividual patient isa function of the severity of usual practice for management to delegate to the his illness (7,8,9). registered nurse it, charge of the patient the iden. The bargaining model suggests that theeen- tification of the patient's needs and the work to tual definition of "appropriate" will be negoti- be performed. Evenif nursing supervision were IMPACT OF ADMINISTRATIVE AND COST FACTORS ON NURSE STAFFING Sl willing arid able to approve musing care plans pend t tireonnursingcareisequaltohis or some other specification of workfor individual perception of the utility of the same number of patients, it:s quite difficult to determine whether dollars if spent for other goods and services.Since these specifications have been met. he will also buy every other good and service Atthepresent time there appear to he no accordingtothe same standard, he will make practical systems for assessing the quality of the alternativepurchasesas,for example, between result. Thus, while the engineer may measure to nursing care and laboratory services, until their his heart's content what he thinks might be done marginal utilities are also equal. This means that for the patient, the nurse on the floor will con- he will purchase nursing care over any alternative tinue to do what she thinks is desirable for the service until he has purchased the quantity of patient. The two decisionsystems can coexist nursing care where its marginal utilityis exactly nicely so long as the engineer does not force the equal to the marginal utility of the lastquantity nurse into a situation where she isroutinely dis- of another service that he purchased. satisfied with her professional performance or her Put another and more concrete way, the pa- work pace. tient will spend $10 for nursing care so long as he feels he can afford the $10 and so long ashe feels that $10 worth of nursing careis at least Economic Approach ,as valuable as $10 worthof laboratory service or any other good or serviceavailable to him. The The other approach to the question of the ap- equilibrium that is reached will be a function of propriate quantity and skill level of nursing care the following: can be called an economic approach.Although the patient's income, because this determines itisconceptually atleast equally appealing,it has received very little attention in the literature. how willing he is to spend the money; This approach would assume that nursing care is the patient's preference or tastes, because these a resource sought by a consumer or aphysician determine on what he wishes to spend his on his behalf because he perceivesthatithas money; some utility for him. 'Thus, in some more orless the nurse's income, because this determines imperfectfashion,the amount of nursing care the cost of the marginal unit; that will he judged appropriate can he predicted the state of health care technology, because by economic theories rather than by strictly em- this otp! Ilipes the utility of a given phys- pirical considerations as the engineering approach ic41 qnant4y of tmrsing resource and also would suggest. Under these theories, nursing care the 9114Y of r titergoods and services; has a certain utility, and the utility is generally the general economy, because thisinfluences of the patient dependent upon the quantity rf care received. not only the relative incomes of is difficultto define this relation- and the nurse but also tie relative incomes Although it labor and cap- ship, the marginal utility, or the value of an ad- other resources, including both ditional increment of a given good or service,is ital resources. generally envisioned as a downward sloping curve, To summarizetheeconomicapproach, one as3 howr in figure1.If we can assume for the might assume thatthe appropriate quantity of moment that a typical hospital patientbuys his nurse staffing,than, is determined by people's per- own ntrsing care,thisrelationshipisnot un- ceptionsofrelativeincomesof patientsand reasonable. The firstcare that he receives may muses, by u.chnology, and by thedesirability of helifesaving,:endthereforelie would valticit spending money generally for health care. infinitelyhighly. As the quantity goes up, the Patients, of course, do not purchase their own utility of additional units goes down and even- nursing care. The decisions that must be made malt reaches a value very close to zero. are collectivizedinseveraldifferent ways. The At least theoretically,itis now possible to de- nursing resource is provided to a nursing unit, cide the appropriate amount of care.It will be the quantity of resource availableis sometimes that amount at which his perception of th.: mar- limited, the impact of the patient's personal in- ginal utility of a given number of dollars af ex. come is to a large extentalleviated by third party

90 82 RESEARCH ON NURSE STAFFING IN HOSPITALS

payment mechanisms, and the problem of evalu- professionsprimarilynursesinthiscasewill ating the marginalutilitiesof various hospital bargain for the purchase of additional nursing goods and servicesis assigned to the bargaining services. They will attempt to argue that the util- processbetweentheoutsidepartiesandthe ity of another nursing service exceeds its cost and nurses and other professionals within the hospital. exceeds the value of any other good or service Inthis regard, however,it must be noted that available. Their arguments will center on quality while the physician acts as the patient's agent in and on the desirability of identifiable elements of selecting the qua titles of most of the other goods service that will be foregone if their demands are and services in the hospital,it remains the nurse not met. The administrator and the trustees wil: who makes must of the decisions in regard to bargain toprotect the patient's purse, arguing 'nursing care. that he needs the money or that there are other In some ways,thiscollectivizationis an im- goods or services that have a higher utility.If portant asset to our problem. If the severity of these stereotypes apply and there is "hard bar-. illness is a random phenomenon, itis reduced' by gaining" oraconsiderable impact of adminis- treating patients in groups. The elimination of trative andcostfactors,therewill be certain individual patient income from considerationis predictable results. clearly desirable. There are a number of obvious First, within the same labor market, under the economies thatresultfrom grouping patients. sair, setof third party arrangemems, and for Finally, the data (rem thengineering approach similar groups 01 patients, similar groups of bar- are not discarded. They enter the arguments as gainers would tendto have the same number a way to evaluate the translation of inputsnurs- of hours per patient day. Second, there would be ing hours, to benefitsnursing service. considerable pressure to identify and install the The normative rolesusually assignedtothe least-cost organizational parameters. All those that various decision makers inthe hospital suggest cost less than the value of the nursing iime they how the bargainers might behave. The clinical release would be installed because this 6 to the

Figure 1.Marginal utility of nursing care.

0 Quantity of nursing care IMPACT OF ADMINISTRATIVE AND COST FACTORS ON NURSE. STAFFING 83 advantage of both parties. Again, one would ex- The second case of lack of consensus would he pect consensus to develop am ng hospitals in shn- where the utilityis perceived very differently by ilar situations. Third, given tat nursing careis differentparties to the....bargaining. Under this a high priced and scarce. resource, considerable case the solution would reflect the most powerful ,,ttention would be given to controlling the actual bargaininggroup.One mightexpectactual delivery of nursing care so that it would not devi- solutions to be subjectto sharp variations de- ate from the level agreed on as appropriate. pending on which group isin control. The situ- There are some special cases of this analysis ation would probably be accompanied by con- that must be discussed briefly. One isthe case siderable tension and dissatisfaction. These two where the supply ofnursesisconsistentlyless cases of lack of consensus can be compared to a than the demand attheprevailing wages. As ship. In thefirstcase,no oneisinterestedin has been indicated by Yett (12),this situation where the ship is going and it is allowed to drift. would be characterized by steadily rising hospital In the second case, the ship sails smartlyina nurse wages ataratefasterthan theriseof variety of directions depending on who is at the wages in other nurse employment or other similar wheel. employment opportunities. It also would be char- It is of some interest to consider what happens cterized by an even greater pressure to seek nurse tothe engineering approach under theseeco- saving changes in the organizational parameters. nomic cases. Where there is consensus there will he more interesting case is where there is no be considerable interest in the engineering ap- utilityconsensus asa resultof the bargaining proach to the extent that it rationalizes the agreed process. One possibility is that there is nolard upon decision and is useful in finding least-cost bargaining" or that the bargainers perceive the organizationalparameters. Otherwiseitwill be marginal utility of additional nurses as being more ignored.Italsowillbe ignoredin bothcases difficult to estimate than the impact it will have where thereis no strong consensus. Where the on the outcome. Under this situation,the deci- shipisdrifting aimlessly, no one will wish to sion as to the amount of nurse staffing may be invest the money necessary to evaluate various treated as unimportant and the amount of nurse staffing would tend to be reduced slowly to the nursingactivities. Where thereisa continual point where a strong consensus could be reached struggleforthe helm, allenergies will go toon the undesirable impact ofreducing it further. wards the battle.

The Current Performance

The implications of the engineering approach tion in nurse staffing between hospitals that are taken alone and of the economic approach in the similar in these characteristics is large, then either analysis of theevidence withregardto nurse the bargainers do not care about the results or staffing are quite different. The engineering ap- they cannot reach a stable agreement. Neither one proach would imply that hospitals that have a of these finch J tad be so. 'ally useful.Itis givensetof organizational parameters and that incongruous to say that no one cares about the treat patients of similar averageseverity of illness most expensive single resource thatthe hospital should have the same quantities of nursing care: purchases. Neither is it a hopeful sign to contem- The economic approach would suggta that fac- plate continued tension and abrupt changes in such as tors quite outside the organization itself, the quantity of nursing resource. the extent and kind of third party coverage and the wages of nurses relative to other health care resources, would also affect the amount of nurse Actual Nurse Staffing staffing. Also, the economic approach provides additional Ilnfortunately, the definitive research suggested explanatory power. It suggests thatif the varia- here does not appear to have been clone. There arc 84 BEST COPY AVAILABLE RESEARCH ON NURSE STAFFING IN HOSPITALS a number of empirical problems that make one nopattern between economically similar States. hesitate to attack such research. Such evidence as Michigan is 20 percent lower than Illinois,for we do have is far from encouraging. One finds, for example. The causes arc likely to lie in licensure example, that the number of nurses employed in examination levels, acceptance of reciprocity, and hospitals by the State bears little relation to the the presence of schools of nursing. Rut these arc bed supply of those States (refer to table1). Not items that over long periods of time are within only do the States range from 171nurses per the control of the bargainers. 1,000 beds to 436 nurses per 1,000 beds, a difference Data compiled by Hospital Administrative Serv- of over 150 percent, bat adjacent States vary by ices (HAS) on the number of nurse hours per 10 percent or more. For example, see New jersey at 300 nt,rses per patient day similarly show an extraordinary varia- 1,000 beds and New Yo '4, Pennsylvania, and Dela tion. HAS segregated hospitals by State, district, ware, or Massachusetts, Rhode Island, and Con- function, and size, thus making at least crude ap- necticut. The widest of these variations is between proximations for many of the explanatory varia- Arizona, the second highest State, and New Mexico bles. These data, shown in table 2, are quite recent -424 and 264 RN's per 100 beds. There alsois but include only 200 self-selected hospitals. The

Table 1.-Loca*ion of RW8 employed in hospitnts ift gelation to hospital beds: 1968

RN's per RN's per Location RN's 1,000 hospital beds Location RN's 1,000 hospital beds

United States. 445.243 268 East North Central New England Illinois 27,653 269 Connecticut 8,179 336 Indiana .. 9.432 242 Maine 2,498 257 Michigan 16,338 229 Massachusetts 20,007 313 22,894 281 New Hampshire 2,142 319 Wisconsin __ . 10,357 275 Rhode Island 2,527 289 West North Central Vermont 1,361 289 Iowa - 7,284 353 Middle Atlantic Kansas 5,129 258 New Jersey - 15.715 300 Minnesota 11.668 349 New York _. _ 49,290 239 Missouri 8,594 221 Pennsylvania _ 33,392 287 Nebraska 4.204 323 South Atlantic North Dakota 1,832 294 Delaware 1,242 248 South Dakota 1,711 267 District of Columbia 3,610 236 Mountain Florida 13,079 294 Arizona 3,936 424 Georgia 6,877 202 Colorado 6,084 367 Maryland 8,172 245 Idaho 1,391 352 North Carolina 8,642 243 Montana _ 1,960 425 South Carolina 3,876 206 Nevada 896 322 Virginia 8,707 226 New Mexico 1.580 204 West Virginia 4,037 232 litah . 1,933 384 East South Central Wyoming 796 197 Alabama 4,986 175 Pacific Kentucky 5,104 222 Alaska 713 357 Mississippi 2,837 171 California 44,137 330 Tennessee 6.051 183 West South Central Hawaii 1,775 284 Oregon 5,032 333 Arkansas 2,340 211 Washington 8,186 Louisiana ,978 186 436 Oklahoma 3,741 225 Texas 16,338 218

SOURCE:Nursing Personnel in Hospitals, 1964. P.S. Departmentof Health, Education. and Welfare. Public Health Service, National Institutes of Health, Washington: U.S. Government Printing Office, 1970. IMPACT OF ADMINISTRATIVE AND COST FACTORS ON NURSE. STAFFING 85

Table 2.-Median nursing hours per adult medical surgical patient day, HAS subscriber hospitals, 6 months ending 12/31/71

Bed site

Region Under 50- 75- 100- 150- 200- 300- Over 50 74 99 149 199 299 399 400

New England 14.04 7.07 6.90 6.71 6.56 6 23 6.40 6.46 Middle Athol tic 8.22 6.24 5.91 5.53 5.84 5.60 5.72 5.54 South Atlantic 7.27 6.41 6.80 6.26 6.09 6.12 6.47 6.04 East South Central 6.39 6.60 6.31 6.46 5.98 6.03 5.93 5.8? West South Central 6.62 6.65 7.49 6.80 6.25 6.08 5.42 5.36 East North Central 6.65 6.72 6.73 6.49 6.74 6.48 6.33 5.98 West North Central 7.05 6.58 6.87 6.71 6.93 6.32 6.28 6.51 Mountain 6.78 6.73 6.47 6.84 7.81 6.95 6.53 7.01 Pacific 7.51 7.83 6.92 7.07 6.87 7.14 6.50 7.10 SOL'RCE SisMonth National Comparison. Hospital Administrative Services, (HAS) The American Hospital Association, Chi. cago. patient care on ,nredical-surgical units, not simply The follow:ngchart summarizes the variationwith- for registered nurses. in size gr mtp: Table 2 shows the siiadjusted regional medians. The datashow some pattern. The Pacificregion

Size High Region Low Region Under 50 beds Mid Atlantic 8.22 E.N. eentral 6.39 50-74 beds Pacific 7.83 Mid Atlantic 6.24 75-99 beds E.S. Central 7.49 Mid Atlantic 5.91 100-149 beds Pacific 7.07 Mid Atlantic 5.53 150-199 beds Mountain 7.81 Mid Atlantic 5.84 200-299 beds Pacific 7.14 _Mid Atlantic 5.60 300-399 beds Mountain 6.53 E.S. Central 5.42 400 and over Pacific 7.10 E.S. Central 5.36 tends to be high and Mid-Atlantic low. But both little, at least at the low end of the scale. All the regions are reasonably well supplied with registered htth percentile values are within 10 percent of the nurses, according to the older data on table 1, and lowest hospital, which isan over-400 bed, non- theMid-Atlantic States have thepeculiarity of teaching hospital with 4.4 hours per adult medical being high for small hospitals aid at or near the surgical patient day. The variation increases sub- low for all other sites. Further. one would infer stantially for the 95th percentile, where the low from these data a demapd for. more nursing per- value is 7.15 and the high is one-third more, 10.02. sonnel for very small hospitals, fewer for hospitals The ranges within size groups, however, cover in- up to 150 beds, and fewest for large hospitals creases of 80 to 150 percent. either 300-400 beds or over 400 (seefigure 2). The conclusion the author draws from these There may he explanations for this. The analysis data is that there is no consensus on the question of returns to scale at ,.'aiying levels of service by Carr and Feldstinis suggested by the shape of of nursing t-tility. The reurring wide variations failure of known theoryto thecost curves (13). Problems of reporting and andthe apparent nonrepresentative samples cannot he ignored. explain the cause suggest that nurse staffingis The ranges of lmhavibr for all hospitals in the simply rn)t subject to administrative and cost fac- size groupsis shown intable3. One isforced tors.Uodoubtedly, more careful research would to conclude from these data that size matters very revealconsiderableimpacttr:)inorganizational

94 86 RESEARCH ON NURSE STAFFING IN HOSPITALS BEST COPY AVAILABLE

Figure 2.High and low region medium AM&S mrsing hours per patient day.

9.0

8.0

7.0

13 6.0

5.0 a. k4.0

3.0

2.0

1.0

< 50 50-74 75-99 .i00-149150-19S200-299 300-399 4004 Beds SOURCE : TABLE 2

High MK: Low IMPACT OF ADMINISTRATIVE. AND COST FACTORS ON NURSE STAFFING 87

Table 3.-Percentiles of nursing hours per medical surgical patient days in HAS subscriber hospitals, 3 months ending 10/31/71

5th Ist 3rd 95th lied site percentile quartile Median quartile percentile

Und'r 50 4.68 5.92 6.88 8.02 10.02

50-74 . 4.87 6.04 8.79 7.51 9.52 75-99 4.82 5.97 6,59 7.29 9.13 100-149 4.84 5.79 6.47 7.22 8.63 150-199 4.58 5.54 6.31 6.94 8.09

200-299 _ _ 4.76 5.51 6.04 6.82 8.18 300-399 4.61 5.55 6.11 6.70 7.93 7.45 Over 400 . 4,40 5.22 5.74 6.38 8.98 Teaching _ _ 4.53 5.67 6.36 7.23 SOURCZ: Hospital Administrative Services (HAS). prepared on request of author.

parameters and environmental factors, but still, have also been available and have been demon- substantial random or unexplained varianceis strated to be feasible. Computer aided scheduling likely to remain. systems have also been developed and at least one One study that relates to this question yields of thesehas been marketed commercially (16). essentially this finding. Ingbar, Whitney, and Taylor Both manual and computer scheduling systems performed multivaria:D analyses on nursing hours rely upon the use of "float pools" or reassignment per patient day in 72 Massachusetts community of nurses to maintain desired levels of staffing. hospitali using 1958 and 1959 data (14). At that They have an additional advantage in that they time, five factors measuring the type and scope of incorporate measures of their own success. Thus, services explained 34 percent of the variance. Sur- where these models have been put into effect,it gical activity was the largest, explaining 22 percent. is possible not only to strike a bargain regarding A measure of ambulatory activity(X-rays)ex- the appropriate amount of nurse care but also to plained 6 percent. Measures of ward patients and monitor how closely it has been achieved. Even educational activities explained the balance. The manual systems are susceptible to a variety of kinds use of nursing students explained an additional of monitoring, either of total hours delivered or of 20 percent of the overall variance. Forty-six per- hours per patient day delivered. The HAS system cent of the variance remained unexplained. This cited above provides routine data on this question rather large residualisnot suggestive of clear together with standards of comparison for individ- consensus. ual hospitals.1 So do several regional systems of manpower planning and control such as CASH (10,17). Contrcl 7'cchnolagy It seems reasonable to conclude that the technology for monitoring and controlling nurse staffing Although available data on actual nurse staffing is feasible. There is no evidence, however, that it and distribution of nurses reveal substantial varia- is widely applied-say in as many as 20 percent of tionsinact..1 performance,itappears thatat U.S. community hospitals. Without extensive sur- least crude technologies are available that could veys itis difficult to make this conclusion defini- permit more careful control of this resource. Pro- tively. Yet the existence of continued variation in gressive patient care concepts of grouping patients the very measures that would be used for control according to the severity and kind of need have is strongly suggestive of a lack of what one would been around for many years, and at least the in- call a closely bargained outcome. tensive type care units have proved popular. These The same argument may be extended to the help in controlling severity-of-illness variance and various organizational parameters. Other than in- establishing appropriate care, but they do not tensive care units and coronary care units, these eliminate tt.,: staffing problem (15). Manual methods for maintenance of control of nurse staffing ' HAS, The American Hospital Association, Chicago. Illinois.

96 88 RESEARCH (.)N ,NURSE STAFFING IN II)SPITALS have not even received enough attentionto he tends to the conclusion that the trustees, Adminis- properly validated.' Often,illlact,the goals of trators, directors, and nurses who are central to installing these systems, such as Service Unit Nlan- the bargaining process do not appear to have acted agement System, appear to he different from a goal in a manner that either an economic or an engi- offreeing nursing resources (18). Thus changes neering approach would dictate. To return to the in organizational parameters have only infrequently sailing analogy, the shipisadrift. The adminis- reached the level of consensus in the bargaining trative and cost factors of nurse stalling have not process, and when they have, the reasons have sometimes been tangential to the direct problem had any perceptible impact. It might be desirable of nurse staffing. to explore some of the assumptions and implica- The evidence, unfortunately,iscircumstantial tions of the',odel in order to identify potential and unsatisfactory, but on a number of fronts it reasons that this is so.

Difficulties in the Control of Nurse Staffing

If we postulate economically rational bargainers Activities Study (PAS),exist and are routinely working within the framework of feasible tech- used (21). Research in medicine tends to validate nology, there seems to be au identifiable set of the procedures against outcomes. A body of vali- potential and actual difficulties. Empirical speci- dated procedures exists as a result of the decades fication of these is again difficult, partially because of intense effort. the documentation of why things do not work in Some efforts are underway at Michigan to develop the healthfieldisquite limited (19). Such dif- procedo,' and outcomes measures of nursing per- ficulties as are known, however, can be classified formance in a manner analogous to PAS.= In this as conceptual, technical, and professional. approach, the activities and the outcomes that clinical nursing judgment suggests are desirable can be specifiedfor patients classified according Conceptual Difficulties to their diagnoses. Some desirable activities would be common to every patient. The list of quality Cone eptual difficulties obscure the bargaining imliators compiled by CASH and developed and process and impede the solution because confusion used by others is an example (22). Other activi- exists in the bargainers's mind asto what the ties would be quite specific to the diagnoses, as bargaining issues are. The most serious conceptual the attached prototype for diabetic patients slug- difficulty in nursing appears to be the lack of clear definitions of nursing roles and the lack of meas- g6ts (see figure 3). ures of quality of nursing activity.Quality, of Preliminary work on this approachisalready course, is closely related to utility. Although" there underway.2 The specification of diseaserelated items of care that can be reliably assessed by an are some problemsinassuming thattheyare synonymous, the measurement of quality clearly independent observer appears feasible. Work this sharpens the understanding of utility. Concepts summerwillinvestigatewhethereconomically and methods of measurement of quality need to practical surveys will yield sensitive and reliable be much more clearly developed. assessments. The effortisstill some years from the point of universal day-to-day use.It will be In medical (thatis,physician) care,itispos- sible_ to identify structural, procedural, and out- a costly instrument untilit can be. reduced to a limited utunber of representative indicators. More comes measures of care. Although structural meas- important, itappears that nursing staffs will re- ures turn out to be of limited use pod outcomes (wire considerable reeducation to orient themselves measuresareincompleteproceduralmeasures to this form of purposive professional behavior, existin considerable detail, atleastfor the in- The effortis consistent with the philosophy of patienthospitalportion (20). Whole systems of management and analysis, such as the Professional smith. II L. and Horn, B.I.., personal communication. IMPACT OF ADMINISTRATIVE ANDCOSTFACTORS ON NURSE STAFFING 89

Figure 3.Prototype outcomes quality assessment instrument, patient observation section, diabetes mellitus. Patient Observations I.Have you been giving your own insulin? YES_ NO (a) (If YES) When did you begin? (b) (If YES) How competent do you feel? VERY__ SOMEWHAT_ NOT AT ALL____ (c) (It FONIEWHAT or NOT AT ALL) Why? _ _

(d) What material and aids do you have? 2.Have you been selecting your own diet? YES_ NO_

(a) (If YES) When did you begin? . (b) (If YES) How competent do you feel? VERY_ SOMEWHAT_ NOT AT ALL___ (c) (If SOMEWHAT or NOT AT ALL) Why?

(d) What material and aids do you have?

3. Have your been doing your own urine testing? YES_ NO (a) (If YES) When did you begin? (b) (If YES) How competent do you feel? VERY SOMEWHAT_ NOT AT ALL (c) (If SOMEWHAT or NOT AT ALL) Why?

(d) What material and aids do you have? 4.Injection observation (a) What insulin did the doctor tell you to take? Type Strength Units Freq (b) What unit syringe do you have or plan to buy? 40 80 40/80____ (c) Observation of patient's selfinjection: Touch needle YES_ NO_ Clean vial top YES__ NO Insert air in vial YES__ NO Correct dosage YES NO If NO, specify Cleanse site YES_ NO If NO, specify Injection: into subcutaneous tissue YES NO If NC), adipose YES_ intramuscular YES_ Post injection: Gently wipe site with spongeYES NO__ Comments

(d) Observed condition of injection site: Hardness YES_ NO Inflamed YES_ NO Bruised YES_ NO Comments_ _

5.Urine testing (a) What tests has your doctor asked you to use? Clinitest Acetest Freq Dyestick Ketostix Timing of testing___ __ TesTape (b) Describe procedure you will be doing: Empty bladder YES_ NO--__ RESEARCH ON NURSE S'IAFFING IN HOSPITALS

Figure 3. (con.) Drink water YES___ NO_._ Void and test YES__ NO (c) Test time with respect to meal time: Prior (minutes)__ Skipped ( (d) Observation of patient's selfurinalysis:

Sugar test CORRECT READING _ INCORRECT READING.__.. Acetone test CORRECT READING__ _INCORRE.:T READING Comments

6.Diet knowledge (a) What are your doctor's orders concerning your own diet?

Calories___CHO_ _Protein__..__ Fats__ . Comments

(b) Observation of patient's menu selection (see attached copy of menu) Correct calories:CFI() YES_ NO__.. Protein YES___ Fats YES___ NO____ Comments

(c) What materials do you have pertaining to your diet to assist you with diet selection? Exchange ___Diet selection sheet_____ General food discussions_ Diet record form______Exchange instructions_____ Other____

7.Foot care observation (a) Cleanliness ADEQUATE__ NOT ADEQUATE___

(b) Toenails (cut) STRAIGHT CURVED__ LONG _ (c) Dryness NONE SOME SEVERE_ _ (If, SOME or SEVERE) Isthis condition the SAME__ WORSE____ BETTER__ than a few days ago? Desquamation NONE._. SOME._ SEVERE._ Daily use of lubricants YES__ NO_____ (d) Calluses NONE__ SOME__ SEVERE__

(If SOME or SEVERE) Are they the SAME___ WORSE____ BETTER_ . than a few days ago? (If SOME or SEVERE) What arc you doing about it?

8.Foot care information For each of the followinr,, is it or is it not recommended for your feet? Wearing wool socks Metal arch supports Wearing loafers Walking barefoot Wearing elastic garters Wearing sneakers Hot water bottle Foot powder SOURCE: R. L. Smith and B. j. Horn, The University of Michigan, Bureau of Hospital Administration. (This figure is a draft and should not be reproduced.)

99 IMPACT OF ADMINISTRATIVE AND COSTFACTORS ON NURSE STAFFING 91 clinical nurse specialitation, Imwever, and it should first task is for the hospital to arrange a predictable be useful soon in at least research or nonrepetitive supply of nurses and other nursing personnel for applications. In the meamime, most of the existing each day and shift and to set that level somewhere quality measures, such a.; large part of the CASH near the point where an appropriate quantity of instrument, have been tumid to be lacking in re- service can be delivered. Thisis the schet4ding liability or sensitivity or validity..3 In the absence problem that arranges days on, days off, and shift of workable quality measurement systems, debate assignntents for t' ;eh individual employee, subject on nurse staffing takes the character of specula- to wage and hour legislation, union contracts, ab- tion, and a consensus from the bargainingprocess senteeism, and (Ls'ttl number of personnel report- becomes much less likely. ing. The most corttnonly reported solutions have Also, thereis a semitechnical, semiconceptual been cyclic schedules that claim to permit thenurs- difficulty in nurse staffing as to the meaning of ing department to make individual work plans the average number of hours per patient day in a relatively economical manner(23). deemed appropriate for a group of patients of The major difficulty with these appears to be in given illness severity in a given set of organiza- the variability of human needs. Sometimes these tional parameters. In an uncontrolled situation are expressed in the form of a request for excep- there is considerable daily variance in the amount .dons from agreed upon schedules. Other times they of nursing care required on a given ward. This is appear simpl) as absenteeism. The result isthat traceable to both illness severity and census varia- the schedule seems always to be in tra6sition and tion and introduces the possibility of a significant it never reaches a steady state. The Management ambiguity. The appropriate level may be perceived Monitoring Systems Demonstration(24)going on as that level which on the average is within the at Michigan indicates that for one demonstration working performance of a given set of personnel; hospital nursing department this problem is cen- that is, the level that can be sustained- by a work tral and disabling. It is simply not possible to pre- group under varying conditions over a reasonably dict the quantity of nurse manpower that will long period of time. appear on any given day, even though considerable Alternatively, the appropriate level may be per.. effort is expended by the nursing supervisory perceived as that amount necessary for the individual sonnel. patients on a given day. The average of the latter Warner of the Michigan group is working on a is likely to be smaller because it does not include new approach to the scheduling problem that will allowance for occupancy variation. Ifit were the make an optimal assignment by mathematical pro- chosen level, the variation would have to he con- graming methods based upon nursing personnel's trolled or there would be substantial quality de- own preferences as to work times(25,26).This cline and an increase in complaints from people may reduce some of the inherent causes of dissatis- who are responsible for giving care. Theconcep- faction with long range schedules, as well as drasti- tual difficulty is that in many people's minds the cally reduce the amount mf middle management two problems have never been separated. If one time involved asks a director of nurses what staffing levels she The second technical problem can be described thinks are required by her patients one cannot as an assignment problem. Given that a known or accept the answer at face value. It may contain closely predicted number of nursing personnel are allowances for occupancy fluctuation. available for distribution to a described inpatient census, one must assign them to nursing units in Technical Difficulties a way that maximizes the overall utility subject to certainconstraints. This problemhasreceived While feasible technology for nurse staffing exl3ts, considerable prior attention. As was noted above, feasible models for itisfar from perfect. There are two or jx;ssibly its solution exist, both in a three problems that the technology attacks, but manual and in a computer oriented mode. there are residual difficulties;in each area. Tile The manual modes appear to be uneconomical. They make large requirements on personnel and 'Smith, R. 1... unpublished paper. require considerable dexterity of the operator. In

100 92 RESEARCH ON NURSE STAFFING IN HOSPITALS theMcPherson Community Health Center,for according totheir severity, a scheduling system example, the director ofnurses handled theas- resulted in predictable numbers of nursing person- signment problems quite well. Her assistant, who nelreporting each day and quality evaluation was forced to do it2 days a week, had extreme schemes were available, it would become possible difficulty, as evidenced by less desirable solutions to evaluate the tradeoffs between transfers and ap- and much greater expenditures of time and effort. propriate staffing. The result of this evaluation A semiautomated solution, which uses the com- might be that fixed staffing for floors of stabilized puter to suggest an apparently optimal solution census and illness severityisthe cheapest possi- for review by the responsible nursing supervisor. ap- bility. In this case, the assignment model would pears to he the method of choice. One of the most be unnecessary. important technical difficultiesinthis approach Itisdifficultto evaluate the severity of the from the administrative and costviewpointis technologicaldifficulties. To he surethere are the specification of the objective function. This is opportunities for improving the technical models, the element that will determine the relative utility but the nature of the technology is such that this of various possible assignments and select the one is likely always to be the case, In the meantime that makes the greatest 'contribution. thereislittle evidence to support a conclusion There are a number of possibilities. One that that the presently available modelsAould not con- the Michigan group is attempting to explore would tribute to the control of nurse staffitig. attempt to use the floor supervisor's own preferences for additional staff. While this evaluation is too subjective to provide much improvement over Professional Difficulties present practice directly, if certain more objective predictors of that evaluation can be determined, Day to day work on the Management Monitoring Systems Demonstration with two community hospi- such as a Connor type severity of illness scale (/7), tals in Michigan leads the investigation team to the objective measuresC111 be usedas a proxy identify a problem concerning the ability and un- utility function for day-to-day assignments.' Even if the Michigan approach provides an im- derstanding of nursing supervision thatistoo provement in the assignment technology, another serious to leave unnoticed. The conceptual and technical issues involved in the problem of nurse portion of the objective function retnains for sys staffing ;kre obviously not simple. It has been rare teniatic exploration. The assignment model assumes for the project staff to encounter any significant that at least sonic fraction of the nursing personnel insight into these problems by nursing supervision are available for assignment on any two or more in our demonstration hospitals. nursing floors or units. It is well known that there Of course one would not expect expertise in are some costs involved intransfers or reassign- economics, work measurement, mathematical pro- ments. These costs appear to include attitudinal graining, or sociological analyses of decision making costs on the part of the employees, training cost in hospitals. On the other hand one might expect necessary to orient personnel for two or more work at least intuitive comprehension that there probably assignments, and quality costs involved in the lack is an appropriate level of care for each individual of knowledge of patients for a recently transferred person. patient, that the consumer or his agents have a These costs can he minimized by reducing the reasonable interest in the definition of what is ap- variation in number of patients and severity of propriate, andthatindividualvaiiations tend illness,which will have the 'resultof reducing to average out when patients are taken as groups. the number of transfers required. Thus, stabilizing In each of these situations we have encountered the occupancy level of the floor and utilizing dif- what appears to he an absence of understanding. ferent floors for different purposes in the progres- In both of our demonstrations therehas been sive patient care manner seem to he desirable. If serious (inestion raised by both research staff and a situation existed where occupancy were reason- administrative staff as to the capability of nursing ably stable, patients were assigned to patient units supervision to deal effectively with improved con- Trivedi. V.. dnctoral dissertation, in progress, trol of staffing. At the very least it appears necessary

101 IMPACT OF ADMINISTRATIVE AND COST FACTORS ON NURSE STAFFING 93 either to begin rather extensive and fundamental form of training. It is impossible to tell how wide- retraining or to supplement the nursing super- spread this problem is, but thereis nothing to visory structure with managers with a different justify optimism.

Validity of the Economic Bargaining Yodel

The implementation of the engineering model apparently have not resisted any of the numerous alone is that carefully developed measurement tech- solutions that seem to be reflected in the variety of niques would yield consistent values for a given current practices. severity of illness, and these would be adapted The problem presented by this conclusion is that smoothly and uniformly by the hospitals across there seems to be conflict with the normative values the United States. The only cause of residual varia- about the appropriate rolesof the bargaining tion after a reasonable period of time would be parties. One assumes that trustees and adminis- variations in severity of illness. Since individual trators should be concerned with costs, that doctors hospitals each admit several thousand' patients each and nurses should be concerned with quality, and year and make their admissions tinder atleast that the four of them will bargain in a more or less roughly comparable systems,differences inthe effective manner. This may be where the problem average severity of illness would tend to be quite lies in the control of nurse staffing. In fact, trustees small as would the differences in staffing patterns. evidence little interest in the control of hospital This obviously is not consistent with the facts. costs, often do not have the proper information The economic bargaining model brings several necessary to show interest in costs (27), and cer- additionalpossible explanations into play, but tainly cannot be described as having been effective one would expect it to lead to uniformity, at least in the control of costs. among economically similar situations. Although The nature of the prevailing cost reimbursement evidence for this is not fully explored, the outlook contracts and the sharing of costs through third is distinctly unpromising. Short of abandoning the party coverage provides little direct incentive for economic bargaining model (a step not to be taken cost control. The record of continuing inflation lightly),the most reasonable conclusionisthat speaks for itself. So, unfortunately, do the planning hospitals are operating in the special case where no decisions of hospital trustees. There has been no utility consensus results from the bargaining proc- tendency toward improved occupancy of American ess. More specifically, the bargaining parties per- hospitals, little activity toward mergers, and fre- ceive the marginal utility of nurses as very indeter- quently documented cases of almost deliberate minant or diffic dt to estimate. competition and over expansion of hospital facil- Thus they have allowed nurse staffing to wander ities. These decisions are very largely within the in an essentially uncontrolled manner. The rea- realm of trustee authority. They imply that, at sons for this appear to lie in the conceptual dif- least up until the very recent months, control of ficulties,particularly the lack of anyDbjective costs has not been a primary activity of hospital measurement of utility and in the technical and trustees. professional problems. If, however, either party in A less important factor in the attitude of the the bargaining process had strong convictions, a bargaining parties may have been the prevailing consensus would develop around those convictions. belief about the existence of a "nursing shortage." No such movement is perceptible. Neither nurses If all the parties approached the bargaining table nor doctors, for example, have come forward with with the attitude that no matter what solution is demands asto the tequired quantity of nurse reached for nurse staffingitisinadequate, no staffing,s and hospital administrators and trustees serious bargaining will result. Not much has been heard lately about the nursing shortage. Both of '"The American Nurses' Association does not recommend thedemonstrationhospitalshaveestablished any specific formula, ratio, or numerical concept which cotild

be applied generally. . ." American Nurses' Association, RequirementsforInpatientHealth Car' Services." The Committee on Nursing Services, "Statement on Nurse Staff Association: New York, March I,1967. 94 RESEARCH ON NURSE STAFFING IN HOSPITALS guidelines as to Cie appropriate average nursing east Michigan have also commented that they no hours per patient day and both appear to be expect- longer have chronic vacancies in nursing. There is ing their nursing department to stay at or below the no way of knowing whether this is also true in guideline. Other hospital administrators in south- other States.

Some Speculations on the Future

If the model and the facts are not misleading. example, requires that hospitals publish informa- one can infer that the question of nurse staffing tion about their rates and costs. has been a nonissue in most community hospitals. One occasionally encounters activity at the State The author is not willing to predict that this will level that appears to be in the exactly opposite change. Although there has been a great deal of direction. A bill was introduced in the Michigan public "viewing with alarm" of hospital costs, there legislature in 1971 that would have effectively re- has been relatively little serious activity aimed at moved all control of either costs, quality, or utili- their control. One must look to the political arena zation from Blue CrossinMichigan!' In one for firm signs of such action. The evidence is mixed version it specified that Michigan Blue Cross was at the Federal level and, on the balance, negative obligatedto pay audited costs of any licensed at the State level. hospital. The bill was eventually passed-in a less Federal action, for example, could have effec- drastic form at the same time as a bill to provide tively forced national franchising:it could have franchising orcertificationof needlegislation. moved more strongly against the economic posi- Fortunately, the latter act restores many consumer tion of pathologists and radiologists: it could have protections the former destroys. pasitd the so-called Bennett Amendment.'for pro- The trend overall seems in the direction of in- fessional peer review or even something stronger creased control of costs. Itis easy to predict the than the Bennett Amendment; and it could have result from the economic bargaining model if this moved much more strongly towards prospective trend solidifies. Increases in nurse staffing will be rating for Medicare payments and uniform na- strongly resisted and hospitals will begin to notice tional payment schemes for Medicaid. As of this the range of behavior within their own regions. writing none of these things have come about. The This will 'lead administrators and trustees to raise clearest sign has been the 6 percent limit on price questions about the staffing levels in all hospitals increases imposed by the Price Commission. Even except those at the bottom of the range. there, there is a provision for exceptions. It will be To the extent this occurs,itlooks as though several months beforeitis dear how firm- that_nurses will be unprepared. In the absence of a guideline is to be. The permanence of the Com- quality measure they will be forced to rely upon mission is also a question. arguments based upon the subjective assessments of At the State level, it would appear that a small utility and the existence of the "nursing shortage." minority of States feel the cost of hospital care The evidence of the range of current performance to be a central political issue. After some years will he used against them in the subjective quality experience in New York withtheMetcalf-Mc- arguments. They may do better with the arguments Closky Act. franchising or certification of need of about the nursing shortage, but itis likely that hospital e-pansion has been successfully introduced people will begin to say that if the shortage has in only a handful of other States. Insurance com- existed for all these many years, perhaps it isn't missioners who might insist on more control of as bad as we initially thought it was. costs through third party payment mechanisms have In any case, a move toward tighter cost control generally not done so. Dennenberg of Pennsylvania will result in considerable pressure to install and and a few others have attracted publicity partially improve thetechnology. Thus interestinthe because they are the exception rather than the modest for nurse sc eduling and nurse assignment rule. Steps toward public utility type regulation of hospitals have been minimal. Only California, for 'H. B. 4030, Michigan Legislature. 1971.

103 IMPACT OF ADMINISTRATIVE AND COST FACTORS ON NURSE STAFFING 95 will continue and probably increase. So will in- component of hospital costs. The resulting atten terest in the evaluation of the organizational para- tion will lead to calls for improved technology, meters, including the appropriate roles for various better defense of present practices, and evaluation skill levels, supervisory structures and service unit ofproposalsfor improved organizational para- management, and related proposals. The evaluation meters. of all these, both organizational parameters and In each of these areas the central itembe- technologicalschemes,willdependonbetter comes the development of measures of quality of quality instruments as will the defense of present actual nursing performance. Advances on this front or increased levels of nurse staffing. will clarify nursing roles, aid in the evaluation of The conclusion seems to be that although the organizational parameters, permit the improvement impaCt of administrative and cost factors on nurse of scheduling technology, and improve the knowl- staffing is at the moment relatively insignificant, edge of allparties to the bargaining processes. there is quitc a powerful chain reaction that might In the author's opinion itis the most critical of occur if there is a basic change in the attitude the various opportunities and is worth substantial of the American people toward the cost of hospital Federal support. A 5-year multidisciplinary and care. Given attention to this itis inevitable that multiuniversity effort measured in millions of dol- there will be attention to nurse staffing,iffor lars each year would be entirely appropriate to no other reason than that it is the largest single national health care needs.

104 96 RESEARCH ON NURSE STAFFING IN HOSPITALS

References

(1) Scott, W. R."ProfessionalsinHospitals: (12) Yett, D. E. "The Chronic 'Shortage' of Nurses. Technology and the Organization of Work." A Public Policy Dilemma." In Klarman, Organization Research on Health Institu. H., Empirical Studies in Health Econom- lions, Georgopoulos, B. R., (ed).Institute ics. Johns Hopkins Press, Baltimore, Mary- of Social Research, University of Michigan, land, 1970. p. 357 ff. Ann Arbor, Michigan, 1972. (13) Carr, W.J.,andFeldstein,P.J. "The (2) Trites, D. K.,, et al. "Influence of Nursing Relationship of CosttoHospitalSize." Unit Design on the Activity and Subjec- Inquiry, vol. IV, no. 3, September 1968, p. tive Feeling of Nursing Personnel." Final 48 ff. Research Report, Rochester, Minnesota,un- (14) Ingbar, M. L.,etal."Differences in the dated. Cost of Nursing Service: A Statistical Study (3)Freeman, J.R. "Quantitative Criteriafor of Community Hospitals in Massachusetts." Hospital Inpatient Nursing Unit Design." American Journal of Public Health, 56, 10 Doctoral Dissertation, School of Industrial (October 1966), pp. 1708-1709. Engineeting, Georgia Institute of Technol- (15) Griffith,J.R.,etal.The McPherson ogy, Atlanta, Georgia, 1967. Experiment. The Bureau of Hospital Art. (4) Revans, R. W. "Morale ;IndEffectiveness ministration, University of Michigan, Ann ofGeneralHospitals."Mclachlan,(ed.) Arbor, Michigan, p. 179 ff. Problem; and Progress inMedical Care, (16) Medinet Corporation. "Nuriing Staff Allo- Oxford University Press,for the Nuffield cation." 100GalenStreet,Watertown, Provincial Hospitals Trust, 1964. Massachusetts. 1969. (5)Pfefferkorn,B., and Rovetta, C.A. "Ad- (17) Edgecumbe, R.H. "How CASH Helps ministrative Cost Analysis for Nursing Serv- Nurses ImprovePatientCare." Modern ice and Nursing Education." American Hos- Hospital, 106, May 1966. pp. 97-99. pital Association and National League for (18) Jelink,R.C.:Munson, F.; Smith R.L. Nursing, Chicago, Illinois, 1940. ServiceUnit Management: An Organiza- (6) --"Departmental Indicators." Hospital Ad- tional Approach to Improved Patient Care. ministrativeServices,AmericanHospital The W. K.Kellogg Foundation,Battle Association, Chicago, Illinois. Creek, Michigan, 1971. (7)Connor, R. J. "A Hospital Inpatient Classi- (19) Perrow, C. "Interorganizational Research in fication System." Unpublished Doctoral Dis- the Fieid of HealthA Radical Criticism." sertation, Johns Hopkins University, Bald- Paper prepared for the Johns Hopkins Con- more, Maryland, 1968. ference on Interorganization Researchin (8)Jelinek, R. C. "Nursing, the Development Health, sponsored by DHEW, October 1972. of an Activity Model." Unpublished Doc- (Mimeographed.) toral Dissertation, University of Michigan, (20)Donabedian, A. "Evaluation of the Quality Ann Arbor, Michigan, 1964. of Medical Care." Milbank Memorial Fund (9)Warner, D. M. "A Two Phase Model for Quarterly, vol. XLIV, no. 3 (July1966). Scheduling Nursing Personnel inaHos- Part 2. pp. 166-206. pital." Unpublished Doctoral Dissertation, (21) ProfessionalActivityStudy.The elm. Tu lane University, 1971. mission on Professional and Hospital 'Activi- (10) Commission for Administrative Services in ties, Ann Arbor, Michigan. Hospitals. Nursing Procedure Manual. The (22) c.f. Committee for Administrative Services Commission. Los Angeles, California, 1968. in Hospitals. A Quality Control Plan for (11) Taylor, F. W. The 'Principles of Scientific Nursing, Los Angeles, California,1965. A Management. Harper and Sons, New York number of modifications of this approach 1911. are in existence.

105 IMPACT OF ADMINISTRATIVE AND COST FACTORS ON NURSE STAFFING 97

(23) Morrish,R.A.,andO'Connor,A.R. Arbor, Michigan, contributed paper: The "Cyclic Scheduling." Hospitals, MHA, vol. Operations Research Societyof America, 44, February 16, 1970, p. 67 ff. There are a 41st National Meeting, April 28, 1972. number of similar publications. (26) Sanders, J. L. "The Scheduling of Nursing (24) HS 00228. "Demonstration of aHospital Units forSmallHospitals."Contributed Management Monitoring System." Granr by paper: The Operations Research Society of the National Center forHealth Services America, 41st National Meeting, April 28, Researchtothe University of Michigan, 1972. Program and Bureau of Hospital Adminis- (271 Neuhauser; D. "The Relationship Between tration, Ann Arbor, Michigan, J. R. Griffith, Administrative Activities and Hospital Per- Project Director, 1969-73. formance." Center for Health Administra- (25) Warner, D. M., Bureau of Hospital Admin- tion Studies, Research Series 28, University istration, The University of Michigan, Ann of Chicago, Chicago, Illinois, 1971.

106 DISCUSSION OF MR. GRIFFITH'S PAPER

Discussion Leaders Miss Ruby M. Martin Dr. Frank A. Sloan Director of Nursing Services Department of Economics Ohio State University Hospital University of Florida Columbus, Ohio Gainesville, Florida

Miss Martin

I think Mr. Griffith has left us with the im- nursing unit without any thought as ..t) what they pression that nurses spend the bulk of their time are doing. As Iunderstand from our personnel worrying about how much they are going to be department, these are actual hours worked or paid paid in relation to all the people working under for without any thought of who the people are them, and that we really care little about overall or what they do as long as they are assigned to costs. I would have to disagree with him. We are a nursing unit. This can be a very misleading very concerned about costs. As a group we do not figure, and we ought to take that into considera- spend any more time on these kinds of concerns, tion. All hospitals do not necessarily assign person- about what others are going to be paid in tLe nel to cost centers using the same rationale. nursing hierarchy, than administrators spend dis- I agree with Mr. Griffith when he says weced cussing the administrator hierarchy salaries. This to beginI would say continuean extcnsive and is not something about which nursing alone ex- fundamental retraining of our nursing supeisory presses concern. people. There are many instances where thisis Dr. Aydelotte mentioned a void in the literature being done and being clone very well. If you look at she has reviewed that deals with the administrative how people have been assignedto supervisory. and costfactors in nurse staffing. Traditionally positioas in hospitals -traditionally, you will find there has been a void regarding this in the educa- they have been chosen because they have been tion of professionals, both nurses and doctors. The there the longest and not because they have had emphasis, of course, has been on educating the any particular preparatior to assume this respon- professional person in the clinical practice of nurs- sibility.If you look at the majority of hospitals ing or medicine, and more attention needs to be in our country, those around the 100-bed size, this given to the financial aspects of health care in is still the way these people are being chosen. We programs of formal preparation. Perhaps Muriel need to provide opportunity for these people to Poulin would like to talk about this as far as her learn to function as managers of nursing rather programofnursingadministrationatBoston than to criticize them for not functioning in this University is concerned. role. In the HAS program, the American Hospital In his paper Nfr. Griffith does not suggest that Association designed the format for the figures we do this or that we provide managers with a that are reported there. These figures represent different form of background and training. We the manhours paid for all those assigned to a are only muddying the water, and we have had

98 107 DISCUSSION OF MR. GRIFFITH'S PAPER quite a lot of this with all the people added to cation becomes more difficult when there are more the structures within the hospitals,Iseeit even people with whom one must communicate to get with the Unit \tanager System in some instances. tliv job done.Iant sure there are many nurses Instead of relieving nurses of duties, it adds time in the room who will want to speak on this later. to their workload in some other way. Communi-

Dr. Sloan

I was glad to sec that we are discussing papers criteria; there are numerous different instruments. on implementation atthis conference. Yesterday Results 'differ inlarge part because instruments we discussed the very important problems of what differ.I certainly would not like to see this kind should be done or how yon could start from scratch of research conducted at the cost of understanding to staff a hospital. The problem is that one can how hospitals actually. behave, especially with re- devise many schemes on paper, but things may gard to how they select and use inputs for the not work that way, which brings us to the question production of care. of implementation. The third point concerns the discussion in the We face parallels elsewhere in the health sector. paper of different types of staffing models. There For instance, there has been much talkabout is an assessment of whether hospitals really behave HMO's and how desirable they are, but they will as ifthey were implementing this engineering not be successful if producers and consumers don't model, and this notion is rejected. Likewise there like them. In other words, one may design some- isa consideration of an economic model where thing very nice and find outitis not desirable in some sense some administrator or trustee evalu- for less "objective" or "quantifiable" reasons. ates the utility of the service and weighsthis I was also glad to see that you made the follow- against the cost of providing that service. ing considerations explicit, even though many if I would argue with the author about both of not most of us are generally aware of these matters. these models. I think he rejected them too quickly. The first pertains to substitutions. There are sub- He takes an unduly pessimistic st.-nce, and I shall stitutions of various types of nursing personnel discuss that in a moment. that we might want to consider. We have RN's but There is the point that the third party payer we also have other categories. We have capital has to be considered, the point that most hospital that may be in some limited sense substituted reimbursement (or at least a considerable percent- for labor. We actually discussed that very briefly age) iscost-plus, and as long as you have cost- yesterday. plus reimbursement there is no incentive on the Second, with respect to staffingratio,hii;her part of trustee or administrator to look at cost ratios of nurses to patients yield (holding other considerations. There is a very widespread use of factors constant) a product that is of greater value cost-plus in the hospital industry, and, of course, to patients. If one increases the ratio one gets more, itis not the only industry that uses it. Aerospace but there isa question of how much more one is a good example; biomedical research is another. can get. There is no particular ratio of nursing Hut hospitals are one example in which some per hour or nursing hours per patientthatis behavior that we observe reflects very much the appropriate. Rather, there are a number of these fact that we do have cost-plus reimbursement ar- numbers, and the question is. What are they worth? rangements. And when we look at how staffing takes This was adequately discussed in the formal presen- place we should remember that we do have to take tation. that into account. Reimbursement is not something This,ofcourse,getsintothequestionof toleave to the economists and staffing to the assessment of what a nurseisproducing. The sociologists and radial units to the architects. We same issueisfoundinthe area of education must work together. planning where evaluators attempt to gauge effec- This paper has stated that we do have to worry tiveness. In education thereare different success about implementation, that there are sortie patterns

108 100 RESEARCH ON NUltsE STAFFING IN HOSPITALS in staffing that are not explained. I think we have treated differently, and I do not know that I be- to develop formal model., a hit more, ones that lieveit,but thisisthe hypothesis thatisstig- are testable with data. The problem in any kind gestetlthat we should be wondering about the of empirical work is that if you build an inaccur- ownership in addition to the reimbursement ar- ate model or you do not build a complete model, rangement. you always end up accepting the null hypothe- In addition, wehave to consider that there are sis. Here the null hypothesis is that hospitals do differences in consumer incomes. This is. brought not behave in a systematic way with respect to out in the paper. If you go to a county hospital in staffing. If you do not build a complete model, California, Los Angeles County or Cook County, you are always going to have to accept tt at hy- thereis going to be a variation in staffing and pothesis. It is not an easy hypothesis to accept. there is going to he a variation in the consumer's The author does cite some work by Ingbar and ability to pay. There is nothing irrational about Taylor that seems to indicate that staffing patterns it.It may he objectionable in terms of what our reflect the type of service, the case mix, and so social objectives are, but itis not irrational. There forth. In fact, 30 to 40 percent of the variance in will also be different staffing patterns, presumably, staffing is explained by hospital type. The greater if there are great variations in the wages different challenge lies in explaining the remainder of the kinds of nurses are paid. The question of different variance. kinds of licensure arrangements has been brought For instance, we should look at hospital owner- up. There are various types of bargaining situa ship. An economist colleagueI do not neces- tions, in fact there are a host of factors I think sarily subscribe to his researchhas hypothesized should he investigated. They should be investi- that proprietory hospitals would he much more gated primarily so that we may have an idea of consistent intheir staffing patterns than would what kind of staffingisdesirable. We have to voluntary hospitals. His reasoning is that after all, find out why hospitals are not staffing the way we proprietory hospitals are subject to some sort of want them to and determine how we might get profit motive and there exist prices of these various them to go in the right direction. inputs and they somehow maximize something and In closing,I applaud Mr. Griffith for getting come up with a unique set of inputs that they into some very interesting questions. His model want to impletnen. The voluntary hospital faces was not specific enough for the conclusion that so much constraint that therefore you find all we do not see staffing as a random or nonsystematic this variation. process. We should have fur'.her research in this There are problems with that, The differences area with a view toward implementing the staffing in case mix and different kinds of patients are requirements that may be desirable,

109 Nursing Directors as Participants in Hospital Peer Review Boards

Dr. Gerrit Wolf Assistant Professor of Administration Sciences and Psychology Yale University New Haven, Connecticut

Introduction

Nursing personnel comprises one of the largest elsewhere,1 is that of monetary incentive for better departments in a hospital, and next to medical than budgetary performance. These are only two departments, the hospital's most central. In this of the many possibilities, but these are of interest highly important department, the quality of patient tothis researcher because of their use of inter- care is of paramount importance. This may be due personal communication (2)and decisionmak- to the training and interest on the part of nursing ing (3). personnel and supervisors (1)as well astode- Let me illustrate with some problems. Place a mands by consumers. director of nursing on a budgetary review board However, the consumers and payers of health of peer administrators and what happens? What servicesstrain under present hospital costs and kind of impact does the director have on a nursing desire some cost control. Any program incost budget and what is the effect relative to the influ- containment in a hospital must involve nursing ence of "peer" administrators? How does the peer because of its size as a cost center. However, because review process operate in relation to cost contain- cost control is not of the highest importance to nurs- ment and patient care? ing,any cost control program must provide a The peer review notion assumes that a nursing means of satisfyingnursing's needs of offering director is a peer of the head administrator and quality patient care and service. controller. The director, like all managers, must One wq). of trying to contain costs and maintain 'schedule, supervise, control, plan, and encourage. quality patient care is to use a combination of What kinds of interests and skills does the nursing the carrot and the stick, The stick we report here director have to do this, and does peer review is that of peer review in connection with the use ofengineering standards. The carrot,detailed Elnicki, Richard, personal communication, 1971.

101 102 RESEARC:11 ON NURSE STAFFING IN HOSPITALS improve them? If the administrative knowledv in- We report 'here aspects of a 3.year experiment creases,is this seen in relationships with subordi- in cost containment using peer review, engineering nates and in relation to cost and performance? standards. and monetary incentives. We explore Answers to these questions may vary with the the above questions about the nurse in;, peer size of hospital, the social and technical opera- review setting. The important organizing distinction of the review board, the social and technical tionsaretechnicalfeatures outside and inside environment of the board. and the background of hoard meetings and organizationsl problems out the nurse, such as the nurse's training and expe- side and inside the meetings. The next section rience. That is to say that organizational, group, describes the setting and design of the peer review environmental, and individual variablesmayaffect experiment. Sequel sections report results relevant the nursing director's performance and learning to nursing directors for the 3 years of the experi- in peer review. molt.

The Experimental Design

Half of the 3ti hospitals in a small State "vol- side of the table and each of three review hospi- unteered" to participate in an experiment in cost tals queued up along the other three sides with containment by peer review boards. The 18 hos- each hospital staying together as a group. This pitals not in the experiment were used as a control procedure tended to structure the situation as a groupfor economic analyses and comparisons, court of inquiry. If department heads participated, reportedelsewhere.Some wishedtopartici- they, waiting like witnesses, sat around the pe- pate, but others were persuadedtoparticipate riphery of .he room. There was no discussion of as necessary for the sake of the health industry. alternative designs, such as not sitting by groups The 18 experimental hospitals were divided into or constructing circular tables. three setssmall. medium, and largeof six each. Small hospitals had fewer than 100 beds, medium hospitals, 100-250 beds and large hospitals, 250- Subjects 1,000 beds. Each of these three sets was divided in half. From each of the hospitals in this final clas- The board consisted of nine members: three sification,peer review boards were constructed. administrators, three controllers, two nurses, and a During the firstyear, each hoard reviewed four trustee. This means that each hospital contributed departmental budgets (nrsing, medical records, an administrator and a controller and that two of housekeeping, laundry-linen) of each of the other the three hospitals contributed a director of nurs- three hospitals of similar size. The planned life ing and the third contributed a trustee. Most of of the experiment was 3 years with expansion in the members knew each other before the experi- the last 2 years for a greater number of depart-. ment throughtheirprofessions. The larger the meets to come under review. hospital, the longer the tenure in office of each person and, consequently, the more known and respected he was. Because of the division by hos- Setting pital size, board members were still relatively peers. Controllers usually sat next to administrators A meeting of a hoard took place at the hospital from the same hospital, facilitating consultation be- under review and lasted a hill clay. All IS reviews tween them. This structure changed slightlyin took place during the last 3 months of the calendar year, which were thefirst mouths of the next later meetings with nurses, the least experienced fiscal year. Meeting rooms varied from a spartan in budgeting, tending to sittogether. The staff hall with assembled table and chairs toplash coordinator for the group (described in a moment) board rooms. Lunch and coffee break snackswere acting as secretary- consultant, was located near the part of the routine at all meetings, target hospital ill small and medium size hospital The target hospital lineditself up along one meetings, The large hospital boards found the co- NURSING DIRECTORS AS PARTICIPANTS IN HOSPITAL. PEER REVIEW BOARDS 103 ordinator, acting as titanium', sitting opposite the with all other hospitals in the State, for each de- target hospital. imminent. This report is a standard report gener- The chairman was to he selected at each meet. ated by a national association. The second report ing. In the small hospital hoards, the coordinator was a biography of the hospital'soperation. This preferrednotto he chairman.(*OBI la IN tothe included kinds of nursing care and bed comple- board's wishes, and an administrator was elected. ment, or how housekeeping handles the cleaning The full responsibility of the chairmanship became and janitorial jobs. The last report was a systems apparent when the coordinatorfeltthehoard engineering report on each department specifying should accept the responsibility for informing the the number of personnel needed to do the job in target hospital of the board's decision on the first each department. These data were generated by department. This procedure apparently went a long having an industrial engineer visit each depart- way to create connnitment and responsibility by ment, interview various personnel, and then have this group for its decisions. the engineering firin convert the data based on Inthe boards of medium sizehospitals,the standard formulae into personnel requirements. coordinator took astrictlyadvisoryrole. The boards tried functioning without a chairman at Instructions thefirstmeeting. However,itwas apparent to them thata chairmanless group floundered. A Boards were instructed to use any of the infor- controller was elected chairmanfor subsequent mation to come up with a decision concerning meetings, and this responsibility generated more thebudget. The board had theauthorityto extensive preparation on his part. change any part of the budget they saw fit that The coordinator of a large hospital took the did not stay within reasonable historical practices role of chairman, which satisfied his own inclina- of the hospital. The board's judgments were to be tion to get things done and relieved thegroup's made with the hospital under review removed anxiety about what to do. This violation of the from the meeting. Also, the target hospital did intended functioning had its positive and nega- not have the right to bargain, debate, or change tiveeffects. The coordinator positively led with the board's decision, However, it had the right priming questions about the budget, trying to get and responsibility of informing the board of jus- the board to take an active part, while the group tification for the submitted budget. negatively became dependent on thecoordina- The board was not told how to organize other tor le.. leadership. than that a chairman was to be elected from the board at each of the meetings. The board Task Environment could use the staff coordinator as it wanted.

Each board was supplied a staff member, the Incentives coordinator, who helped construct budgets and gather together correlative data. The budget itself, A monetary reason for participating was that a thefirstyear, was constructed for each of four hospital would be reimbursed for performing bet- departments:nursing, medicalrecords,laundry. ter than its budget. Thisreimbursement would be linen, and housekeeping. In the second year ad- distributed at the end of each year andsupplied ministration, pharmacy, and operation of plant by the Social Security Administration andBlue were added. For each department, salary, nonsal- Cross. However, once the experiment started, im- ary, and depreciation figures were generated,line- mediate problems of planning and learning about by-line, according to a standardized recording sys- management took on an autonomy oftheir own. tem. Data sheets reported the expense budget for The opportunity to learn was an inducementalso, the previous year, the actual results of the pre- although the need to learn how to control costs ceding year, and the expense budget for the cam- may have been driven by a fearthat "outsiders" ing year. might force something to be done to control costs. Correlative and supplementary data were of These observations stemmed frominterviews three kinds. One was quartile expense comparisons with board members. Attendance at the meetings

11Z 104 RESEARCH ON NURSE STAFFING IN HOSPITALS yielded some further ideas on motivations after of the experienced hospitals was thicker, food bet- the experiment started. These were desires to learn ter, and visitation of new wings occurred more and desires to look good in front of peers. often. Relevant to the desire to learn...small hospitals This looking-good motive carried through in historically had done little specific breakdown budg- constructing a budget. There appeared to be a de- eting.Participating forced them to budget and sire to construct a tighter budget than that of a learn how tobudgeteffectively.Althoughthe similar hospital. This desire could conflict with large hospital hail budgeted for years, its budget the monetary incentive, because the tighter the had grown sothatthe experiment provided a budget, the less likely the possibility of receiving means for sorting out budgetary complexities. a monetary reward. Also, the turnover in administrative personnel in the experiment's 18 hospitals had produced a Data Collection situation where about one-third of the members were relatively new to their jobs. In these cases, Three kinds of data were collected. 'These were the experiment provided a means forlearning questionnaire, content analysis of audiotapes of all about their own hospital's operation as wellas board meetings, and personal observations and getting some professional feedback to their initial interview. Questionnaires were sent to the 54 par- decision. ticipants before and after the set of meetings each Taking a look at self-presentation motivation, year. Fewer than half the participants responded. we found that for those who had been in their All meetings were tape recorded, and observation positions for a considerable time the meetings pro- occurred in more than two-thirds. Details of ques- vided a means to show off theirfacilities and tionnaires,coiling systems and analyses can be personal skills. The carpeting it the meeting rooms found elsewhere (4,5).

Results

One way ofclassifyingthesocial-psychologi- the first year. Except for the small site hospitals, cal results is to describe the social and technical all hospitals had budgeted using State hospital problems inside and outside the board meetings. association guidelines. However, there were still Where appropriate, we note differences between differences between hospitals on allocation of costs different size hospitals and nurses with different to departments. Even within a hospital, budgetary backgl.ounds. We do not report the details of the procedures changed annually by small degrees. economic results other than to note that as of this The engineers had to hustle to get data on the time, the .first year resulted in no differences be- .frequency of microtasks performed in each depart- tween the experimental and the control group hos- ment. Often the hospital personnel had to supply pitals in their actual financial performance.2 data that were hard to obtain or were inaccurately reported. The result was that engineering stand- Technical Problems Outside the Board ards lacked credibility. Meetings Finally, the development of rules for reimbursement was delayed until the end of the first year Before the boards could meet, budgets that were when they were to be actually used. This procedure comparable and consistent between hospitals had raised decision making problems for the boards, to be constructed. Engineering standards needed to which is taken up in a subsequent section. be set, and rules for reimbursement were called for By the second year, the first two problems of in sonic detail. uniform budgets and complete engineering data Because of the shortage of time and experience, had been, for the most part, corrected. However, these three problems were handled relatively poorly rules for reimbursement had not been developed and were not considered until after the second ' F.Inicki,Richard, personal communication, 1971. year's review board meetings. When they were 113 NURSING MRECTORS AS PARTICIPANTS IN tlosPITAI, PEER REVIEW BOARDS 105 considered, they were treated differently by differ- was used. This course of action was of leasteffort ent hospitals. for them. The small hospitals decided to use a step vari- This mem that between hospitals, comparisons able procedure that called for adjustment of budg- were difficult to make, and withinhospitals, com- ets to volume, not on a linear basis but on a step parison to engineering Aandards was rejected. The basis. For example. for every x-tousand patient reason for the latter can be understood bylooking days' deviation from the projected patient days. again at the nursing department as an example. staffings wo,.1d be adjusted by one FTE. Rewards The engineersconstructedstaffingstandards would then be based on ajustified adjusted based on national norms for microlevel tasks. The budget. nursing department supplied information on how The medium size group talked at length about often each microlevel task had to be performed. using a form of step variable but was unable to The engineers then cranked the frequency of tasks come to agreement. The large hospitals decided and time per task through a formula to come there was no way to adjust budgets and, therefore, up with a standard. This was a classicalindus- gave no rewards or penalties. Some thought there trial engineering approach of time and motion. should be no adjustment for volume. The argu- The problems with time and motion studies are ment used was that a household does not change well known and these problems occurred in this its income and budget for olume but instead setting. The problems were as follows: tightens its kit. Others thought there should be adjustments but people could not agree and said The shortness of the study gave the nursing the problem was too complex. directorlittle help in determining how to The nursing department exemplifies these com- change and implement the staffing change if plexities. Staffing depends on the available part- the engineering standards, calledfor fewer time and full-time employees in different grade nurses. levels, the number of nursing units, and the opportunity to switch personnel from unit to unit The time pressure to get data forced the nurs- on a daily or hourly' basis. It also depends on the ing employees who filled out the frequency of variability of patient demand. With ample part- task questionnaire to fill in fictitious data. time help and relatively continuous units, staffing There was no one engineer regularly available can adjust to volume if thedirector has the skills to help the nurses collect the data and answer to do the juggling act. On the other hand, if a questions. hospital had a few part timers, separate units and stable demand, clear step function adjustments The final staffing standard was not explained could be made. These two extremes made it diffi- to nursing's satisfaction; the engineerssaid the cult for participants, including nursing directors, formulas were too complicated and their ex- to agree on reimbursement and budget adjust- pertise would have to be trusted instead. ment. The standard was a single number with no confidence interval. Neither the nurse nor the Technical Problems Within the Board hoard could determine how far fromthe Meetings standard was really acceptable or unacceptable.

This area includes the kind of data supplied to In summary, had data, hidden procedures. and the hoard, the form that they were in, and the unaided involvement produced arelative rejec- rules for how the data were to be used. The forms tion of the engineering standards by the nurses for data reporting were literally hard to read and and the boards. The hoards'suspicion of the had too much raw data without useful summary standards led them to using the standards only statistics. Also, the use of data, particularly if con- when it was to their benefit. One nurse said, "I tradictory, was left to the boards to work out. The don't believe the standards, but ifit makes my result was that complex data were ignored, and department look good, I'll push the board to use simple decision rule of normal historical increases them so that my budget is accepted." 114 106 RESEARCH ON NURSE STAFFING IN HOSPITALS

Organizational Problems Within the Board with their own hospitals. The general tone was Meetings one of trying to understand the nursing department. This type of strategy was not used as much, The concepts of intergroup relationship and if at all, in the other departments. This may have decision process became clear during the first year been because directors of other departments were as useful for describing peer review. We had ex- usually not present. pected that concepts such as conflict and influence The reasons given for increased costs in nursing between peers might be more important (6).In were salary increases or increased staffing. The lat- the first yearthe strong difficulties of technical ter created confusion when the hospital budgeted problems inhibited these phenomena from clearly for unfilled positions that it really did not expect emerging. Instead,allparticipants were in the tofill.It seemed that nurses would set budgets same boat of grappling for understanding. With according to an idea that was not ever possible no one knowing the rules of the ,game, conflict to attain. The board feared cutting the vacancies could not appear. in case the nurse might actually be able to hire. The decision process for the review of a partic- There was time spent discussing the number of ular budget followed a three phase sequence of nursing hours per patient day and the staffing mix. search, analysis, and decision. These parts have These areas seemed to be ones of hospital philos- previously been noted as essential components of ophy and generated much managerial discussion. a group decision process (7,8,9,10). Search includes The total time spent on analysis of a nursing learning, concept formation, and information gen- budget was 47 minutes on the average. eration.Analysis encompasses problemsolving, The decision phase took about 15 minutes and comparison and contrast, and evaluation. Decision no budget was cut. Some budgets were tabbed for involves the acceptance, rejection, or change of the a further look at reward time for what to do with budget. unfilledpositions(vacancies). The nurses were The intergroup relationship was seen as one not as active in the decisicro phase as in the pre- between a review board and a hospital. This re- vious two phases. In ratings of who contributed lationship was viewed as consisting of the co- to the group's decision, controllers rated highest, occurrence of strategies by board and hospital. administrators next, and nurses and trustees least. The strategies and relationships were thought of The second year, a fixed set of strategies for as varying with the search, analysis, or decision each phase was selected by the researcherfor stage of the decision process. measurement. Participants, acting in the role of The strategies,as measured on the question- board members and then in the role of hospital naire the first year, were not evident to the partic- members, rated their preferences before the meet- ipants before the meetingsstarted, Because of ing for the kinds of strategies preferred. The ob- this uncertainty, the nurses met in a preliminary servers rated the actual use of strategies. Table 1 session to acquaint themselves with the data and reports strategies for each phase and the inter- possible selection of strategies. We observed the coder reliabilities. From these data, we look at the following strategies at the meetings. During the intergroup problem and the influence of nurses. search phase, the board was passive and the hos- Across all budgets we found no differences be- pital made a presentation of the budget. The tween the hoard and the hospital in their prefer- analysis stage usually found the board proceeding ences for joint strategies. This held up forall through a budget line-by-line.Also, much time three decision phases. The key finding, instead, was spent on managerial discussion. The decision was that during the search and analysis stages phase usually saw some discussion and consensus the participants' strategy preferences were nega- on a decision. tively correlated with their actual strategies per- The nursing budget was presented with percent- formed. The negative correlation was highest for age . breakdown for subunits and changes from controllers. During the decision phase the corre- previousyears.Often,background information lation was positive, with the nurses' preferences was supplied by the director on the wage strut= correlating highest with performance. However, ture and staffing patterns. Nurses on the hoard the nurses were rated by their peers as having usually questioned the figures by comparing them the lowest influence over all participants the sec- 115 NURSING ontrcroas AS PARTICIPANTS IN tiosPITAI. PEER REV1EWBOARDS 107

Table 1.-Interater reliabilities for all the peer review experience valuable for increas- strategy items ing awareness in how to manage a budget and the department. Doan! siva! egy Hospital st tegy Learning stage Organizational Problems Outside the Board 1:. rept ion Preset' ta t ion velum Nothing Meetings Croup study 0.90 0.90 ($6 India i(111.11study 0.98 0.23 1.00 In the section on experiment design, we noted No st tidy 0.99 1.00 0.98 nonsystematically some of the organizationsin- Problem solving stage volved in this study.:listing of the organizations tuts tta 1 Niu 1Ia I tagemen t and their relationship at the beginning and dur- increases increases (list ussion ______ing the experiment supplies a further basis for Compare and contrast 0.1r) 0.71 0.33 understanding the experimental outcomes. Line bv.line review 0.74 0.72 0.82 The experiment was designed by the State Hos- 1.00 1.00 0.71 Nlanagentent discussion pital Association with the advice of an engineer- Decision stage ing firm and was supported by the Federal Gov- Sat isliet1 Quest ions Protests ernment and Blue Cross. Hospitals of various s'fes 0.94 1.00 1.00 Accepts _ were involved in the design to the extentti:ey 0.99 0.82 0.84 Delays wished, which often meant that administrators en- Changes 0.69 0.84 1.00 couraged the idea. The professional associations for administrators, controllers, or nurses were not involved. Table 2.-Mean contribution and influence rating Because there was a short lead time in starting according to role (N 16) the experiment, the actual participants had little Role ContriI utiuil Influence Average hand in the design, little time to tome to grips with the experiment. In addition, it became ap- Controller 3.4 3.5 3.5 Hospital 3.3 3.3 3.3 parent that in the large hospitals the administra- Chairman 2.9 2.5 2.7 torswho pledgedthehospital'sparticipation p 0.01 were too busy to participate. Also,the controll2r Coordinator 2.4 2.4 2.4 and nurse were nut sold on the value of the ex- Administrator 2.0 2.4 2.2 periment. The large hospitals dropped out after Nurse 1.5 1.2 1.4 the second year without administering rewards or penalties. and year, averaged over all the budgets (refer to During the experiment the large hospitals be- table 2). came interested in prospectivereimbursment. Al- These second year results support the results though they did not openly endorse this alterna- of the first year that, generally, the nursing direc- tive, privately they spoke of its advantages, which tors were relatively powerless during most ofthe were being able to control costsand place the review but influential during the nursing budget. blame on outsiders. However, this could he due to the diminished The engineering firm had trouble resolving the. attendance of nursing directors from the small and problems with engineering standards. This was large hospitals. The nurses from medium size partly because they had vested interests in the hospitals attended meetings regularly and were design of the experiment and because they as- very involved in their own budgets. sumedthathaving been apparentlysuccessful The third year saw the large hospitals dropping with the sante hospitals outside the experiment, out of the experiment and nursing directors in the failures within the experiment were not their small hospitals not attending meetings, often be- fault.It appeared that hospitals were willing to cause of a turnover illthis position. Of the four use engineering standards in behalf of apriori directors forthe medium sizehospitals,allat- plans they wanted Co institute. Hospitals were not tended the meetings, and three of the four found interested in using engineering standards if the 116 108 RESEARCH ON NURSE STAFFING IN HOSPITALS engineers wentcountertohosp talplan and ministrative experience had left the nursing direc- policy. tor job during tl..e experiment. Their replacements As for the nurses, they did not coalesce as a were hired with better managerial skills. group. Some half of the eight who had some mini- The problems thefirst year were blamed by mal administrative experience and training said thehospital, on the experiment staff and en- they increased their awareness of cost problems, gineers, by the staff on the hospitals and engineers, but they could not identify any changes in the and by the engineers on the other two. The rela- nursing department. The other half thoughtit tionship among these three groups varied, but was a waste of time. The four withlittlead- general:), it was distant to hostile.

Discussion

Looking at all four areas of technical and or- quently fed fictitious data to the engineers because ganizationalproblemsinsideandoutsidethe of low understanding that came from low partici- board meeting, we might speculate on the order pation. This analysis assumes that participation is of importance of the areas. This speculation re- useful both for utilizing resources, such as the quires an assumption. We assume that the thread controllers, and for developing resources, such as linking the areas is that of utilization and devel- the nurses. opment of human resources for the goal of cost These technical problems were carried into the containment. Or more simply, conscious partici- board meeting with bad data, few ways to process pation is the key. We have in casual order then, the data, and rejection of engineering standards. the external organization, the external technical, At this point, only awareness of the problems and the internal technical, and the internal organiza- development of ways to overcome them would tion. have worked. But neither the board's chairman We suggest that the lack of participation of all nor the nurses, among others, showed skills at relevant parties in designing the experiment led organizing, such as listing goals and search, and to low understanding in the execution of technical analysis and decision strategies to accomplish the problems of budget construction and engineering goals. The idea of participation becomes further standard estimates. This low understanding was refined because of the boards. Participation with- criticalinproducing problems intheseareas. out awareness of alternatives can be something Nurses did not participate in design and conse- even less than muddling through (11). NURSING DIRECTORS AS PARTICIPANTS IN IIOSPITAI. PEER REVIEWBOARDS 109

References

(1) Argyris,Chris. Human Relationsinthe (6) Bather, James. Power inCommittees: An Nursing Division. Labor and Management Experiment inthe Governmental Process. Center Report, Yale University, 1955. Chicago, Rand McNally, 1966, (2) Wolf, Gerrit. "A Model of Conversation." (7)Bales, Robert, and Strodbeck, Fred. "Phases Comparative Group Studies, 1970, 1, 275-305. in Group Problem Solving." Journal ofAb- (3) Wolf,Gerrit.''EvaluationWithinInter- normal and Social Psychology, 1951, 46, 485 - personalSystems,"GeneralSystems,(in 495. press). (8)Davis, James. Group Performance. Reading, (4) Wolf, Gerrit. Behavior in Decision Making Massachusetts, Addison-Wesley, 1969. Groups: Budget Approval Boards:1969. (9)Simon, Herbert. Models of Man. New York, Technical Report No. 33, Yale University, Wiley, 1957. DepartmentofAdministrativeSciences, (10) Weick, Karl. The Social Psychologyof Or- New Haven, Connecticut, 1970. ganizing. Reading, Nlassachusetts, Addison- (5) Wolf,Gerrit.BehavioralAspectsofthe Connecticut Hospital Experiment 1970-71. Wesley, 1969. . Technical Report No. 49, Yale University, (11) Lindblom,CharlesE."TheScienceof DepartmentofAdministrativeSciences, Muddling Through." Public Administra- New Haven, Connecticut, 1971. tion Review. 1959, 19,79-88.

118 Social PsychologicalFactors Relating tothe Employment ofNurses

Dr. Virginia S. Cleland Mary Lee Sulkowski Professor of Nursing Graduate Student, PsychiatricNursing Wayne State University Wayne State University Detroit, Michigan Detroit, Michigan

Introduction

the develop- The general purpose of this paperis to examine deleterious effect this has had upon the ment of women andthereby the development of factors that are antecedent or contiguous to will be given decision of a nurse to be employed. Thedecision the profession itself. Consideration is of great signifi- to alternative ways toview "women's place" and to participate in the occupation align itself with cance to nursing and to thepublic. Not only has the manner in which nursing can unsuccessful in some of the currentsocial and psychological forces the occupation of nursing been and women. producing adequate numbers of professional nurses that are redefining the roles of men prepared mem- There are economic and politicalforces that can but, also, large numbers of the best nursing's contribution to bers are utilized in educating nurseswho never suppott an expansion of Nation as the profession work after their first child is born. the health services of this documents that it can assume these newresponsi- With greater knowledge of the natureof the during bilities. differences between nurses who work only have passed those who In1975, one hundred years will periods of financial exigencies versus established nurs- since the first schools of nursing were actively seek long term professional careers, in the United States. However, the next10 years select students ing school faculties might attempt to will afford nursing its mostcrucial tests. Can the who are more likely to work for extendedperiods. profession adapt and grow in theface of social Also, with better u9derstanding of thecultural changes that will require thatit modify its very determination of certain values held by thestud- nature? The new, expandingclinicalrolesin ents, it would be possible tomodify these through nursing demand career continuity,independence the process of socialization into theprofession. of judgment, and initiationof action that are not The general theme of this paper isrelated to a congruent with thefeminine image of employ. view of the traditional role of womenand the meat inactivity,dependency, and passivity.

119 112 RESEARCH ON NURSE STAFFING IN HOSPITALS

Psychological Development

A brief review of selected aspects of the psycho- inent at this stageare the developmental tasks of social development of identity in women and num the childthe cognitive, physical, verbal,and inter- follows. Bardwick's Psychology of Women (1971) personal skills that the child seeksto master. Suc- providesauseful frame ofreference,asdoes cess in this area receives parental approval, and the Branden's The Psychology ofSelf-Esteem (1971). child recognizes anew source of enhancing his The Adolescent Experience(Douvan and Adel- well-being. son, 1966) and extensive study of adolescencepro- vides supportive data. 'Sexual Identity Bardwick (1) has reviewed themost influential formulationsofpersonalitydevelopment. Al- The child seeks a stable self definitionand one though she describes Freud'sgenius as "unassail- of the "givens" hecan readily grasp is his sex, able," she rejects his ideasabout the psychology which he begins to recognizeat the age of 5. At of women as reflecting politicalsentiments rather approximately the sameage the boy begins re- than scientific investigation.Freud saw a male ceiving less parental approvalfor dependent be- dominated culture and believedthat the penis was havior; likewise, emphasis isincreasingly placed on the symbolic source of thispower. Karen Horney his mastery of skills. As parentalapproval becomes rejected the idea thatwomen suffered from "penis a less reliable sovrce of well-being,as independ- envy." Horney believed that thefemale's capability ence and skill mastery are stressed, he beginsto of creating another humanbeing left men with rely less on others andmore on self for his feeling profound envy that could onlybe resolved by of well-being. Hecomes to understand that mas- seizingcultural andpoliticalpower. Bardwick culine identity(masculinity) correlates within- views this as a reaction-formationidea and rejects dependence and achievement andadjusts his be- most of it. The third point of view,currently havior accordingly. He developsinternal criteria popular, is that thereare no real differences be- by which to judge and rewardhimself and, cor- tween the sexes that extend beyond thebiological, respondingly, internalmeans of control. Bardwick and thus all differencesbetween the sexes have believes masculinity becomesa crucial issuc for the merely been imposed bya masculine dominated boy at age 5, that heuses the latency years to society. Bardwick rejects this notionalso. Bardwick formulate his concept of masculinityand comes to takes the position that thereare fundamental bio- a healthy adolescence with a fairly stableconcep- logical differences between thesexes but that tration. ditional role divisionsare too restrictive and that In line with the cultural definition offemininity cultural changes insex roles are necessary. the girl is allowed, in fact encouraged,to remain It is unlikely that thenature of biological dif- dependent. While achievement isimportant to her ferences upon cognition, forexample, can be ac- self concept, she continuesto depend primarily on curately assessed until culturalrestraints have been others for rewards and controls, ineffect for her removed. The general psychosocialdevelopment of self definition. Thus her emphasis ison the inter- boys and girls is outlined below. personal; others rather than self remainher most effective source of well-being.Her skill in dealing with others (interpersonal' skill)becomes impor- ChildhoodA Sense of Self tant. Our culture is more tolerant and indulgent Bardwick states "the basic of girls. Accordingly, there isless emphasis during sense of self, as a the latency years on femininity; separate entity, probably beginsto develop around "tomboys" are al- the age of 2, when lowed. Hardwick believes thatfeminine identity a child begins referring to does not become himself as 'I.' " Previouslydependent on his par- a crucial issue until adolescence, ents for emotional and physical well-being,his sense of self develops to a great extent throughhis Self-Esteem interactions with his parents and isdependent for The child's efforts to gaina sense of self and a its well-being on parental approval, Equallyprom- feeling of well-being refer to the dualprocesses of

140 SOCIAL. PSYCHOLOGICAL. FACTORS RELATING TOEMPLOYMENT OF NURSES 113 achieving. identity and esteem. Branden defines self-esteem. In adolescence, with cultural, parental self-esteem as the "integrated stun of self.confidence and peer pressure focusing on the importanceof and self-respect. Itis the conviction that one is femininity as preparation for the wife-mother role, corn petent to live and worth.) of living(2). heterosexual affiliation becomes crucial: Douvan As the boy develops, his self esteem increasingly and Adelson believe the adolescent girl does not comes to depend less on others, more onself. The have the time, the energy or the motivation to stress on the independent securing of satisfaction focus on identity resolution. Popularity, especially and control foreshadows the development of au- with the opposite sex, absorbs most of the energy. tonomy in adolescence. But the dependence fos- Further, "too sharp a self-definition and too full tered in the girl makes "the other" crucial to her an investment in a uniquepersonal integration self-esteem. Bardwick states: "While the interper- are not considered highlyfeminine (3)," and might sonal rewards remain salient for girls, internalized therefore reduce eligibility. criteria of achievement and personal satisfaction become more important for boys yr Marital Status and Female Identity The girl's identity is bound up not so much in AdolescenceIdentity what she is but in what her husband will be. Someone has spoken of marriage as a "mutual Branden defines personal identity as "the sense mobility bed." We may add that for the girl it of being a clearly defined psychological entity (2)." Douvan and Adelson note that although identity is equally an identity bed (3). Bardwick arrives at the same conclusion: does not begin in adolescence, it becomes critical at this time: The only way to achieve a feminine sense of During this period, the youngster mist synthe- identity,if one has internalizedthe general size earlier identifications with personal quali- norms, is to succeed in the rolesof wife, help- ties and relate them to social opportunities and mate, and mother. ...As a man's self-esteem socialideals. Who the child is to be will be is linked to his appraisal of his masculinity, ap- influenced (and in some cases determined) by parentlyawoman'sself-esteemissimilarly what the environment permits and encourages linked to her feelings of feminity. The difference between them isthat the goal of esteem and (3). The key terms in male adolescent development identity is achieved much later in life by women are "the erotic, autonomy (assertiveness,independ- and is primarily achieved in the traditional in- ence, achievement), and identity," the key terms tense and important relationships ratherthan in for the female"the erotic, the interpersonal, and occupational achievement (I). identity (3)."The onediscrepancyinterms The little girl's potential wife-mother role deter- autonomy vs. the interpersonalshades the entire mines the pattern of her socialization. Although it course of adolescent development. The identity is allowed that she may at some time work outside problem for the girl revolves around "the inter- the home, the focus is her traditional sex role and personal," for the boy "autonomy" is the central the "feminine" qualities of dependence, deference, issue: nurturance, and passivity. She has been success- Identity is for the boy a matter u: individuating fully socialized when she assumes the socially ap- internal bases for action and defending these proved status of "Mrs." Mead has written, "Success against domination by others . . . (A. the girl) for a woman means success in finding and keeping it is a process of finding and defining the internal a husband (1)." and individual through attachment to others Although itis assumed that the little boy will hissocializationfocusesonhis (3). beafather, This feminine dependence on "the other" leads future occupationalrole. He istaughtto be Douvan and Adelson to question the validity and "masculine"independent, aggressive, competitive, relevance of the concept "identity crisis" to the achievement oriented. adolescent girl. In her early years, the girl learns Marriage provides a socially approved arrange- to depend on others for her self-definition and her ment for bearing, rearing, and caring forchildren.

121. 114 RESEARCH ON NURSE STAFFING IN HOSPITALS

But marriage also "offers social recognitions, so- too dependent upon husband and children for her cial place, and designated roles and thus gives its self definition. The belief that ir.otherhoodisa membersa,vitalformofsocialsecurity (5)," drive, anchored deeply in the biological sphere, While thisstatement has implications for both leaves the woman who remains single to be viewed sexes. it has special and essential implications for as abnormal and deviant. This may cause the sin- women, whose primary role is wife-mother, whose gle woman to question her normalcy. Having in- "place- is marriage and motherhood. ternalized the cultural definition of femininity, she A recent article in Cosmopolitan opens with the questions her feminine idertity. Unlike the man. following statements: who has developed internal standards and criteria As an unmarried woman, I am suffering a pro- for judging and rewarding himself,the single found identification crisis. I am a spinster. woman has learned to define and value herself as Spinster is, quite simply, the legal description of she is defined and valued by others. Society neither the unmarried adult female. Why should the adequately defines nor values singleness. unmarried adult male, or bachelor, be a figure Through processes of definition and social rein- of glamor, and the unmarried woman, or spin- forcement the essentially broad patterns of nor- ster, a pitiful lump (6)? malcy in sex roles have become rigid sex stereo- Althoughthepsychologicalimplicationsof types. Once the characteristics of the problem are being single are not well explored in the scientific identified, modification of the process of develop- literature, Sakol focuses on two concepts pertintM ment generally becomes possible by altering the to an understanding of single women. The first, social influences. "identification crisis,"is self-evident. The second, lack of self-esteem., is implied in the phrase "piti- Toward Resolution ful lump." The unmarried woman does not assume the Women's ambitions are not fixedeither bio- traditional feminine role and thereby exists on the logicallyorculturally. They aresubjectto fringes of the social structure. She may work and change in response to circumstances....There contribute to society through her work, but itis arecrises, or criticaltransitional experiences not quite acceptable for her to define herself in which stimulate a reorganization of values and terms of an occupational rolethat is a man's role patterns of aspiration and gratification (7). and calls for "masculine" behavior. A woman may One can question the health of a society that so be a sister, a daughter, a community participator, narrowly constricts a woman's source of identity but these are not primary roles that readily lead to and esteem to one primary role. Mannes credits the self-definition. mass media for women's obsession "with an ideal There is no acceptable status setting name for of femininity as the guarantee of happiness (8)." the unmarried woman... exceptfor the des- Bettelheim feels our view of women is "psycho- ignation "unmarried," which is a designation by logically immature (9)." social default, as it were, not of a social status. Bardwick reflects on the awareness of parents The woman who remains single is ambiguously who encourage their daughters "tr ew the love of placed in the social structure (5). a man and the raising of child, en as the self- By defining woman in terms of marriage and defining, self-esteeming achievement,limiting motherhood,societyessentiallylimitswoman's achievement to the traditionalasks and the need source of status, recognition, and security to the to affiliate (1)." She feelsiti,'dangerous for a wife-mother role. By socializing little girls to de- woman's sense of worth to be enormously depend- pendency and passivity and a focus on the inter- ent upon her husband's reaction to her and to her personal, society effectively circumscribes woman's contribution to his welfare (1)." source of identity and esteem around husband and In the process of resolution, women may con- children. sider Branden's examination of the psychology of The female's dependence on othersforself- self-esteem, contrasting his statements with the cur- definition leaves all women vulnerablethe mar- rent cultural definition of femininity. Branden ried and the unmarried. The married woman is states: SOCIAL. PSYCHOLOGICAL. FACTORS RELATING TO EMPLOYMENT OF NURSES 115

A strong sense of personal identity is the product men tasit human being dependent on "the bio- of two things: a policy of independent thinking logical process of fathering a child (10)." and the possession of an integrated set of valueS Nfead states the problem: (2). Every known society creates and maintains arti- Further in his discussion: ficial occupational divisions and personality ex- A psychologically health man does not depend pectations for each sex that limit the humanity on others for his self-esteem: he expects others of the other sex (4). What needs to be questioned is not a woman's to perceive his value, not to create it (2). desire to marry and have children but rather the Branden briefly mentions: fact that her self-definition primarily Focuses on . . . theincalculable damagethathasbeen her biological role. It is time to question the rele- wrought by the conventional view that the pur- vancy and desirability of archaic conceptionsof suit of a productive careerisan exclusively "masculine" and "feminine" that arbitrarily limit masculine prerogative, and that women should one's personal potential to biology. not aspire to any role or function other than What is needed is a more human definition of that of wife and mother. A woman's psycholog- male and fethale. In a recent article. Ulmer dis- ical well-being requires that she be engaged in a cusses three sex role systems. In the"patriarchal long range career: she is not some sort of second system," the male is the provider and therefore class citizen, for whom mental passivity and de- superior. This system may be efficient within the pendence are a natural condition (2). family but LimierfeelsithasnegativeCOH5C- Before she car find identity and esteem in her Times for society. The "Complementary sex role work the woman must have new values that ensue system," although claiming men and women are from a fresh, healthier self-conception. Branden equal, retains an emphasis on basic differences be- believls "the cause of authentic self-esteem is. . . tween male and female. The "human role system" the rational, reality directed character of mind's acknowledges the differences between the sexes but thinking processes (2). He designates "five inter- sees them as secondary (II).This system proposes connected areas that produce enjoyment of life: the possibility of balancing male and female sex productive work, human relationships, recreation, roles and working toward a new "human role." As art and sex (2)." While one may disagreewith his human beings we,ti reality, play many roles. 'Co categories, itis worthwhile to note he does not focus so exclusively on one limits our humanness. refer specifically to marriage and motherhood. A society threatened with extinction may need to To take another perspective, Baker notes that impose such limits. Our society not only needs the marriage and parenthood arc not a man's "primary rich complexity of. individual human potential,

avenue to personal identity" . . . noris his fulfill- but it can not really afford to do without it.

Implications for Nursing in Resolving Rigid SexRole Conformity

Career Choice and keep a job but without the sort of commitment that will make her either toosuccessful or In the childhood years a girl is encouraged to unwilling to give up the job entirely For marriage achieve in school. But adolescence brings a cul- and motherhood (4)." A recent study by Constan- tural emphasis on Femininity and feminine be- tinople reveals that college reduces conflict for men havior. Passivity, dependency, and deference are since it equips them for occupational success but qualities that conflict with those that spell aca- increases conflict for women. The emphasis on demic success;for example. competition, aggres- competitionand occupationalcompetence and siveness, achievement. Commenting on the girl's achievement does not correspond to the feminine educational dilemma, Meade noted that girls are ideal (12). expected to display enough ability to be "success The natural congruence between the traditional ful, but not too successful; enough ability to get role of women and the occupation of nursing has 1a 116 RESEARCH ON NURSE STAFFING IN HOSPITALS made iteasy to recruit girls for nursing. When selves professionally attrative to males. This is the nursing programs were relatively inexpensive and traditional approach that women have always used, students 11-1 paid, in tuit:on and service, for their 11 that were to happen, nursing would merelyre- education ny the time they graduated, the fact semble more closely the professions of teaching, that many chose not to be employedwas not a social work, and library science whereinwomen serious problem. Nursing has continuedto con- are dominant numerically but men hold all of the done this practice when itcan no longer be de- power positions. The image of nursing will not fended. be altered simply by recruiting moremen into the Vaillot has written that the committednurse is feminineoccupation. Rather women who are one who views nursing as a means of achieving nurses must free themselves from the narrow norms being, as a vital way of expressing and definingthe of lemininit- self. Although the subjects of her study overwhelm- To conforn. to the mean of the norm leadsto ingly ranked marriage and motherhoodaspri- dull,unimaginative,steeotypedbehavior. The mary, she postulated that this does not preclude important ialchanges are brought about by commitment because students see marriage and persons exploring the outermost edges ofthe motherhood as a natural extension of theircom- range of behavior, which in turnis defined by mitment to nursing (13). Douvan and Adelson one's personal system of values. Women whoare demonstrate that girls who choose feminineoccu- alive and growing will developnew insights and pations such as nursing view these occupationsas values.Thisinvolvesgrantingtheindividual extentions of the traditional feminine role and as nurse the rightto be ditTerentwithout being a means of expressing feminine qualities (3).If labeled as deviant. Female nursescan be helped to nursing is an extension of the traditional feminine live their full human potential so that nursingas role, if the primary commitment is to marriage and an occupation will change and become a field at-. motherhood, commitment to nursing is secondary tractive to both men and women. and temporary. Astin, Stmiewick, and Divech, intheir over- Extrapolating from tables in Facts about Nurs- view of studies on career determinants, have re- ing,1969, andliA"s 1966Inventory (the most ported that there is general agreement that women recent national survey of registered nurses), it can with career orientations tend to be high achievers be estimated that out of every 100 ?flurried gradu- and in aptitude testing show high need for achieve- ate nurses 31.5 are inactive and not registered, 31.5 ment and higher levels of endurance, introspection, are inactive but registered, and 37 are employed. and independence. These same tests tend to de- However, of these 37 nurses who are employed, scribe career orients women as having a mascu- 10 work only part time. Thus, out of every 100 line orientation. The authors conclude, graduates from schools of nursing who marry, only 11. unfortunatelythepresentclinicalinter- 27 nurses are employed full time at any one time pretation of a person who has interests and/or (14,15). talents in areas that are descriptive of the op- Nurses, individually and collectively, must rid posite sexisthat such interests and aptitudes themselvesofrigid sex stereotypedthinking. reflect one's lack of acceptance of his/her sex- Women who happen to be nurses must learn to see uality and that the person is in some kind of themselves as Iminan beings who have a contribu- psychological conflict. This interpretation war- tion to make and who have an identity greater rants reconsideration in light of evidence that than thcifeminiue selves. Nursing as a human great numbers of women desire careers that have service is so tied to the nurturant role of women traditionally been viewed as careers for men that nursing's professional role can expand only as (16)." the societal role for women expands. Nursing as an The entire practice of classifying occupations as occupation and nurses as women must seek new feminine or masculine needs to be restudied. Itis definitions and new roles. Tile two processes must !gised upon culturalbiases and produces poor he concurrent and mutually supportive. science. The Strong VocationalInterestBlanks To suggest that nursing be made acceptable to have been cited for their sex discrimination. It is males is to suggest that female nurses make them- notonlythatcertainscoreconfigurationsfor

124 SOCIAL PSYCHOLOGICAL. FACTORS RELATING TO EMPLOYMENT OF NURSES 117 women will suggest psychiatric social work and for ing students primarily oriented to the traditional men, psychiatry, but thattheportrayalofthe role, and they view this orientation as precluding women's vocations are biased and restrictive (17). serious career commitment (20).Mayes, Schultz, There is need to develop new vocational aptitude and Pierce considered student satisfactionwith instruments that reflect potential for nursing ca- nursing as it influences commitment. They found reers of the1970's and 1980's rather than the students consider the career vs. marriage conflict a 1930's. source of dissatisfaction and interpret this as an obstacleto commitment (21).Freeman,etal., document studies thatidentify marriage as the Career Without Marriage primary goal of nursing students, a goal they be- A generation ago women often chose between lieve prevents many from a serious commitment to nursing. marriage and a career, or atleast, these were The prospect of marriage and children per- seenastheprincipalalternatives... .The choice for career in place of marriage is prob- meates every aspect of nursing; no aspect of the ably no longer a consideration of young women. profession can be completely understood apart The new decision is for marriage alone, or mar- from the influence of marriage plans and their frustration (19). riage with career (18). Nursing can bea viable means of achieving The single nurse internalizes the cultural defi- "being" and self-definition. However, as long as nition of femininity and desires the socially ap- the traditional role model predominates in the proved stunts of marriage and motherhood. Coping social and psychological development of young with her "failure" to achieve woman's traditional women, women will be incapable of achieving role, the single nurse under 30 faces acrisis of "being" through nursing. Their commitmer,ts will identity and esteem. The dilemma delays, if not be half hearted. Nursing has never explored the totally precludes, a serious commitment to nursing. nature of the problem whereby itsservices are In1966,there were over150.000 unmarried provided by women who essentially remain aloof nurses in the United States. Of those under 30, from a deep occupational commitment. 95 percent were actively employed inthe profession compared to 64 percent of married nurses ina comparable age range (11). Unencumbered Marriage and Career by the responsibilities of marriage and motherhood, is the single nurse more likely to concen- The research interest of the senior author has tratefull energy on her nursing career? Is sEe focused upon the decision of married nurses to likely to possess a higher level of career commit- engage in employment. The Barnard-Simon the- ment? Or does failure to achieve the traditional ory of organizational equilbrium provides a rela- role pose problems that preclude authentic com- tional explanation of the high level of job turn- mitment? over and underemployment within nursing. One There are few studies of single women. Free- of the central postulates of the Barnard-Simon man, Levine, and Reeder note that historically theory states, "Each participant will continue in middle aged unmarried nurses have been the most an organization only as long as the inducements committed to the profession (19). In a 1968 study, offered him are as great as or greater (measured Baker found single women achieve "adjustment in terms of his values and in terms of the alterna- and fulfillment through creative contributions to tives open to him) than the contributions he is society" but the mean age of his subjects is 51 asked to make." Another postulate of the same (10). Althoughitis assumed that women who theory states, "An organization is 'solvent' and will pass age 30 without marrying review theirlife continue in existence only so long as the contribu- goals and expectations and readjusttheirself- tions are sufficient to provide inducements in large image, there have been no studies of how this is enough measure to draw forththese contribu- achieved. Similarly, there has been little explora- tions (22)." tion of the dynamics of "being unmarried" for the To a very great extent nursing is dependent up- 21 to 29 age group. Davis and Olson found nurs- on the employment of married nurses with child-

125 118 RESEARCH ON NURSE STAFFING IN HOSPITALS ren in order to keep the vocation "solvent." Yet existing skills rather than upon advancement these are the very nurses for whom thereare the or growth (24)." fewest employment inducements and the greatest The goal of this research has been to explain personalcontributionsassociatedwithemploy- or predict the employment behavior of married met, Simultaneously, the ready availability of the nurses. This has seemed appropriate since these housewife role has made it easy to leave the field are the womenwhohave an option to whenever the expected contributions appear to he exercise (25). too great and the inducements insufficient. Thewriter'spreviousresearch (26)supports From a research point of view, the problem has that of the other investigators, that financial need been to make the theory operational; thatis,to and absence of young children are the best predic- be able to measure the contributions and induce- tors of which women will choose to be actively em- ments as perceived by the individual nurse in order ployed (27,28,29). However, there are large numbers to predict when and under what conditions she of nurses who choose to work in the absence of will move toward labor force participation. The financialneed and before children are grown. writer has been the investigator fortwo projects Knowledge of the differences between nurses who inthisfield of research (NU 002 Hi-Decision to seek employment under these conditionsversus Reactivate Nursing Career and NIH 70 4128- those who do not could provide valuable insight Recruitment,ti li/ation, and Retention of Part- for the maintenance of a nursing work force. time Nurses). In the former study, use was made Two of the variables hypothesized to havean ofa Consequence Questionnaire developed by effect would be the strength of the self-concept of 'Rosen and modeled after Vroom's cognitive model the nurse and the family ideology subscribedto of motivation (23). by the couple. The influence of the self-concept This instrument was designed to measure the has been discussed. Whether the family ideology expectancy ami valence of the consequence of a (husband-wife power relationships) is "pa- change in work status. The nurse listed the ad-Itriarchal,""complementary"or"equalitarian" "antages and disadvantages of changing her work could be predicted to influence the extensiveness status (inactivetoactive or active toinactive) of the work role involvement of the wife. and attached to each an estimate of the probabil. Under the patriarchal system, employment of by that the consequence would occur. The total the wife would not be expected except inin- consequence score was determined by thealge- stances of financial distress. if the couple subscribe braic sum ofthepositive and negative consu- to a complementary ideology there would be em- quences alter earl: was multiplied by its probabil- ployment of the wife if she wished to he employed ity. 'Ellis instrument did not measure the theory's and if the work role could be made compatible concepts adequately and thus was not a good pre- to her family role. Vacillating conditions in the dictor of changes in employment status, However, home or in employment would be expected to pro- the listing of advantages and disadvantages when duce irregular employment participation. Finally, grouped by demographic variables such as age, the equalitarian sex role system could be expected education, and employment status resulted in an to correlate most highly with serious long-term excellent reservoir of descriptive data. careerinvolvement. The equalitarianmarriage From that study the writer concluded that "em- w .yidafford the housewife-careerist the greatest ployment of married nurses can most often be personal support for the full development of her characteri/ed by the principle of immediate grati- professional potential plus the greatest sharing of fication, namely: home and family responsibilities. Not all adjust- rapid job turnover and change of work sta- ments for home or family complications should he tus when family problems arise, imposed upon the wife's employer. Increasingly acceptance of dead end jobs that are compati- such occurrences willhe classifiedasparental ble with family needs and demands, rather than maternal responsibilities within equal- work goals that focus upon maintenance of itarian families.

126 SOCIAL. PSYC,i0LOGICAL FACTORS RELATING TO EMPLOYMENT OF NURSES 119

Conclusion

The common delay of a serious commitment to mother. There is need for controlled studies of nursing until the issue of marriage has been re- the effectiveness of such socialization. solved causes many female nurses to waste, pro- The passivity,deference, and dependency in fessionally, those age years between 21 and 30. herent in femininity and inherent in nursing limit These are the same years in which society dictates professional potential. Woman's need for affilia- that young men will become vocationally estab- tion and approval makes her susceptible to others, lished. Nurses under 30 years of age should be. whethersheisseducedby sexorauthority. come an important concern of the profession for Energy directedtowards securing esteem from itis the manner in which these women resolve "others," be they medical or administrative per- their own identity (within or outside of marriage) sonnel, diminishes energy to be directed toward and establish some sense of self-esteem and self. change. Acceptance of the status quo results. To worth that will determine their long term involve. obtainaclearlydefinedprofessionalidentity, ment and contribution to nursing. The woman nurses must develop an caupational orientation. who views herself as a personal failure can never Traditionally, women have not defined themse1ves be a professional success. There is some evidence in a work role. Nursing independence requires an to suggest that women who marry after the age of achievement orientation along with the qualities 25 are more likely to combine career and marriage. of aggressiveness,assertiveness, and autonomy It is probable that the woman's identity as a per- qualities that threaten feminine identity. son, independent of any spouse or child,is suffi- Nursing needs to question and challenge a tra- ciently developed by this age for career involve- ditional role concept that limits a woman's poten- ment to be viewed as personally satisfying and tial,creativity, and humanity. New deF---' ions, rewarding. new alternatives need to beexplored. As lung as Certainly thereis evidence from the feminist woman's primary definition is wife and mother, movement that identity can be significantly in. nursing will be primarily defined as a feminine, fluenced long. after adolescence. "Rap" sessions in nurturing, mothering profession. An emphasis on schools of nursing might well be used to help the "human role" would allow women and men, students(graduateorundergraduate)perceive within nursing,to seek broader self-definitions, their options for achieving full human potential, untapped avenues of participation in society, and including or excluding theroles of wife and improved ways of delivering health care.

1..47 120 RESEARCH ON NURSE STAFFING IN HOSPITALS

References

(1) Bardwick, J. M. Psychology of Women. (New and Careers, (New York: Washington, D.C., York: Harper and Row, 1971), pp. 5 -16, Human Services Press, 1971), p. 4. p. 175, p. 189, p. 188, p. 151. (17) Schlossberg,N., and Goodman, J.Resolu- (2)Braude!), N. The Psychology of Self-Esteem. tion to the American Personality and Guid- (New York: Bantani Books, 1971),p.110, ance Association: Revision of Strong Voca- p. 173, p. 174, p. 204, p. 131, p. 133. tional Interest Blanks (Detroit: College of (3)Douvan, E., and Adelson, J. The Adolescent Education, 1971). Experience. (New York: John Wiley and (18) Epstein,C.F.Woman's Place.(Los An- Sons,Inc.,1966), p.157,p. 347,p.348, geles: University of California Press, 1971), p. 349, p. 18. p. 94. (4) Mead, Margaret. Male and Female. (New (19) Freeman, H. E.;Levine,S.; and Reeder, York: Wm. Morrow and Co., 1949),p. 323, L. G.(eds.). Handbook of Medical Soci- p. 372, p. 320. ology.(EnglewoodCliffs,N.J.:Prentice- (5)Perlman, H. H. Persona. (Chicago: University Hall, 1963), pp. 200-201. of Chicago Press, 1968), p. 102, p. 100. (20) Davis, F., and Olesen, V. L. "The Career (6)Sakol, J. "The Case for Spinsterhood: Bet- Outlook of Professionally Educated ter Dead than Wed?" Cosmopolitan, 1971, Women7 Psychiatry, 1965, 28, pp. 334-345. 171 (6), pp. 92-100. (21) Mates, N.; Schultz, M. N.; and Pierce, C. M. (7)Fogarty, M. P.,Rapoport, R., and Rapo- "Commitment to Nursing. -How is it port, R. N. Sex, Career, and Family. Lon- Achieved?" Nursing Outlook, 1968,16(7), don: Allen and Unwin Ltd., 1971, p.197. pp. 29-31. (8)Mannes, M. "The Problems ofCreative (22) March, J. G., and Simon, H. A. Organiza- Women." In S. M. Farber and R. H. Wilson tions. New York: John Wiley and Sons, Inc., (eds.), The Potential of Women, (New York: 1958, p. 84. McGraw-Hill, 1963), p. 123. (23) Rosen, H., and Komorita, S. S. "A Decision (9)Bettelheim, B. "Growing Up Female." Har- Paridigm for Action Research." Applied Be- per's Magazine, 1962, 225; pp. 120-128. havioral Science, 1970, 19, pr 509-518. (10) Baker, Jr. L. "The Personal and Social Ad- (24) Cleland, V., and Bellinger, A., et al. "Decision justment of the Never-Married Women." to Reactivate Nursing Career." Nursing Re- ournal of Marriage and the Family, 1968, search, 1970, 19, p. 451. 30, p. 478, p. 437. (25) Cleland, V. "Role Bargaining for Working (11) Linner, S. "What Does Equality Between Wives." American Journal of Nursing, 1970, Sexes Imply?" American Journal of Ortho- 70, pp. 1242-1246. psychiatry, 1971, 41, pp. 747-756. (26) Cleland, V.;Bellinger,A.;and Meek L. (12) Constantinople, A. "An Eriksonian Measure "Husband's Income and Children's Ages as ofPersonalityDevelopmentinCollege Motivational Factors in Nurse Em- Students." Developmental Psychology, 1969, ployment." in V. Cleland (ed.), Part-time 1, pp. 357-372. Nurses, (Detroit:Center for Health Re- (13) Vaillot,M.C. Commitment toNursing. search, Wayne State University, 1971). (Philadelphia: J.B. Lippincott Co., 1962). (27) Nye, F., and Hoffman, L. The Employed (14) American Nurses' Association, Facts About Motherin America. (Chicago: Rand Nursing. (New York: American Nurses' As- McNally, 1963). sociation, 1971), p. 9. (28) Shea, J. R. Dual Careers: A Longitudinal (15) RN's 1966 Inventory. (New York: Ameri- Study of the Labor Market Experience of can Nurses' Association. 1969). Women. (Washington, D.C.: Department of (16) Astin, H. S., Suniewick, N and Diveck, S. Labor, Manpower Research Monograph No. Women; A Bibliography of their Education 21, 1970).

123 SOCIAL PSYCHOLOGICAL. FACTORS RELATING TO EMPLOYMENT OF NURSES 121

(29) Women's Bureau, Handbook on Women Michigan. (Date from a questionnaire, the Workers. Washington, D.C.: Department of findings of which have not yet been pub- Labor, Women's Bureau Bulletin 294, 1969). lished). (30) Personal Communication with Dr. Robert (31) Cleland, V. "Sex Discrimination." Ameri- Nfendohlsolm, Associate Professor, Lafayett can Journal of Nursing, 1971, 71, pp, 1542- Clinic, Wayne StateUniversity,Detroit, 1547. DISCUSSION OF DR. WOLF'S AND DR. CLELAND'S PAPERS

Discussion Leaders Miss Marjorie Simpson Dr. Mario Bognanno Nursing Officer in Research Assistant Professor Department of Health and Social Security Industrial Relations Center London, England University of Minnesota Minneapolis, Minnesota

Miss Simpson

I would like to discuss the two papers h. re- solutely to address me as "Miss" because clearly verse order, starting with Dr. Cleland's. She has this was insulting. To this day I am on the records touched on the universal problem of recruitment of the hospital as "Mrs." I do not believe this selection and wastage of stafffor nursing posi- would have happened had the ward sister been tions. Her arguments are very good; however one English. or two points did occur to me. It should be possible today for single women, The first is whether we need to follow slavishly for men, and for married women to mingle to- the career pattern that men set. I think we have gether in the nursing profession, finding their own to consider the resources we have and the needs position. But for the married woman with chil- we have to meet. We should I am sure be happy dren, I personally have swallowed hook, line, and to have men with us as nurses and I believe that sinker the things researchers have identified in rein my own country we have gone farther than lation to the deprivation that children suffer when you have here in incorporating them into the they are deprived of their mother's care in the profession, not as men, but as nurses. early stages of their lives. In this way, of course, Single women living away from home in medi- my public health biases come across very strongly, eval times were known as "anilepi" women. They and I cannot reconcile the idea of the nurse who were thought to have no viable life of their own should be the health teacher setting the example nut throughout their lives they were expected to of leaving the children. This I think you may be dependent on a male member of the family, strongly wish to dispute with me. This is my bias, usually the senior male member. They held posi- which I am now declaring. dons as daughters, sisters, or aunts. They were not But I do wonder if we couldn't cash in on this independent. This has lingered with us, but not situation, whether it is biologically determined or Ithinkto the extent that we haven't broken culturally determinedI don't know which itis. away from it for women to hold happy, successful In nursing there are very many positions at the positions in their own fields of work. foot of the ladder. There are far fewer higher up, This varies greatly from culture to culture. I was, and I am not sure that wastage for child rearing for instance, astonished recently when I happened is entirely dysfunctional. If we are preparing our to be in a hospital. I had as a ward sister a very nurses for health care as well as for sickness care, excellent nurse from Singapore who refused ab- we may be able to rear a much healthier future Id()1.22 DISCUSSION OF DR. WOLF'S AND DR. CLELAND'S PAPERS 123 generation if the mother understands the import- going on with the idea that in 1974 the transition ance of child health. This is not only within her will be made smoothly from a tripartite organiza- own family. As a member of the community in tion of the health service to a unified organization. her peer group she can have a gt eat influence on One small segment of this scheme is looking at other mothers, and we should make it possible for the management of areas, and it has been quite her to contribute in voluntary service or in other clearly stated that the nurse director, whatever ways to the health service during the time she is she will be called in this new setup, will have to rearing her own family. be able to develop and control the nursing budget, One of the problems is that we prepare nurses not only for the hospital but for the area as well. for a caring role that they have to change at This is another thing for which we are not at some stage in their career either for a much more present well prepared, and certainly for us this technical role or for an administrative or academic is one of the things we shall need to look at. role. And I do wonder whether we could use this The other feature that came out so clearly in gap to train or retrain the women probably of Dr. Wolf's paper was the problem of action re- about age 35. search. It is not too difficult for the outsider to go Now, there is experimental work going on in in and describe a situation. We have many, many the training of older women. We have done far of these studies. It is when you try to work with less in thinking about the possibility of retraining people who are affected by the study that the prob- our nurses for a rather different role when they lems start to arise. We should congratulate Dr. come back. There is quite a lot of talk at present Wolf most warmly on his courage in coming here about the need for people in many occupations and telling us of the problems we shall meet. to retrain at 40. We might examine this possibility Those of us who have tried already know the diffi- in relation to nurses who have a natural break culties to some extent. in the late 1920's. He gave a most masterly exposition of the This leads us rather closely to the %object of things that are going to happen and that we need .the second paper. Here we saw all too clearly to look at to arrive at a situation in which the the problem of the nurse unaccustomed to the research worker can genuinely help the people on budgeting, management approach. This will need the field and not merely find ways in which to de- to be incorporated somewhere inhertraining period. In England, as you probably know, we are scribe them and criticize them. Itis not difficult moving into a unified health service, and 1974 to go in and describe how badly a thing is working. for many of us has a ring both of challenge and of It is a very different proposition to go in and help anxiety. We have a critical path analysis program people to adjust.

Dr. Bognanno

I would like to make a few very general com- study of the effect of peer review on the efficiency ments about each paper and then bring to your (lower cost, quality constant) with which health attention another body of research thatiscur- manpower, particularly nurses, is utilized by the rently being undertaken by a handful of econo- hospital is a very worthwhile problem area. mists in the area of nurse staffing. Dr. Wolf's study tried to get at answers to ques- In my mind there is little doubt that the nurse tions like the following: (1) how does the peer staffing research dollar will yield a considerable review p mess operate in relation to cost contain- return ifit results in the successful development ment and in relation to patient care, and (2) if the of operational schemes to motivate hospital econ- nursing director is on the peer review board, does omy. Two-thirds or more of most hospital expendi- her administrative knowledge increase; if it does, tures are allocated to salaries, and we all know isthis knowledge manifest in relationships with what has been happening to hospital costs in the subordinates and in relation to cost and perform-. United States over the past 15 years relative to ante? costs in other sectors of the economy. Thus the Well, these were the questions to which Dr. Wolf

131 124 RESEARCH ON NURSE STAFFING IN HOSPITALS and his colleagues tried to find answers. In con- of labor force participation begins to decline. Dr. currence with the previous discussant, I too think Cleland and Mary Lee Sulkowski see the entire these questions still go unanswered. However, in distribution of participation rates shifting upward our continual search for answers to them we can if it were possible to, say, create the same labor learn from Connecticut's experimental design and market role for women as that which currently from the many *research problems encountered dur- exists for men. That is, such a change in thinking ing the study as discussed in the concluding sec- (roles) would resultin more human (nursing) tion of Dr. 1Volf's paper. resour "es being allocated to market work. The ideas that 1)1'. Cleland and Mary I PC Sul- Thi,- brings me to the last point I wanted to kowski develop are rooted insocialpsychology. make regarding research currently being conducted From their literature review it would appear that by me and a few colleagues like Drs. Stuart Alt- considerable agreement exists among socialpsy- man, Lee Benham, Jesse Hixon, and Donald Yett. chologists regarding the point that a woman in We have been studying the relationship between the United States today does not engage in (at various measures of nursing labor supply and a least duringearly years) an occupational number of socioeconomic variables that impact commitmentjob achievement fotheridentity on the former. In general, we have not considered and self-esteem. Rather, the latteris found in the effects of change in sociopsychological variables the woman's role as a wife and mother. Speaking (as discussed by Dr. Cleland) on the nurses' labor as an economist, I consider the formal develop- supply and laborforceparticipationbehavior. ment of these theories to be excellent, and I am Rather, we implicitly assume that at a point in persuaded to agree with their specific explanation timein cross sectionthere is little or no labor of the work behavior of professional nurses. market role variation among nurses that could In effect, what Dr. Cleland and her colleague account for variations in the nurses' labor market are saying is that the supply of nursing resources behavior. iu society will increase, holding everything else This assumptionisrealistic enough when it constant,if we can negotiate or bring about a comes to cross sectional studies; however, through change in the social status of women as they re- time, cultural changes do occur and their effect on late to labor market work. In this instance the market work behavior must be taken into account. change being recommended isto broaden the As an aside, I should note that this is precisely woman's source of identity and self-esteem from the point where Dr. Cleland's study complements the mother-wife role only to that which also in- our work in this area. Indeed, one of the positive cludes her role as a professional in the labor aspects of this conference is that it has given us market. the opportunity to gain interdisciplinary insights If, on a graph, we label age on the abscissa into a common research problem. and the labor force participation rate on the or- The variables we have used in analyzing the dinate, then for nurses we would see a bimodal nurses' market work behavior include husband's labor force pattern develop. The probability of labor force participation would be relatively high earnings, the nurse's wage, family size, age of chil- during a woman's younger years after college; this dren, and the age of the nurse, to mention a few. probability would decline during her childbearing I am confident that most of you will find the and childrearing years, and then following these partial relationship between these variables and years it would increase until the woman enters the time a nurse spends in the labor market most her retirement years at which time the probability interesting. Evaluation of Quality of Nursing Care

Dr. Mabel A. Wandelt Professor of Nursing Wayne State University Detroit, Michigan

Introduction

In any project to evaluate the quality of nurs- either with all aspects of the program or, frequently, ing care, whether for research, educational, or with the complete absence of a program. administrative purposes, those directly responsible There are four sources of dissatisfaction with for nursing care of patients must be involved. evaluation programs in which nursing staffs are Though it means beginning on a rather negative involved. Perhaps the fundamental source of dis- note, it seems to me that it may serve us to identify satisfactionis failure to recognize the magnitude some of the difficulties these persons encounter as and complexity of the task. The second most basic they are involved in evaluation endeavors. It has source is the imprecision with which many inherent been my experience that nurses, of whatever level terms are used. The third is the failure to dis- of administrative hierarchy and inrelationto tinguish between and to make explicit the reasons and purposes for undertaking evaluation. The whatever purpose for evaluation, express dissatis- fourth is the lack of neatness possible in any eval- faction with evaluation. They may be very satis- uation endeavor where human action, interaction, fied with all other elements of a program, or vari- and reaction are involved. ous nurses may identifyparticular areas about There of course is interrelationship among these, which one or another hap mild unease, but when and the sequence of listing is not meant to indi- evaluationis mentioned thereis unanimous ex- cate greater or less importance, except for the pression of dissatisfaction. There is dissatisfaction primacy of the first one.

Understanding and Precision in Use of Terms

For reasons that will later be clear, 1 shall begin of sufficient concreteness and specificity, in con- discussions with the second listed source of dis- trast to abstractness and generalization, so that a satisfaction. Among the difficulties in relation to common understanding is achieved among those use of terms is failure to define the terms on alevel using the term. Perhaps the source of greatest

125 133 126 RESEARCH ON NURSE STAFFING IN HOSPITALS difficulty is the interchange of words with similar warranted. For example, recognition of the dis- or closely allied meanings. These arc often used tinctmeanings of measurement and evaluation as though they had identical meanings in one might well expedite the work in allprojects for sentence but had distinctly different meanings in evaluation. Failure to recognize. differences in the the next sentence. meanings of the .two terms leads to attempting to My protest is not that every word or term had begin in the middle or near the end of the process but one meaning or thatthere are not words of planning evaluation. Measurements arc identi- having two or more meanings. The point I wish to fied without establishment of a base on which to make is that interchange of words or use of one judge their relevance to the purpose for which word for two or more connotations ina single the evaluation is to be undertaken. general context frequently presents problems of communication. In most such instances itis possible to designate distinguishing definitions of the Standard, and Standard of Measurement confusing terms. Further,itis reasonable to suggest that, in any one context under consideration, Another area of difficulty derives from the inter- each word be used only inits defined meaning, change and failure to distinguish between mean- and that two terms, which in other context might ings of standard and standard of measurement. have interchangeable meanings, not be so used in Actually, except for the undesirable connotation the particular context. of average inherent in- the concept of norm,it For example. of immediate concern to us are might be well to adopt the use of the term "norm" the words "evaluation" and "measurement." You where ordinal flythe term "standard"is meant. may propose that, since their meanings are quite The connotation of averageishere considered distinct, itis unlikely that these two words would unfortunate because frequently, when standards be used interchangeably. Indeed,itistrue that are set, the intent is that the goal or what is de- their meanings are distinct, but it is also true that sirable is something above average. The latter does persons frequentlyfailto make the distinction seem to be implied by the concept of standard as they use them. A most common phrase, "evalua- whenitisdefined as "a level of excellence." tion and measurement," would seem to indicate The failure to distinguish among the meanings that many do not use the terms with precision. of standard, standard of measurement, and unit The structure of the phrase seems to indicate of measurement is frequently observed in nursing that individuals do not recognize that measurement literature, particularly where nurses are struggling is a prerequisite to evaluation. with evaluations of clinical competence.Inal- Delineation of varied usages of terms pertinent most any article on the subject, there is lengthy in considerations about evaluation and possible description of the various standards of measure- reasons for varied usages, along with the accom- ment to be used for evaluations of nurses with panying confusions or limitations in communica- various levels of educational background. tion, would involve detail beyond the scope of It is proposed that there be different standards this paper. Clarification and suggestions for allevi- of measurement because the same level of compe- ation of some of the difficulties may be served by tency cannot be expected for nurses with varied defining some of the involved terms, followed by educational preparations. What is meantis that identification of some difficulties in usage. To pre- there must be different standards, different levels clude confusion in the meanings of the various of competency that will be acceptable, not differ- terms employed in this paper, I have included at ent standards of measurement by which to ascer- the back a list of definitions. tain those levels. It seemed a good idea to start with a definition 1 am confronted with these misconceptions re- peatedly in relation to the Slater and Qualpacs of definition. Some of the definitions are excerpts scales, in which the standard of measurement is straight from Webster. Many, however, are con- "care expected of a first level stall nurse." Those structed on the basis of logic and fitfor the pur- considering use of the scales, whether education poses of the context in which they will herein be orservicepersonnel,askabout changing the used. Brief discussion of some of the words seems standard of measurement in the events of using

134 EVALUATION OF QUALITY OF NURSING CARE 127 the scales to measure competencies of nursing stu. ment! Her use of clinical judgment is accepted dents, staff nurses, or clinical specialists. They ob- without question when she usesitto instruct. viously are unaware of the meaning of a standard But when usedto evaluate the results of her of measurement being "an object which serves instruction, itis immediately something less than to define the magnitude of a unit." They fail to satisfactory. distinguish between two concepts: Perhaps it would helpif we looked at what (a) the concept of standard, which denotes an the words mean. To judge is "to form an idea accepted level of excellence and which implies or opinion abolit; to think or suppose." Subjective the appropriateness of establishing varying means "of, affected by, or produced by the mind." levels of excellence for various levels of stu- Actually,subjective judgmentisa redundancy: dent orforthe nurse staff with various all judgment is subjective; there is no judgment qualifications, without subjectivity. All nurse actions stem from (b) the concept of standard of measurement, judgments. Professional judgment has basis in fact which denotes an object that, under specific but would be nothing without the catalyst of conditions, serves to define or represent the subjectivity. magnitude of a unit and that, in turn, per- Always there is the quest for objective evalua- mits comparable measurement to he made of tion. Here we have blatant examples of impre- similar phenomena. It provides the unit of cise use of terms: subjective judgment is a redund- measurement to be used for determining the ancy, and objective evaluationis an antithesis. degree of attainment of an established or ac- Evaluation isthe term usually used by those cepted standard. expressing dissatisfaction,despite thefactthat they are usually talking about subjectivity in the measurement aspectofthe evaluationprocess. Subjectivity and Judgment Possibly some of the discomfort and condemnation occurs because the denotation of subjectivityis Another difficulty frequently identified by per- used instead of that for subjective. The diction- sons considering the use of tools for measuring ary definition of subjectivity is "The tendency to clinical competence of nursesisconcern about consider things primarily inthe light of one's subjectivity in measurement. In a way,it seems own personality. Concern with one's own thoughts remarkable that nurses have so much discomfort and feelings." Whereas subjective means "Of, af- about the matter of subjectivity and "subjective fected by, or produced by the mind." The defini- judgment" in relation to measurements of clinical tion of subjectivity is undoubtedly what leads to competence displayed by nurses. This is particu- the comment: "They evaluate a staff nurse on larly remarkable in relation to clinical instructors whether or not they like her as a person." and supervisors. Their whole effort in clinical in- Might nurses be helped if they were reminded of struction, guidance, and supervision is aimed to- specificsof what they are really about? That ward assisting nurses to make judgments. evaluation is a generalization descriptive of a judg- The professionalnurse'sforteis her clinical ment based on many measurements. The measure- judgment. In 01 of her clinical supervision, the ments are of the nursing care actions of theindi- instructororsupervisorusesher own clinical vidual being evaluated. The measurements on judgment to determine when a nurse being super- which the evaluative judgments are based are not vised is making a correct judgment and when she measurements of attributes or qualities of the may need guidance to improve her judgment. person. The person happens to be the vehicle for This is accepted with no problem. Now let the the entities being measuredthe performer of the supervisor redirect her conscious focus away from actions. The actions are being measured, not the guidance and instruction and toward evaluation for the purpose of accounting for the nurse's dis- perfomer of the actions. play of clinical competencea large part of which Also, nurses might develop confidence from a is professional judgmentand she, and all about reminder from Abraham Kaplan when he said, her, question the goodness of the evaluation she "The new treasonofthe intellectualsisthat might do because itis based on subjective judg- we have shared and even contributed to the cur-

135 128 RESEARCH ON NURSE STAFFING IN 110SPrAIS rent lossoffaithinthe power of the human not to have discovered that detailed specificity of mind to cope with human problems, faith in the terms can reveal whether an objective, as stated, worth of reasoned discussion,faith even in the actually says what those who constructeditin- possibility of objective truth.- It seems to me that tended it to say. They seem not to know that the nurses needto recognise specifically what their only assurance of relevant criterion measures' being concerns are and to have the confidence and four- identifiedisvia detailed and specific definitions. age to know that they are capable oc making in- There I commend to you the process of develop- formed judgments based on observed facts. ing a three-phase definition of the crucial terms where in the first phase is a general or dictionary de- Criterion Measures finition. The second phase is a pertinent general definitionthe general meaning of the term in the Another whole area in which nurses neglect to context of the concern of the evaluators. In de- use definitions and precise meanings of words to lineation of the pertinent general definition, the assist them isin identification of criterion meas- criterion measures will evolve. The third phase is ures appropriatetosecuring measurements on a for-instance definition, in which an example of which to base valid and relesant evaluations. When a single relevant measurement is described. This faced with the need to plan for evaluation, they process of definition ensures a systematic approach begin by identifying observations in terms of the to assurance that criterion measures and actual overall objective or purpose of the program to be measurements to be sought will be relevant to the evaluated. They fail to approach the task at even purpose for which the evaluation project is under- the suhobjectives level, not to mention entering taken. The crucial terms to be so defined are those at the level of constructing precise definitions of that stipulate the outcomes of achievements of the each crucial word in the subobjectives. They seem program to be evaluated.

Failure to Make. the Purpose Explicit

The general concept that evaluation is done for data germane to the purpose are to be obtainea the purpose of "gathering information on which For example, ifthe purpose for evaluating out- to base improvement" is useful as far as it goes. It comes of a new procedureisto determine its can he used to serve as a criterion on which to effectivenessinpreventing skin breakdown, the screen potential relevance of observations to be data to be gathered will be rather different from made. But thereare other purposesfor t:oing those gathered should the purpose be to determine evaluations of quality of nursing care. Among them the numbers and mix of staffing required to per- are quality control,theobligationof account- form the new procedure as recommended for a ability, continuing equation, and determining re- set of patients. lationships among a variety of potentially related Other factors alluded to earlier as encompassed variables. Here .again there may arise difficulties inthis broad area of difficulty, failure to make due to lack of clarity about meanings of words. purpose explicit, may be considered under the Reasons and purposes for doing evaluations are three-part framework for evaluation of human ac- sometimes confused. Particular measurements may tion programs: structure,process, and outcome. serve the intents of persons with various reasons When planning evaluation of quality of nursing for doing evaluations. For example, data that will care it is CSACIIIIII1 to be aware of what, specifically, provide for determining relationships among vari- isto be measured. What isto be measured, of ables will serve persons whose interestis research. course,is identifiedinthe declarationofthe the interests of the therapist. and those of the purpose for doing the evaluation. It is conceivable administrator. Therefore, failure to identify speci- that a particular purpos, would dictate measure- ficallythe reasonsisnot apttoinfluence the ment of all three components of care: structure, quality of the outcomes of evaluation. 41k. process, and outcome. If this is the case, the plan- On the other hand, the purpose for which the ners should he fully aware if the intent, They data will be gathered must he made explicit,if must also approach the task with definitive plans

136 EVALUATION OF QUALITY OF NURSING CARE 129 for measurement of each of the components. This complishecl through utilization of a single tool or awareness and approach will ensure cognizance procedure for gathering data. that few, if any, evaluation projects can be ac-

Lack of Neatness

The impossibility of accomplishing an evalua- of care, the setting and a situation of care, the tion of nursing care with a single procedure or attributes and qualifications of persons administer- instrument of data gathering is part of what I ing care, and the outcomes of care in terms of time mean by the lack of neatness of the process. If and efficiency and in terms of welfare of the nursing care were unidimensional, if there were an patient. established standard of measurement for this di- These and many more factors make for lack mension, if there were a validated, scaled instru- of tidiness and directness in the accomplishment of ment for making the measurements, evaluation of evaluation of nursing care. Perhaps the most im- nursing care would pose few if any difficulties. portant element in' dteviation of these difficulties But nursing care itself is multidimensional. There is the recognition and acknowledgment of their is no consensus about what many of the dimen- existence. The recognition of them and the nature sions are, not to mention consensus about stand- and influence of each should lead to organization ards or units of measurement for them. When for orderly approach to the task of evaluation. evaluation is considered there will be concerns They should lead to effective and efficientse- about the various facets of care, such as the process quencing of the components of the total task.

Failure to Recognize the Magnitude and Complexity of the Task

There was reason for elaboration of the other cisions must be made before any data are collected, sources of dissalsfaction with evaluation projects and many and varied data must be gathered and before addressing..the failure to recognize the mag- organized. In turn, further hard work of thinking nitude and complexity of thetask.Essentially, and decisions must be done before evaluative state- facets of the other sources as elaborated identify ments can be delineated, before there will be in- and describe components of the task. For example, formation to serve as bases for change, to fulfill there is the need to make explicit the reasons and the purpose for doing the evaluation. purposes for doing the evaluations, the need for Since research in evaluation of quality of nttrs- precise definitions of terms in order to reveal both ins, care must involve those directly' responsible the accuracy or lack thereof of the identification of for the nursing care of patients, the researcher who objects to be measured and the criterion measures is cognizant of the difficulties experienced by the relevant to them, and the determination of the understand and precise data germane to the purpose to be served by nttises and who assists them to the evaluation. cope with them will advance his investigation and There needs to be recognition .hat much hard incidentally will assist nurses with an important thinking and planning must be done. Many de- administrative task.

WhereIn Evaluation of Quality of Nursing Care

Variables of primary concern to investigators of described by Donabedian. The ultimate concern is nursing care of patients are those composing the the relationship of all other variables to those structure, process, and outcome of patient care as identified as outcomes of nursing care. In this

137, 130 RESEARCH ON NURSE STAFFING IN HOSPITALS context, all other variablesstaffing, facilities, philo- of nursing careinterms of patient well-being. sophy, policy, staff performanceare causal varia- All of the comments about difficulties in evalua- bles.Outcomespatient well-being,patientim- tion apply here. Some of us have wrestled with the provementare effects variables. Itis within this ideas of the magnitude and complexity of the task. broad framework that research in evaluation of We have faced the need to define some of the quality of nursing care must be pursut terms. In these attempts we ran into problems of Historically, research in evaluation of nursing making explicit the purposes for evaluation: Whose care has progressed along a continuum of the purposes are to be of concern? During a confer- concrete to the more-or-less abstract. There have ence lastfall, a group of us recognized and re- been successes in development of measurements of conciled ourselves to one difficulty:the task of structure:numbers ofpersons,time,facilities. measurement of outcomes of nursing care will not Among them are the "how to" techniques and be accomplished through development of a single instruments developed by the Division of Nursing measuring instrument. during the 1950's and early 1960's: how to study Recognition of the magnitude and complexity nursing activities in a patient unit, how to study of the task should be sought, not as an excuse for nursing service in an outpatient department, how immobility but rather as a caution that the task to study nursing supervision, and others. cannot be approached all of a piece. Progress to There have been successes in development of date has been made by individuals and groups measurements of process: nurse actions,interac- isolating single facets of the larger task and pursu- tions with and interventions on behalf of patients. ing information about each. It may be expected Among these are three developed by the faculty that continuing work will follow that pattern but of the College of Nursing, Wayne State University: always with the plea for increased communications (a) Slator Nursing Competencies Rating Scale, among the seekers. Communications will promote formeasuring competencies displayed by the work on individual projects and preclude un- a nurse, necessary overlap or repetition. (b) Quality Patient Care Scale,for measuring The goal to be sought through research in eval- the quality of nursing care received by a uation of quality of nursing care must be capability patient while care is ongoing. of determining relationships among variables of (c) Nursing Audit, for measuring the quality of structure, process, and outcome. There are availa- nursing care received by d patient after a ble to us, not finished or perfect but usable, meas- cycle of care has been completed and the urements of the first two. There are some measure- patient discharged. ments identified as potentially usable for the third Thesethree instruments and suggestionsfor area. But it seems to me we need to recognize their uses, have been described in some detail in the magnitude of the task and proceed in a system- an article in Hospital Topics, August 1972. And atic fashion by beginning with determination of there have been beginning studies to develop meas- criterion measures of outcomes of nursing care. urements of outcomes of nursing care. The criterion measures must be delineatedin There must be continuing work in all areas along terms of patient well-being. the continuum. But it does seem to me that we Work has begun. There is extensive time and have reached a point where a concerted effort is distance to go, but the beginnings have been made warranted in the area of measurement of outcomes and the direction is discernible.

138 EVALUATION OF QUALITY OF NURSING CARE 131

Definitions

Definition.An arbitrarily imposed description Judgment.The mental ability to perceive and that allows common understanding. distinguish relationships or alternatives; the critical Evaluate.To ascertain or fix the value or worth faculty; discernment. of. The capacity to make reasonable decisions, espe- To examine and judge. daily in regard to practical affairs of life; good sense, wisdom. Evaluate implies considered judgment in setting Subjective.Of, affected by, or produced by the a value on a person or thing. mind or a particular state of mind. Measure.To ascertain the dimensions, quantity, Of or resulting from the teelings or temperment or capacity of. of the subject or person thinking, rather than To mark off, usually with reference to some unit from the attributes of the object thought of; as, of measurement. a subjective judgment. Evaluation.A generalization describing a judg- Subjectivity.The tendency to consider things ment based on many measurements. primarily in the light o5 ont'N own persmatity. Evaluation.A process by which we gather infor- Concern with one's own thoughts and feelings, mation as a basis for improvement. Objective.Of or having to do with material ob- Quality.A characteristicor attribute of some- ject as distinguished from a mental concept, idea, or belief. thing: excellence, superiority, degree or grade of Having actual existence or reality. excellence. Uninfluenced by emotion, surmise, or personal prej- Standard.Something used by general agreement udice. to determine whether a thing is as it should he. Based on observable phenomena; presentedfac- An agreed upon level of excellence. tually. An established norm. Measurement.An objective term: the ascertain- Measurement.A comparison of a single pheno- ing of a dimension through observation of a phe- menon with a standard of measurement: the re- nomenon. corded number or symbolthatrepresentsthe Evaluation.A subjective term: the ascertaining magnitude of the phenomenon in terms of the or fixing of a value through considered judgment. magnitude of the standard of measurement. Criterion measure.A quality, attribute, or char- Standard of judgment.An object which, under acteristic of a variable that may be measured to specific conditions, serves to define, represent, or provide scores by which subjects of the same class record the magnitude of a unit. can be compared in relation to the variable. Unit of measurement.A precisely defined quan- Variable.A measurable or potentially me; ura- tity in terms of which the magnitude of all other ble component of an object or event that may quantities of the same kind can be stated. fluctuate in quantity or quality, or that may be Judge. To form an idea or opinion about (any differentillquantity or quality from individual matter). object or event to another individual object or To think or suppose. event of the same general class.

139 132 RESEARCH ON NURSE STAFFING IN HOSPITALS

References

(1) Donabedian, Avedis. A Guide to Medical Quality Patient Care Scale. College of Nurs- Care Administration. vol. 11. Medical Care ing, Wayne State University, Detroit, Michi- AppraisalQuality and Utilization. Ameri- gan, 1970. can Public Health Association, New York, (5) Wandelt, Mabel A. Guide for the Begr,ing New York, 1969. Researcher. New York: Appleton-Century- (2) Phaneuf,Maria C. The Nursing Audit: Crofts, 1970. Profile for Excellence. New York: Appleton- (6) Stufflebeam, Daniel L., etal. Educational Century-Crofts, 1972. Evaluation and Decision Making.Itasca, (3) Wandelt,Mabel A. The Slater Nursing Illinois: F. E. Peacock Publishers, Inc. 1971. CompetenciesRatingScale.Collegeof (7) Wadelt, Mabel A., and Phaneuf, Maria C. Nursing, Wayne State University, Detroit, "Three Instruments for Measuring Quality Michigan, 1967. of Nursing Care." Hospital Topics, August (4) Wandelt, Mabel A., and Ager Joel. The 1972.

140 DISCUSSION OF DR. WANDELT'S PAPER

Discussion Leaders Dr. Elizabeth Hagen Mr. Richard G. Gardiner Professor of Psychology Operations Research Specialist Teachers College Hospital Association of New York State Columbia University Albany, New York New York, New York

Dr. Hagen

In all the definitions that are given, thereis describelevels of evaluationineducation. He no definition of quality. During the past day and pointed out that most of the evaluation of quality a half, whenever we have talked about quality of of education was in terms of structure, the condi- care, values and value systems have been reflected tions under which education takes place, or of although no one has made them explicit. Many process, how the education was being conducted. of the arguments about what should be studied The product or outcomes of the educational proc- as indicators of quality of care arise from conflicts ess,according to Tyler,were rarely appraised in values. We would do well to make the value because they were believed to be too difficult to systems that underlie our selection of criteria for measure. Tyler criticized people for assuming that judging quality explicit. because certain structures or processes were pre- One problem with giving definitions is that you sent the desired outcomes must be present. We always want to start haggling with the definition. know that the assumed high relationship between I do not want to start haggling. However, I am process or structure and outcomes in education going to use the phrase "criterion variables" be- does not exist, but in education we still have most cause one of the immediate problems is to identify of the emphasis placed on structure and process the variables that are relevant indicators of quality and too little on outcomes. I am bothered when I of care and then to talk about how to appraise see the health care field repeat the mistakes of those variables. educational evaluation. Although the three terms We ought to make a clear distinction between may be useful in classifying variables used to judge the term "criterion variable" and the term "stand- quality of health care, they do not provide an ards." Theoretically a criterion variable can and adequate framework for studying quality of health should have a range of values from low to high. care. Standards imply that you are seeking some definite To make progress in the evaluation of health value or score on a particular variable in order to care, perhaps we should stop asking the global say the performance reflected by the variable is at question, "What is the quality of nursing care?" least satisfactory or adequate. and start asking more specific questions. The more The threeterms"structure,""process," and specific question will permit better identification "product" that Dr. Wandelt used are creeping into of the value system being used and will provide the literature on evaluation in the health care area. a better frame of reference for judging the rel- I believe these terms were first used in a paper evance of criterion variables than will the global by Tyler in1931. He used the three terms to question.

133 141 134 REsEARCH ON NURSE STAFFING IN HOSPITAL s

A number of frames of reference can be used comes might resultinthe methods of practice to study nursing care or, if yoo prefer the term, becoming ends in themselves or being used more. quality of nursing care. One frame of reference for the benefit of the profession than for the bene- could be what the profession of nursing would like fit of the patient. the practice of nursing to be. If you could identify A third frame of reference that could be used what. professional nurses deem to he "good prac- is consumer or patient .expectations of the health tice," you could appraise the extent to which the care or nursing care system. This frame of refer- practice matches these expectations. Or if you could ence differs from the second one mentioned in the identify the conditions necessary for delivering what previous paragraph. In the latter, patient outcomes is considered "best nursing practice," you could are judged by the health care professional in terms appraise the extent to which these conditions are of what the professional wants to achieve with the present. patient in terms of his health or well-being. In other A second frame of reference that could be used words, itis the extent to which the status of the to appraise nursing care ispatient outcome. To patient meets the expectations of the health care ascertain that nursing practice is congruent with prokssional. professional expectations of that practice does not In the third frame of reference, itis the extent guarantee that desired patient outcomes arc present. Although no one would argue that the ex- to which the status of the patient meets patient pectations of the profession concerning the desira- expectations or the extent to which the service ble practices of nursing are set without regard to meets patient expectations for that service. Patient what happens to the patient, we have very little expectations may be reasonable or unreasonable hard data on the relationship between nursing in the eyes of the-professional; however, they are practice and patient outcomes. To locus too much very realto the patient and cannot be ignored attention on practice and too little on patient outin the appraisal of any health care system.

Mr. Gardiner

I have simply a set of questions that hopefully stay recovered? It may be that we need to con- will provoke some response. First, isit not likely sider the relative costs and benefits of taking meas that points at which we make measurements of tires at any of these points to determine where we quality go a long way toward determining what ought to focus our attention. Where do we want ends up being defined as quality? Wecan measure to begin measuring to define quality? it either at the input level of theprocess or at the Why measure? That has been alludedto.Is output level. If it is measured at the output level, quality relevant to changes in staff? Harvey Wolfe are we not measuring the ability of the profes- asked the question and I think it's a good question. sionals to carry out their tasks; not whether, in At least two things indicate that if we are looking fact, the tasks are carried out, but whether they at quality simply as a reflection of marginal or have the ability to carry them out? small changes in levels of staffing, maybe the staffing We have said the psychological social needs of can be done independently and we do not need the patient are critical toward an evaluation of to examine quality. quality. I suspect this is equally true of the psycho- First,undercrisisorstresssituations(short logical social needs of the nurses who respond to staff), the quality of care delivered to the sickest changes in the level of staffing. This is,infact, patients probably remains a great deal different equally as important in their view as the quality from the quality of care delivered to the patients of care that they provide. who are not as sick. Do changes in staffing levels At the output level, we are judging the ultimate have an effect on the ultimate outcome? Not how effects in the patient's welfare or well-being. Does the patient perceives he's being treated or how the he recover? Does he recover sufficiently,orto nurse feels, but does ithave an effect on the what degree does he recover? How long does he fact of recovery? Second, are there not sufficient DISCUSSION OF DR. WANDELT'S PAPER 135 episodes of illness that are, themselves, self-limit- of care, the quality of medical practice, the quality ing? That is,will the patient get better despite of the ancillary services and all the other services the quality of nursing care delivered? that are provided in the health care system? Don't That seems to beg the question of why we want we have to decide a priori what the cost benefits are to measure quality. If we want to measure quality or at least estimate them to determine what aspects for purposes other than staffing,it seems to be of quality we want to investigate and what we want relevant. Is it relevant for evaluating the effects of to measure? changes in staff, at least marginal changes in staff? These are all considerations that must be un- Moreover, can the quality of nursing care be con- dertaken if we are to perform a realistic measure- sidered independently of other aspects of quality ment of quality of nursing care.

143 Psychosocial Factors in Hospitals and Nursing Staffing

Dr. 'Reginald W. Revans Senior Research Fellow and Scientific Advisor Fondation Industrie-Universite Brussels, Belgium

Introduction

In preparing this document I am only too aware ble more general measures may be to improve of our general ignorance of the psychosocial fac- conditions throughout health services as a whole. tors that influence the behavior of those who serve One way of changing a local situation is by ad- complex human institutions. After a year as rap- ministrativt. surgery, transplanting the person in porteur general to a group of management experts charge. But like surgery carried out in other con- nominated by the governments of the Organization texts, many subtle equilibria may be disturbed of Economic Community Development countries that were neither forseen nor even understood. I to study the relations between management and have therefore relied in this paper upon the blander labor across the industrial world, I am particularly resources of administrative and managerial psy- aware of may own ignorance. chotherapies. To what extent can we treat the In 1964 I concluded that hospitals are institu- anxiety situations as offering scope for learning? tions cradled in anxiety. In consequence, the most What are the conditions under which the daily urgent task of nurse administrators andall others work, however charged it may be with anxiety, upon whom they rely is toameliorate this anxiety. may become a process of autonomousdevelopment, No research work since that time has modified this of learning by doing, of social reinforcement? conclusion, although much has been done to sug- I give here two sets of illustrations from ac- Since the gest how the anxiety may he tempered. tivities in and around London. One set offers the anxiety varies significantly from one hospital to results, good and bad, of what was called the another and even from one ward to another in the HospitalInternal Communication (HIC) Proj- same hospital,itisclear that anxiety issignif- icantly a product of the local situation. Itfol- ect. The other describes aseries of discussion lows that only changes in the local situation are groups among nurses of all grades from matrons to likely significantly to affect anxiety, however desira- students at the King's Fund Hospital Centre.

137 144 138 RESEARCH ON NUKSE STAFFING IN HOSPITALS

Earlier Studies size and age and set in identical semblance ofsur- rounding population. The resulting comparison of A number of us at the University of Manchester, like with !E,e showed conclusively that hospitals Manchester, England, did researchon what began that could keep their nurses, the studentnurses, as a study of nurse wastage, or dropouts among could also keep their senior nurses. Even their girls who were qualifying to become nurses. matrons stayed longer, and this was also associated Manchester is the largest city in industrial Lan- with a short length of patient stay. The hospital cashire. When the Industrial Revolution took place with a highly stable nursing staff discharged their it left behind it in Lancashire 14towns that, with patients most quickly. the exception of Manchester,are all about the We examined this in many details. We found, sante size. They are allidentical; the same per- for example, that hospitals with short length of centage of people do all different kinds of jobs. patient stay had very short waitinglists. You The ratable value for head of population,com- would thinkpatients go throughthe hospital munity tax inhabitant is remarkably similar. Sowe quickly because many people are waitingat the have 14 separate towns scattered around Lancashire door to take their beds. But that isn'tso. The Mai are more homogeneous than any other 14 hospital that discharges patients :quickly hasa towns that you can draw anywhere. shortwaitinglist.Astheoperationsresearch We had thus assembled 14 ecologies that are people know, if the stay time is short themean distinct butidentical. In each of these was a time of waiting is small. hospital. We therefore decided we would examine This led us to examine very carefully a sample the structure of these hospitals, their metabolism of five hospitals. We examined them in great de- rather than their anatomy, to determine how suc- tail. We asked what ward sisters did. The head cessful they are in digesting those who enter them, nurses, as I think you call them, seemed to us to both nurses and patients. be the key people in the system. We examined Researchers in the extremely homogeneous en- the communication patterns of these hospitals. It. vironments of the Lancashire towns suggested that was clear thatthe causative variables that ex- hospitals with serious nurse staffing problems had plained staff stability and patient recovery were difficulty in retzining all grades, not student nurses very closely connected with the intelligibility and alone. If the senior staff and ward sisters were friendliness of the communication inside the hos- stable, were the nurses in training. In like man- pitals. We actually established methods ofmeasur- ner, instability reflected itself down the line from ing what communication was. senior staff to nurses in training. Again in this There is not only the dimension of fact but homogeneous Lancashire sample, hospitals with also perception, anxiety, fearcallit what you stable nursing staffs seemed to have short lengths of like. My thesis was that if we could improve the patient stay. A stable staff seemed to make for general level of communication in hospitals we efficient patient recovery. might do two things: improve staff stability and It was suggested that the hospital factor that reduce length of patient stay. Thiswas the hy- significantly accounted for this relation of effec- pothesis that appeared in the published report, tiveness to stability was the quality of the human Standards For Morale. 1 tested it outside of Lan- contacts within the hospitals. These contacts are cashire by examining 15 hospitals all over the of two kinds, factual and emotional; both are im- United Kingdonr. Scotland, London, Yorkshire. portant. A nurse may need to know something We found high concordance among five things. about a patient's treatment and so have need for One was the communication between the ward factual guidance. She may he afraid to ask for sister and the power structure; how approachable such guidance and thus need emotional support. the ward sister thought the doctors, the matron, When we tested these suggestions in the actual and the administrator were. What the clarity of study we found some interesting results, and I communication was down from the ward sister to believe subsequent research is beginning to suggest the nurses and the patients was another considera- the results we found there are probably universal. tion. We measured both of these in the 15 hos- The hospitals we could compare were similar in pitals.

145 PSYCHOSOCIAL FACTORS IN HOSPITALS AND NURSING STAFFING 139

NVe also measured the average length of stay of that improved communications could improve the all qualified nurses in the 15 hospitals. We com- attitudes of stall and the effectiveness with which pared this with the mean period of discharge of the hospital is run. surgical and medical cases for a limited number It was the view of the author that the improve- of rather simple. straightforward, uncomplicated ment of human contacts, both factual and emo- cases, like asthma, bronchitis, and pneumonia.Out tional, both operational and personal, was best hypothesis that good communication was in some achieved by action and responsible cooperation in way associated with staff stability and patient stay efforts to understand and to improve those con- was borne out and confirmed when we shifted away tacts where they existed. Since the HIC Project from the Lancashire semblance. started such attempts to improve human experi- All of this appeared in SfundardS For Morale. ence, the term "action-learning" has been intro- King's Fund Hospital Centre of London thought duced. This dual nature suggests that to under- it a good idea to follow up some of the statements stand a proposition one must be able to apply it made in Standards For Morale. The thought was in practice.

The HIC Project

NVe persuaded 10 hospitals to join a consortium. wheedling information out of people, measuring The expe.ment design was straightforward. We opinions. And during that period we sent them to chosefront each hospital two teamsofthree work for a week in a hospital other than their people. The senior team included the nursing own. superintendent, the chairman of the medical com- During that week they tried to put into practice mittee (senior doctor), and the hospital secretary- some of the skills, the efforts, the practices, the treasurer. techniques that they had learned while at the sea- The junior team consisted of an assistant to the side and tried to identify within a week what matron of some order, a deputy matron or a senior seemedtobe the problems of communication departmental sister, the hospital secretary - within their host hospitals that gave the hospitals treasurer, consultant, (but normally he was a reg- most concern. istrar, the level below the consultant). When they came back we had further confer- For the senior team, the top three, we ran a 3- ences. Each of the receiving hospitals hadbeen day conference. We talkedto them about the acquainted by its visitors with what seemed to be nature of internal communications in hospitals as the outstanding problems. The receiving hospitals we then understood them. We got together totalk chose particular projects that they would work on about their perception of how the hospital . held with their own team. For the next 3 or 4 years together as an organism, what the network of com- these hospitals among themselves selected quite a munication was that turned a large mass of people large number of projects that they worked on, into a staff. What was it about the hospital that sometimes with the advice of the teams that had gave it this quality of nurse stability orlength of visited them in the first place, sometimes with the patient stay that discriminated that hospital and help of a small team that I assembled at Guy's other hospitals? Hospital. Much more frequently they worked with Looking back on it, I can see we spent far too the help of students of sociology, nurses in train- short a time working with these people, the top ing, all kinds of people, university students who three people at the hospital. L.4t the junior team intended to be nurses or who were interested in spent a month together in a hotel at Hastings, and the sociology of hospitals. A very large numberof there we tried to go a little farther. We tried to people lent a hand with the study of what was give them a smattering of methodology: how to going on in these 10 hospitals. conduct an interview, to draw a flow process chart, to design and score a questionnaire. They inter- The Problems of Perception viewed eachothertopracticetheirskillsof This direct invocation of the psychosocial genre

146 140 RESEARCH ON NURSE STAFFING IN HOSPITALS of the hospital turned out to be the most evasive the solution of the problems, two things happened: and uncertain departure. A large number ofper- their attitudes toward the hospital as a whole im- sons at all levels of responsibility in the hospitals proved, and the system itself improved. and, indeed, in the central research team guiding They improved the communication system at the project, confessed that they never understood the level of attitudes. They improved. itatthe what it was trying to do. Only whena group of level of operations. They found that if people be- persons at a hospital really determined to see what can.e operationally involved in attempting to im- they could learn of their problems by energetically prove the system that was giving them anxiety, tackling them did the mission of the project be- they could, in fact, improve the system and they come clear, both intellectually and emotionally. could change attitudes toward the institution. Such,of course,istrue ofallpsychosocial phenomena. There must be some leaven to en- ergize the lump. Even reviewers of the books that Psychological Factors have appeared about it say its aims werenever made clear, despite the fact that these wereset "... the feeling of people concerned with this forth in writing for all to read. A resume of the study has been that a happy hospital is the best aims of the HIC Project is attached as appendix I. guarantee of a nurse completing her training. The Suchaction-learningprojectshavesincebeen practical problem is to achieve this by attention to started in many other places, mostly outside of the difficulties that beset sisters and the sisters' Britain. In appendix II the responses to such ef- attitudes to the junior nurses, by providing efficient forts in an American university are analyzed and nursing management, by helping the nurses to should be helpful to those concerned with action- organize an adequate social life, and by arranging learning in hospitals. support for the nurses when they have emotional Now, what did this effort actually achieve? I problems. Some of these are easier to do than think itis fair to say that where from the outset others, but a start has been made with allof the senior team in the hospital believed in the them... (1)." goals of the project, then the results were highly Here, in less than one hundred words, is the positive. The attitudes of the staff were favorably substance of this paper. More fully, it describes a changed and the project on which they worked series of efforts, made both within hospitals and achieved much of what itset out to do. Where outside them, to identify, understand, and resolve there was no confidence, where there was skepti- some of the human problems bedeviling the pro- cism from the start, then verylittlehappened. vision of patient care. Whether the results that Sometimes, in the course of the study, thepersons flowed from these efforts were promising as well concerneddoctors,nurses,and administrators as meager, or whether they were simply meager were converted. They saw what these notions. on because the approach was misconceived, is a matter communication really meant when they themselves of opinion. The efforts were either part of the became engaged in it. HIC Project or part of a series of monthly dis- The union of these ideas in studying communi- cussion groups organized at the Hospital Centre cation in this sense anti the scientific operational for nurses from hospitals in and around London. research development have produced quite a re- Both of these efforts were strongly supported by markable result. Where people had become in- the King Edward Hospital Fund for London and volved and were expected to contribute ideas to the first by the Ministry of Health.

Some Illustrations of Psychological Development

"Our results indicated that the attitudes and ministration, in turn, have some effect on ward opinions of the nurses in the ward and those of sisters and charge nurses. We attempted not only nurses in charge of wards are closely connected. to alter procedures and attitudes at ward level but Without a doubt, the attitudes of the nursing ad- alsoto involve the senior nurses more closely,

147 PSYCHOSOCIAI. FACTORS IN HOSPITAI S AND NURSING STAFFING 141

hoping for sonic modification or change in the (e) to enhance the public image of the hospital. tipper levels of nursing hierarchy. These nurses Following detailed explanations of the survey were involved in atleast 30 meetings to discuss by the HIC teams, comments by department staff different aspects of their work. These became in- members were solicited and 1,750 of these were re- creasingly uninhibited and healthy. At two meet- corded in 38 interviews. When agreement had been ings held with H1C team members to discuss the reachedas to organizationand as to physical or whole project with especial reference to the senior structural problems, these were referred to the re- nurse's involvement. many nurses spontaneously sponsible administrative, medical, or nursing pol- said that the meetings helped them to reflect on icy makers for action. In subsequent meetings with their own ideas and attitudes. Some gave examples HIC team members, many solutions were contrib- of how they had attempted to overcome problems uted by the staff and were put into practice im- in their particular areas, indicating some changes mediately. Redeployment of staff resulting from had occurred. How far attitudes changed throughthis has not only improved patient care but has out the hospital could he assessed by repeating the provided an improved training atmosphere for survey in the future (2)." student and pupil nurses. Nursing responsibility We give in this section a few illustrations of how has been more dearly defined, and discussions have the involvement in action-learning studies seemed beeninitiated concerning medical and nursing to those at the cutting edge of the hospital system. policy and procedure. A great deal of the success The examples are in the alphabetical order of must be attributed to the fact thatthe chairman name of the hospital from which they aredrawn. of the hospital subcommittee gave her support The firstwas written by the matron andher throughout the study, attending many feedback deputy at Edgware Hospital.t meetings and several presentations by HIC teams of other hospitals participating in the project. There are several outstanding items of interest Accident and Emergency Department in this account. First, the impending retirementof the head of the department offered theoppor- In response to complaints received about the tunityto change the system.Second, the exercise accidentand emergency department, many of was planned both tomodify a system and train them apparently justified, the HIC decided to in- the staff.Third, from the very outset it was seen vestigate. Finding that the casualty department and by the subordinate staff to be supportedby those in the casualty theatre were controlled by different established positions ofpower. Fourth, the mean- heads of departments (one of whom was about to ing of the term"communication"became clear retire) with followup dressings controlled by a only after vigorous efforts had been made toim- third, the team decided to perform an in-depth prove it. study of the problems, conducted in a way calculated to involve all grades of staff in the de- Hillingdon Hospital (3) partment. The team's objectives were as follows: "My own view, for what it is worth,is that the (a) to provide a comprehensive casualty service HIC Project suffered from toomuch pressure to despite the physical limitations of the depart- produce results by its techniques, whenin the hos- ment, pitals there may have been more pressingprob- (b) to establish continuity of treatment and pa- lems that had to be solved by othermethods. This, tient care, Ibelieve, was the case at our hospitalwhere, (c) to provide an improved practical compre- during most of the project period, we were con- hensive training fcr student and pupil nurses, cerned with the overriding task ofcompleting, the (d) to reduce patients' waiting time to a mini- equipping, moving into, and commissioning is only since the move mum, new hospital building. It has been completed that we havebegun to use Vick.Pauline, and Merchant, Barbara. "Study of the HIC techniques to solve a realproblem, the one Hospitals:Corn. AccidentandEmergency Departmert." staffdescribed by my col- municationJ, Choice and Change, 1972. The following account concerning nursing is taken entirely from this chapter. leagues.

148 142 RESEARCH ON NURSE STAFFING IN HOSPITALS

"There was a lack of definition of thepurpose 10 in the HIC Project. It is marked outas having at the Hastings course. It seemed to be a collection its own research units, includingone to promote of interesting lectures withan insufficient central operational research within the hospital itself.It theme, I believe that the hospital peopleon the is a unit with a wide reputation and, for example, course could have contributed more by cooperat- is lending its computer to the researchers into the ing with the research workers in devisingways in services for the mentally handicapped. which communication in hospitals might be im- Mr. Michael Fahey, the deputy housegovernor proved, and I think too much emphasiswas placed of the hospital and author of most of the section upon demonstrating the connections between poor of the report concerning the London Hospital,at communications and other short-comingsatthe the outset shared Mr. Bardgett's confusionas to expense of not devoting sufficient time to devising the objectives of the project. practical means of improving matters." "The hospital first became aware of the WC These are entirely fair and constructive criticisms Project through the King Edward's Hospital Fund of a poorly tinted effortto use what might in other for London. It is difficult tosay what members of conditions have beenreasonablysuccessful. A the team expected of the project,.since itspurpose major upheaval may not necessarily be the best was not entirely clear. It might well be that this time to try out new methods of treating old prob- was because they were not able to attend the two lems. The opportunityso apparent in the previous briefing sessions Meld at the beginning of the proj- example was missing here, Indeed, themove into ect. The mystery was not noticea!,te dispelled dur- new premises is seen as countereffective. Another ing tite progress of the month'scourse at Hastings hospital reported that thesame pressure was both where a similar confusion wasapparent. Members necessary and constructive. of the team formulated a working hypothesis that "One other aspect of the central team's role has the aim of the project was to instill methods by been of great importance to all teams. This was which to perceive and study problems withina the pressure exerted on hospitalteams to report large organization, and that the success of thisat- back on their activities andprogress. Without this tempt would be evaluated. In that event, this hypo- added stimulus many ofus would have done very thesis proved to he not too far from the truth." little, especially in the early stages. We stillre- At least one person has given a lot of thought member that after the training course at Hastings to the implications of these lines, and their lesson we were expected, within a couple of months, to ha5 been well marked fir future applications of state our intentions for the future. Howmany action-learning. The project chosen at the London teams would have any future had we been allowed Hospital aimed to improve the emergency admis- to proceed at our leisure (4)?" sion of patients. This, contrasted with the previous extract,em- It is of great interest that the London Hospital phasizes the profound differencesof perception should have tested the hypothesis underlying' the that can arise between colleagues engaged in sim- entire project, that responsible involvement in the ilar missions. The lesson for the psychotherapist identification, analysis, and treatment of perceived is to choose well his occasion. Finally, theobjec- problems should lead to changes of general atti- tion made by Mr. BardgettthattheHastings tudes towards working in the system manifesting course lacked.definition of purpose runs through the problems. In the words of the hospital'sown many commentaries. As is shown in appendix II, report, "There were highly significant changes in it is a very general objection. Butwe note that the attitude in all those areas where alterations to the staff ,Edgware Hospital grasped what communi- system had been made. In marked contrast, in the cation meant when they triedto improve it. There one area where no changes had been carried out is, unfortunately, no royal roadto explaining it .. .opinion remained remarkably stable." in advance. Our next illustration makes thepoint more clearly. North Middlesex Hospital The London Hospital The report of Mr. Alderson, the assistantgroup This was the only teaching hospitalamong the secretaryatthis hospital, shows the difficulties

149 PSYCHOSOCIAL FACTORS IN 1-10SPITAIS AND NURSING STAFFING 145 that confront those vho ny to stimulate action- ately relevant evidence, that their everyday opera- learning projects for the first thne %vit bout being tions can be seen as learning projects may prove able to illustrate their goals with visible evidence insoluble. Without such belief no opportunity can of success elsewhere. be engineered to give the hypothesis the more- "Perhaps the outstanding feature of both our than-favorable launching thatitwill always de- earlier studies was the general lack of enthusiasm mand. on the part of the staff concerned, comb:ned with a certain amount of skepticism and even suspicion. Several ls Hospital . . . Perhaps what we need, to achieve greater staff participation, is the completion of a survey show- We have quoted from the report of this hospital ing tangible proof of success, for itis on results at the head of this section. Mr. D. J. Dean stun- that are of obvious and lasting benefitthat the marites the total experience. "Not all of those as- majority of staff judge our efforts in this field. . . sociated with the project felt able to contribute "So although our use of HIC methods has not or become involved to any great extent. Much of been very extensive, we are not entirely disen- the workload fell on one or two members together chanted, but would prefer not to undertake more with a university student. For the efforts of people than we can cope with, even though the help of involved in such a project to be successful, itis expertsisat our disposal. As to the future, we essential that they not only feel themselves to be a have hopes of being able to continue to do some- team but are seen as a team by the hospital as a thing to meet staff wishes in the catering field, and whole. For this reason the initial stages of prep- there are plans for a small departmental project aration are of crucial importance. upon which our thoughts have not yet fully crystal- "If this form of mahagement is to succeed, the lized. hospital managers must be fully conversant with "Beyond this, we shall continue along the path the ideas behind HIC and be involved fully in already mapped out, but perhaps with diminished any projects that are undertaken." enthusiasm and no longer any great belief in the The physician superintendent of this hospital, theory that a problem has only to be evaluated to now at Rochester, New York, helps toclarify what be solved." the HIC Project should have done and very much NIr. Alderson was not alone in his views. Dr. in terms of psychosocial factors in nursing. G.S.A. Knowles, the senior medical member of "Communication occurs in two principal forms: the team, had equal difficulties. are "I must admit that I joined the project with (a) Propositionalfactsandpropositions considerable skepticism. The stream of verbiage transmitted, (b) Emotional feelings and attitudes are trans- emanating from the central team overwhelmed me both quantitatively and qualitatively, and it did mitted. "I hope the Hospital Internal Communications not matter how oftenIreadit,Istill did not understand it properly: the individual words had Project will highlight for readers the importance meaning, but when strung together,they com- of people timethe way people's opinions are listened to and dealt with or ignored the way their nunicated little. . . . feelings are ruffled or respectedthus producing "As far as I am concerned, there have been few positive results to get excited about so far. Staff the enthusiastic and willing worker or resulting in of all grades tend to be uninterested or suspicious. the sour deadbeat. "Object time, on the other hand, is time that is Unless there is some money available to make the spent dealing with things, and although account necessary changes that are suggested by an investi- ants, carpenters, and committees can deal with gation, further effort is a waste of time." material things easily,they concern aseparate Dr. Knowles speaks here for a large number of dimension of hospitalfunction and, when the his colleagues, for many senior nurses and, pre- chips are down, the least important one. sumably, also for a majority of hospital adminis- "If the simple platitude that tr2.s matter and trators, The problem of convit.7ing senior mem- people count is again emphasizf u and if a more bers of the staffs of institutions, without immedi- reasonable approach to the happiness and well- 144 RESEARCH ON NURSE STAFFING IN HOSPITALS being of staff and patients is observed, then this evident that during their training many nurses study will be most worthwhile andmy own .support rapidlydevelopedfeelingsofresignationand for it well rewarded." hopelessness about the possibility of changing anything, or of obtaining redress, and these feelings West Middlesex Hospital were principally responsible for our scheme not being conducive to personal counseling. The team at this large hospital tackledmore "At the discussion meetings, the feeling of people problems than did the teams atany others and concerned with this work had been that a happy made'a %Aim effort to get to the heart of their hospital is the best guarantee of a nurse completing serious problems of nurse staffing. her training. The practical 1...oblem is to achieve "At the inception of the HIC Projectwe were this by attention to the difficulties that beset sisters undergoing a testing time over certain nursing and to sisters' attitudes to the junior nurses, by problems, and like everyone else wewere short of providing efficient nursing management, by help- nurses. This provided apowerful stimulusto ing the nurses to organize an adequate social life, ameliorate the lot of junior nurses,a first move we and by arranging support for thenurses when believed,inreducing nursewastage. ...The they have emotional problems. Some of these are topics discussed by the nurses emphasized their easier to do than others, but a start has been made preoccupationwiththehospitalenvironment; with all of them." they fell into six main categories: lack of social life and activities, defects of communication, personal problems, professional matters, Conclusion patients, and simple administrative deficiencies. .. Such, then, are some perceived results of the "Interesting manifestations were observed in the HIC Project. They suggest that, given the interest groups. Young girls in their first year of training and support of those in positions of authority, the discussed in a mature fashion defects and deficien- nurses and other staff within the hospital system cies in the hospital that affected their lives, and may work together to improve both the function- although they .felt strongly about some of their ing of that system and their attitudes toward their complaints, they discussed them with moderation work as a whole. We shall see that the concept of and insight, showing an ;-4pressive perception of autonomous learning, or of self-development from the hospital structure and of the feelings of their the group examination and discussion of one's superi. rs. own problems and progress, whatever attention it "Or. the other hand, sisters and consultants, even had attracted in the past, still deserves to be more those taking the nurses' discussion groups,were deeply developed as a form of psychosocial ther- see unapproachable Outside the groups. It was apy.

A Questionof Attitudes

During the 1960's the interestin the human "It may not be an over simplification to state problems of British hospitals was heightened by that the care of patients in hospitals . , hinges the publication of a number of criticalbooks, on the two main factors. The first is the nature, some of which were widely read. Following one of quality, and amenities of the accommodation utey these, Sans Everything, a group of people at The occupy, and the second is the morale of the staff, Hospital Centre of the King Edward Hospital Fund involving as it does their personal attitudes towards for London agreed to stage a st ties of discussion patients in their care. the latter the more im- groups for nurses from both general and psychiat- portant, but inevitablyitisinfluenced by the ric hospitals in and around London. The aims of former." these were suggested in the following paragraph We may observe in passing that all researchers from a report of the East Anglian Hospital Board. into the subject have revealed highly significant

151 PSYCHOSOCIAL FACTORS IN HOSPITALS ANI) NURSING STAFI ING 145 differences, among otherwise comparable hospitals, to the hospital that precipitates in visible social in the measures of their nurses' attitudes towards actiontheslower changes detectable,butnot their work. Itis the existence of these differences obviously so, in the analysis of attitude change that suggested the HIC Project. Cannot a process reported below. of organisational learning be set in train so that, Our ward sister concludes: "I would not like to in the hospitals with the less favorable attitudes, give the impression that discussions of this type something is done to correct the conditions out of cure all illnesses of the ward, nor that they dramat- which they arise? The group at The Hospital ically alter our attitudes towards each other and Centre was less ambitious and confined itself to the patients. But one thing I do know; discussion organizing a series of monthly discussion groups, and frank appraisal like this can make a consid- eventually spread over 4 years, that gave hospital erable dent into this problem of attitudes. . .It nurses of all ranks the opportunity to discuss their isworthwhile continuing, andIfeelthat any problems andtogive eachother supportin ward sister would gainif she were prepared to ameliorating or enduring them. give the time, energy, and thought to initiating The original invitation to join the group had such a group on her own ward. I warmly recom- in it these words: "There can be little doubt that mend it to anybody really concerned about people, the attitudes of nurses towards their patients and ether patients or visitors, or whether our col- towards those with whom they work are an im- leagues who work alongside us each day." portant element in the standards of patient care This woman once more demonstrates what some achieved. Very littleis known about the nuncs' of us have long known and what a few in the HIC understanding of their patients' needs, nor do we Project actually learned: that one can solve prob- know enough about their reaction to the day-to-day lems both if one wants to solve them and if one problems of ward administrationasitaffects recognizes that, in so emotional a situation as a patients and staff." hospital ward, one may in oneself be a substantial The illiscussions set out to develop both under part of the problem. standing and knowledge. To what extent they suc- The fundamental question remains: Why do ceeded is hard to define, as responses were more some senior nursing staff think as didthis ward than contradictory. A rigorous attempt to measure sister, and why do others regard not only the attitudes during and after one series of seven dis- HIC Project but the discussions organized at the cussions showed highly significant changes for the Hospital Centre as all a waste of time? Even more, better. A ward sister who did not cven attend the why do some regard the discussions as destructive? meetings but read of them in the nursing press The following extract from the Hospital Centre was so inspired that she set upin her own ward record shows how difficultit may be for some her own discussion group. She had been 12 years senior nursing staff to change their attitudes: "This rein charge of an acute male rnedli.41 warn of 26 debate... wasthe culmination of 'several beds staffed by two consultants, three housemen, quests for students to speak up; the result was and about 15 full-time and part-time nurses, aux- personal attack rather than reasoned debate on iliaries, and domestics. the points the students had raised.. .That it She said, "In the Kings' Fund Reports I was was upsetting there is nodoubt. How much so particularly impressed that frank discussioncould can be gauged by a principal tutor apropos a modify attitudes. I decided to initiate ward meet- student with things to say: 'Empty vessels make ings. The first problem was time. A ward sister the most sound. I have often found that the most has little time on her hands: thisis the firstat- vocal student verges on the psychopathic.' " titude to throw out. One has to make time. Want- ing to do it means one has to find the time. Having TheCaKteasts Revealed made up nrte's own mit-id, then it is possible to carry it out. I have proved this." It would be hard to find two attitudes more As I see it, this is the key to all attitude change apparently opposed than those of the wardsister and to all learning. Our problem is to know what who was convinced of the value of frankdiscussion it is about this particular sister and her relation merely by reading of the Hospital Centre's efforts 1.5Z 146 RESEARCH ON NURSE STAFFING IN HOSPITALS

in the nursing press, and those o{ the -principal Nursing Discussion Groups and Attitudinal tutor, that guiding light for the future, who iden- Changes tifies a willingness to ask questionsas a mark of the psychopath. Yet the contrastappears again At the end of the first monthly series of dis- and again and at many stages. cussion groups, lasting 7 whole days inall and The group discussions at The Hospital Centre attended by 82 nurses, Dr. Toni Caine, Principal leave us in no doubt about the immenserange of Psychologistat Claybury Hospital, undertook a personalities involved. Itis as if some nurses are modest evaluation of their effects. He prepared, so accustomed to authoritarianism and depend- from a content analysis of topics raised in such ence that no efforts of a psychotherapeutic kind groups, a questionnaire with five broad .categories. can ever change them, atleast not in the par- One of these could be described as "Getting in- ticularhospitalsettingsinto which theyhave volved with the patent." Its termswere "A nurse grown. The point is made by Dr. Tom Caine in should take care not to show too much interest his summary of the first discussion series. inpatients' deeper problems in order to avoid "We have seen that the differencesare in terms getting involved.. .. of attitudes to discipline and organization, atti- Patients should be discour- tudes to the required degree of personal involve- aged from developing feelings towards staffmem- ment with the patient, attitudes to a formal versus bers.. . .Staff being too friendly towards patients an informal approach, attitudes to free communi- makes for poor discipline on the ward. . . Pa- cation and a questioning of the fundamental scien- tients should not callnurses by their Christian tific status of work, including the value of the names." formal diagnosis. We areallaware thatthese Another man sectiontested the nurses' views questions form the cornerstones of the training of of their relations with the doctors, A thirdsec- nurses. A nurse trained in the beliefs and attitudes tion was on discipline, cleanliness, and efficiency. of one type of nursing, indeed ofone institution, About two-thirds of the total number ofnurses may find that she is required completely to reverse who went through the seven monthly discussions them on transferring to another. were able to complete the questionnaire twice. The "My own feeling about the conferenceon patient care and our analysis of the data is that we second test was given several months after they have thrown up fundamental problems without had returned to their hospitals. yet finding the answers. . . . All patients awl ell Score comparisions were rigorous. The usual staff are people and not just diagnostic categories five-point scale of strongly agree, uncertain, dis- or nursing grades. The full recognition of this is agree, and strongly disagree was used, but in seek- possibly the most therapeutic attitude of all. Per- ing change of attitudes only reversalswere marked. haps the most important thingsare to find out In other words, changing from stronglyagree to how we lose sight of this truism, howwe can agree was not considered sufficient indication that relearn it and, having done so, whatwe can do there had been any shift of attitudeat all, only about it," disagree to agree or viceversa would be admitted. All I feel able to add to this is that Dr. Caine's observations apply not only to nurses in British There was a highly significant (below 0.1 percent) change' in attitudes toward hospitals but to all the other huniangroups, in more involvement with allfivecontinents, with whose action oriented thepatients andlesssubservience toward the education I am now concerned. As Henry Wallace doctors. Indeed. for both classes ofnurses, general observed 25 years ago, this is the Century of the and psychiatric, therewere improvements in at- Common Man. Whatever his sexmay be, he is titudes in all of the five broad categories recognized at least demanding to be treated as an individual in the questionnaire and basedon the content person. analysis of the discussions.

153 PSYCHOSOCIAL FACTORS IN HOSPITALS AND NURSING STAFFING 147 Appendix I

The Aims and Intentions of the HIC Project

1. Normally the demands. for order and economy quite significantly, in collecting data and in dis- oblige hospital staff to develop procedural rules cussing the meaning of the data and theirrel- for overcoming operational problems. If no precise evance to the tasks of those involved with them. rules existitis expected that those in authority Beyond this committed core there should he a will use precedent or judgment to settle whatever supporting force of partisans, not deeply involved need may arise. Even if the powers and duties of but willing to give help when asked forit and staff have not been set forth in detail there will thus ready to be involved should the project be exist within every hospital an expectation as to extended into theirterritory. Beyond these,all who should handle the emergent problem. other employees in the hospital should atleast 2. In such conditions staff tend not to examine know that the project is going on. the origins of problems, only to allocate responsibil- A well designed project should have the follow. ity for dealing with them or to attribute blame ing qualities: should they not be dealt with. Thus there is little (a)It should demand the cooperation at all stages true learning,certainly no education,thefirst of a group of persons whose tasks are inter- objective of witich is to encourage the ability to locked. A study assigned to one or two indivi- ask those questions most likely to lead to the solu- duals, to produce a report for others to imple- tion of a problem. Education aims to develop the ment, isuseless. The choice of project,its general capacity to inquire, to show people how to depth and scope, the methods of study to be Let about finding out. Since, it seems, the principal used, the forms of coordination between in- defect of many hospitals is a lack of awareness of dividuals, the nature of any report to others their internal problems, education inthis sense outside the group, the methods of putting is of great importance. The project aims to provide into effect the findings of the 'project, and so it. forth, are all essenti:-.1 subjects for agreement 3. We are, as I see it, primarily trying to help between the team leaders and those whose those in charge of hospitals and of their services work will be woven into the study. at all levels to use better their own resources in (b) It should demand not only cooperation at all identifying and tackling their working problems. stages within the group but also an appeal to We choose this approach both because we believe sources of information or ideas outside the that these are the greatest resources that the hos- group. These may at first he limited to purely pitals have available and because we also believe technical affairs, such as asking for help from the hospitals possess at alllevels a considerable the WC team on methods of interview or of fund of latent goodwill and ability. Without ex- data classification. Or it may turn out that the ception, all research shows that, given the organ- working group, in order to make real prog- ized opportunity of working through a problem of ress, will need to share the experiences of concern to them personally, those who serve even another hospital or even to inquire about the in the most stressful conditions and who may have treatment of similar problems inindustry. seemed most incapable of coping with them can For this reason the hospital team should, at riseto unexpectedheightsofinsightand a fairly early stage, be in touch with any local resolution. college of technology of commerce with a de- 4. Any firstproject should on this account he partment of management or administration. !wen as a growth point, as a kindergarten in coop- (c) It must be limited in time; from the outset eration. It should involve few people, say up to 10, there should he agreement as to when par- ' Rem+. R. W., Hospitals: Communication, Choice and ticular stages of the work should he com- Change, 1972, pp. 130-132. pleted. Once such a project is started the 154 148 RESEARCH ON NURSE STAFFING IN HOSPITALS

interest aroused can be so great that any ing out an idea that seems to him to contri- waiting time may lead to disillusion and hos- bute to the definition or solution of the tility in the very places where enthusiasm and problem, cooperation have once been generated. (d) to sense the effect of his words or actions: (d) It should, when completed, suggest another he learns how this behavior has affected others project, elsewhere in the same hospital but and thus can test the adequacy of his self- involving largely different persons. understanding. 5. In work on the project, the framework of (e) to generalize from this learning: in the proj- authority should be temporarily thrown away. ect he will begin to sense the accuracy or Those involved in it are faced with a new situa- otherwise of his own impressions and of those tion; how to define the need, how to collect in- of others; he will see how he and others formation how to work with others, how to measure respond to different attempts to influence and progress ancl, above all. how to employ their own to be influenced;:.ow differing goals and abilities. It is as if, by stripping away the hypotheti- desires,if contrasted and evaluated freely, cal framework of authority, we are trying to under- leadto learning and to improved understand how individuals think and act as the persons standing of communications. they are. They are being given an opportunity, 7. Itfollows that to be successful the groups like children whose teacher has suddlnly left the must be highly informal and the imposition of the classroom,to compare notes among themselves traditionalframework of authority upon their as to how a problem can best be solved. project must be avoided. I have little fear, now 6. The project should teach each who works that I know some of the senior staffs of the hospitals on it at least the following: personally, that such imposition will deliberately (a) to observe how the group respond to the occur once experience of open discussion has been originally understructured situation, gained. Nevertheless, its removal must be accom- (b) to question what behavior on his part will plished by slow degrees; senior staffs must gain best help himself or the group, confidence in free exchanges before subordinates. (c) to act, whether in words or otherwise, of his In this I feel both the HIC and the hospital teams own volition, as in collecting data or in try- have a role first to learn and then to enact. - PSYCHOSOCIAI. FACTORS IN HOSPITALS AND NURSING STAFFING 149 Appendix II

An Experiment in Active T earning

In 1971 I was approached by the dean of the structure and available options ofthecourse.

Business School at Southern Methodist University, . . . "Fromm's (1941) discussion of freedom may Dallas,who, with the support of somebut a be used to help understand the variety of student minorityof his faculty, was endeavoring to change responses to the course. Fromm distinguishes be- the principles upon which the school was run. In tween freedom from constraints and freedom to the words of Roger Dunbar and John Dutton, two develop facilitating structures that allow individual of the dean's active supporters: achievement. The absence of many of the expected "AttheSchool ofBusinessAdministration, instructor demands provided freedom from many Southern MethodistUniversity,vigorous efforts of the usual course constraints. HoWever, students have been made to educate students for an active often showed little freedom to develop new be- role insociety. As a step inthis direction,all havioral responses to utilize the new freedom from undergraduate course requirements but one have expected instructor demands. Some got angry be- been abolished. Thus stw'ents now must design cause they did not perceive the instructors as will- their own learning prc.,am instead of fulfilling ing to tell them what to do and a number remained a course sequence dictated by the faculty. In this antagonistic throughout the semester. Others be- way an attempt has been made to transform the came angry because they could not reconcile the school itself into an environment which challenges freedom that did existinthe course with any the student to act effectively. One required course constraints whatever; they may have felt that the remained as a setting in which to force students requirement that they examine. the consequences to consider what isinvolved in designing one's of their own choices contradictedtheidea of own learning experience. This requirement ap- freedom to make choices. peared necessary to the faculty in order to assist "Still others, originally skeptical, eventually saw students to adopt a more Active style of learning the purpose of the course. Frequently, these in- behavior, for students at the school tended to be- dicated they only regretted it had taken so long have passively in learning situations, waiting for in- to become aware of the objective. Finally, those who structors to generate material rather than seeking were already highly motivated to learn reported information for themselves." that as a result of the course they were even The experiment soon ran into some of the prob- more enthusiastic about their work at school. In lems of the HIC Project: a lack of intelligibility summary, responses to the course appeared to vary followed by resentment. In the words of its or- gteatly and the instructors believe they probably ganizers: reflected whether students were able to formulate "From the instructor's point of view, the purpose new behavioral responses to cope with new de- and the design of the course were clear enough mands of the learning environment." and reflected their beliefs as to what was necessary Nothing could describe our experiences with for student learning to occur. Specifically, for stud- the HIC Project more faithfully. M at S.M.U., ents to learn, they would have to be active and moreover, there were members of the staff who responsible for their own behavior. However many were unable to grasp this idea of freedom to un- students felt lost and confused with this unfamiliar dertake one's own learning. structural design. Some responded with curiosity, exploring what could be done in the new situation. But others became passive, sullen, or even violently Two Extremes of Students angi y because they could not understand what was expected of them. These disaffected students seldom In the HIC Project no effort was made to dis- found helpful the extensive written handouts discriminate between two types of hospital staff as tributed inthe course concerning the purpose, the S.M.U. studies discriminate between acceptors

156 150 RESEARCH ON NURSE STAFFING IN HOSPITALS and rejectors of the new learning program. Some was a significant positive correlation between learn- of the central team were awareas, indeed,is ing and a desire to be noticed by the teacher. shown here in the main paperthat some of the In the light of our HIC Project experience, these hospitalstaffshad more difficultytitanothers results are very plausible. Both the Severalls and itsgrasping the notion of self-directedstudies. the Middlesex Hospitals stress, moreover, that the But at S.M.U., where conditions for quantitative attitudes of the junior nurses ultimately reflect evaluation were possible, a deliberate search was those of their seniors (a result also reported in made toidentifythe extremes among the 412 Standard For Morale). It therefore looks as ifour students who took the course. Of these there were first task is to persuade those in the senior posts chosen 32 manifest acceptors and 32 manifestre- of the hospitals to act as manifest acceptors. Any- jectors ofthis new method: the discriminating thing effective in the way of change will demand, criterion was the persistence with which they waited at alltimes, the support of those in authority. for guidance from the teacher and abstained from In the many programs of institutional change experimenting with their new methods. The mani- with which I have been concerned since the HIC fest acceptors neither waited nor abstained; the Project, I have been more than careful to involve manifest rejectors did both. the coalitions of power responsible for the partici- It was then shown, as a result of a survey in- pating institutions. Not enough attentionwas given volving a sample of 238 students of which the two to this in launching the HIC Project. In our latest sets of 32 are therefore about one quarter, that attempt, to stage a development program for the therewerecategoricaldifferencesbetweenthe nationalized industries of Syria, we have not only two sets. Among the manifest acceptors there were secured the personal interest of the Prime Minister significantpositive correlations between learning but allocated to him an operational role and a time- and efforts to assess the feelings of the other stu- table to go with it. Since the program was designed dents and of the teacher. Learning among the mani- at his invitationhaving heard of its success in fest acceptors was thus coupled withan awareness another Arab countrywe may he confident that, of others in the group. Among the manifest re- should our efforts fail, it will not be for lack of high jectors there was no such correlation: instead there level interest. Impact of Computerized Information Systems onNurse Staffing in Hospitals

Donald W. Simborg, M.D. Chief, Clinical Systems Development Johns Hopkins Hospital Baltimore, Maryland

Introduction

To describe the impact of a technological de- From these four items of information, we can development such as computerized information sys- termine the nursing staff needs on any nursing tems, it is necessary to define and understand the unit at any time such information is available, problem area to which it is applied. Nurse staffing To answer the second question, how can the in hospitals is a broad and complex topic. For assignments of personnel be made tofitthose the purposes of this discussion, nurse staffing will needs, we must have the means to provide the refer to the following two problems: items of information sought in the first question How many of each type of nursing personnel on a timely basis and must then have the ability are needed on inpatient nursing units? to optimize the distribution of available staff ac How can the assignments of personnel be made cording to those needs. This optimization can oc- to fit those needs? cur on a short-term, dynamic basis, or on a long- I would like to discuss what impact computer- term basis. The advantages and disadvantages of ized information systems might have on these com each approach will be mentioned later, I would ponents. To answer the first question, how many like now to discuss how all of this might be ac- of each type of nursing pesonnel are needed on complished and what role a computer system might inpatient nursing units, four items of information play. are required: We have not solved the nurse staffing problem (1) What kinds of tasks need to be done by the at Johns Hopkins. Howeve:we have partially nursing staff? solved a related problem that I think will provide (2) How many of each task are there to be done some basis for solving the staffing problem. I would on any given shift on any given nursing unit? like to digress for a while to describe some of the (3) Which nursing personnel type should do each work we have been doing the past year. task? Approximately 2 years ago we became inter- (4) How long does it take to do each task? ested in the general problem of management of

151 1b8 152 RESEARCH ON NURSE. STAFFING IN HOSPITALS an inpatient nursing unit. What is the process by instance, an attempt was made to determine the which the nursing stall carry out their duties, how exact point on the flow chart of that order that ac- well does this process work, and what might be counted for the failure. As well as could be de- done to improve it? In general,nurses do two termined, 14 percent of the problem could beat- categories of tasks: those that the doctors tell them tributed to the physician's original order, 4percent to do and those that they decide themselves need to transcription of the order, 22 percent to written to be done. Let us first look at what the doctors communication of the order subsequent to the tell them to do; i.e., the carrying out of physicians' initial transcription, 58 percent to the execution written orders. (actually nonexecution) of the order by the nursing At our hospital. the nurses use the traditional staff, and 2 percent to notification errors. (Noti- kardex system of carrying out orders. Thereare fication refersto those specific ordersin which about a dozen different kinds of orders eachcar- the physician requests to be notified of a certain ried out in a different way. As an example, figure 1 finding if it occurs.) shows what happens when a physician writesan Another study performed was a traditional nurs- order to have a patient weighedon a periodic ing activity study in which random observations basis;e.g., every other day. The basic steps are were made to see what nurses did with their time the writing of the order, the initial transcription, while on the ward; i.e., when they were notsup- the written communication onto some working posed to take a break. Our results are similar to document,the executionofthe order by the other such studies and are shown in tal,le1. A nursing staff, and the communication of there- nurse spends about 30 percent of her time at the sult back to the physician. Figure 2 shows this bedside and about 40 percent of the time doing in- same process for a vital sign order; i.e., the order direct care activities such as paperwork, supply to measure the pulse, temperature, respiration rate. functions, and preparing medications. and/or blood pressureat some giveninterval. From these observations of the problems it was Figure 3 shows this process for an intravenous possible to determine some requirements for their fluid order. solution. This process of determining functional The point of these figures is simply to illustrate requirements to solve problems is9 creative or the complexity of our manual "system" of carrying synthetic process, whereas everything' prior to this out orders. Of the12 categories of orders we point was an analytical process. Thus, as in any studied, no less than 13 different documents are creative process, there is tremendous latitude as to used to keep track of these orders. One naturally the types of solutions that could be derived, and asks, "Why was the system designed this way and there is considerable judgment as to the initial how well does it work?" It is clear that the system solutions to try. was not designed atall.Rather,itevolved.It For each problem identified in the analysis phase evolved over 60 to 70 years' with changes a number ofpossible alternative solutions were frequently by different people in different decades considered. The details of this process will be re- to meet certain expediencies. ported elsewhere. The results ofthissynthetic In answer to the second question, how well does phase were aset of14functional requirements it work, we made some observations. Duringa 2- for a new system. These are shown in figure 5. Let week period on one acute inpatient medical ward. me call your attention to requirement 6, to allocate all physicians' orders were examined. Each order nursing staff dynamically according to needs. This was read and evaluated,thetranscription was is the subject of this entire conference, yet note read,thewritten communication ontovarious that itis only one of 14 requirements we had for documents was checked, and on site observations solving the problems of managing our wards. This were made to determine whether or not the order particular requirement is related primarily to one was tarried out. This was done 'for the nonmedica- of the problems found in the analysis phase, the tion orders only. fact that 58 percent of all orders not carried out Figure 4 summarizes the results of these studies. were due to the lack of execution by the nursing Of the 1,592 observations made, about 15 percent staff. of the orders were not carried out. In each such One reason the nursing staff do not execute

, 159 IMPACT OF COMPUTERIZED INI.ORMATION SYSTEMS ON NURSE STAFFING IN HOSPITALS 153

Figure 1.Flow chart of weight order. BEST COPY AVAILABLE

Weights

Order Reviewed by RN

Pink ticket Treatment if special kardex

Weighed by To weight book special each morning

Recorded in Nursing staff weighs special chart each morning as per weight book

Recorded in weight book by nursing staff

Transferred to graphic sheet by ward clerk

160 154 RESEARCH ON NURSE STAFFING IN HOSPITALS AVAILABLE BEST COPY Figure 2.Flow chart of vital sign order.

Order Reviewed by RN

Treatment Pink ticket for Pink ticket card file TPR Sheet BP,special for notifica- TPR ,frequent tion TPR J

Page made in Nursing staff To bedside for BP book reviews and special nurse pink ticket(s) performs

Physician Nursing staff Special nurse notified if records on TPR performs necessary sheet and in BP book

Ward clerk Special nurse graphs records on special sheet

T=Temperature, P=Pulse, BP=Blood Pressure, R=Respiration, RN=Registered Nurse

161 IMPACT OF COMPUTERIZED INFORMATION SYSTEMS ON NURSE STAFFING IN HOSPITALS 155

Figure 3.Flow chart of intravenous fluids order.

BEST COPY AVAILABLE Intravenous fluids

Order Reviewed by RN1 t

Transcribed Call to.I . A .Mnurse Occasional to 4-part therapy if prepares order to I .V. ticket start needed I .V. work- kardex sheet

Gold copy 3 parts to I .,/ .s re- remains on I .V.therapy ported in ward nursing report

Checked Yellow copy White copy Pink copy Bottle hung PRN for billing for adminis- to pharmacy by nursing (infrequent) I .V.therapy tration I .V. for additives staff therapy

Thrown Solution Solution Bottle away in sent to sent to regulated 1-3 days floor (if floor with by nursing no additives) additives staff and with pink and type - I .V.therap tick& label written label

.ImisharrirsomrrippiiPatient billed

16 .0 BEST COPY AVAILABLE 156 RESEARCH ON NURSE STAFFING IN HOSPITALS

Figure 4.Physicians' orders study 1,592 OnWard Operational Observations 234 FailuresI5% I.Physician -14% 2. Transcription-4c', 3. Written communication-22% 4.1.1xecution-513% 5. Notification-2%

Figure 5.Functional requirements of a new system to carry out physicians' orders

I.Providea legible.convenient summary ofthephysicians' current orders ina place where heislikelyto look at themdaily. 2.Provide arot iew process atthe input phase of physicians' orders in which an RN verified the 'input. 3.!lave physicians write orders on renumbered order sheets. Provide a mechanism to account for every order number, 4.Eliminate manual communication. 5.Schedule nursing tasks so as to reduce peak workloads. 6.Allocate nursing staff dynamically according to needs. 7.Provide a single document from which nursing staff can see at a glance all tasks needed to be performed at any given hour. A.Prot ide a mechanism by which the head nurse can monitor what tasks need to he done in the coming hours as well as what tasks were not carried out during the previous hours. 9.Indicate notification orders in the place where the specific patient recording must be made. 10.Make all recorded patient observations at the bedside at the time of the observation. II.Avoid transcription of recorded patient observations. 12. Summarise allrecorded patient observations in a convenient format in a single location for the physician to review at the time of morning rounds. 13. The nets. system should require less nursing staff administrative time to maintain than thepresent system. 14. The costs of the system must be appropriate for the benefits provided.

16J IMPACT OF COMPUTERIZED INFORMATION SYSTEMS ON NURSF STAFFING IN HOSPITALS 15i orders isthat they are too busy, hence the need output is listed in figure 7. Again, 1 do not wish to to staff appropriately. There are many other tea._describe the entire system but just that part which sons why nurses 10 not carry out their tasks that may be relevant to nurse staffing. The most im- would not be solved by appropriate staffing, hence portant information in this regardisillustrated the need for other requirements. I would like to call by one of our output documents called the "Team ymn attention to several other ofthe require- Action Sheet." An example is shown in figure 8. ments on the list,particularly7,8,13 and H. Our unit consists of three nursing teams each with We needed to provide the nurses with a simple approximately. 10 patients. For each hour of the way to know what tasks were scheduled to be done day for each team,the action sheet showsall at any given hour of the day. This is again re. tasksthat must he performed on any oftheir lated to the execution problem, since some orders patients at that hour as determined by the orders. are not carried out because people forgot and others This output document, as well as all others we because some.people may never have become aware produce. was not designed to solve a nurse staffing of the need to carry out the order inthefirst problem. The primary goals of this system were place. Also, there is no easy mechanism for the as follows: head nurse on the floor to monitor or supervise (a) reduce errors in transcribing and communi her staff. Finally, any new system that we implement eating physicians' written orders, should reduce nursing. staff indirect care time, or (b) reduce failures in carrying out appropriate at least not increase it. And, of course, the hospital nursing tasks, must be able to afford the costs of the system. (c) reduce nursing staff administrative time. With these and other requirements in mind, a new system to carry out physicians' orders was We evaluated the system on the one nursing designed and implemented on one of our acute unit by repeating the studies I showed you earlier medical nursing units. This is a computer based on afloor with the traditional kardex system. system that I will describe only briefly. This sys- Table 2 summarizes these repeat studies. As you tem meets all of the requirements listed except can see, each of the original goals of the system that we do not yetstaff our floors dynamically has been met. according to needs, nor do we reschedule nursing Now does all of this relate to the problem of tasks. We feel that we have the means to develop nurse staffing per se? Can this system and other a dynamic staffing system based on what we are now computer information systems have any effect on doing. that? To answer these questions, let us look again Figure 6is a diagram of our current system. at the questions which we said earlier neede:! to Twice a day we produce output documents. This be answered.

Number of Each Type of Personnel Required

Kinds of Tasks To Be Done develops a nursing care plan as usual. In addition, however she writes a set of nursing orders in the We can see From the output shown in figure 8 order book just as the physician does so that re- (and some other output documents not shown) minders will appear on the Team Action Sheet what kinds of tasks nurses do related to physicians' to perform those tasks as well as the physician written orders. The nursing staff, however. perform generated tasks. Table 3is an exmple. If the other tasks or duties not related to the care of the patient has a nasogastri tube in place, the nurse patient such as bed changes and patient hygiene. will write the older "NC Tube Care." This will Other tasksthat the nurses initiate are specific automatically generate the message shown in table for the particular problems of their patients. 3 at the time indicated. We have looked atthesetasks and have in These messages and times were derived by the stituteda procedure in which the nurse makes nursing staff as minimum procedures consistent an assessment of the patient's nursing needs and with good nursing care. Nlodifiation of the "stand-

1 64 BEST COPY AVAILABLE 158 RESEARCH ON NURSE STAFFING IN HOSPITALS

Figure 6.General flow chart of new system.

Physician writes Original copy Unit Clerk order on pre- remains in manually numbered order order book action sheet sheet if necessary

Carbon copy Order entered Hard-copy record collected by onto terminal of input produced Information Clerk on teletypewriter

Orders transmitted Punched paper from paper tape tape produced over telepnone to computer

Orders c.ciF.ed by Copy of computer accepted orders program given to nurse for verification

Errors returned jCorrections Output documents to Information retransmitted 1101.produced by Clerk to computer computer

Figure 7.System output documents I.Action Sheets A. By Team B. Special Patient C. By Function I.Diet List 2.Utility Room List 3.Weight List 4.Specimen List 5.Intake and Output Sheets 6.Patient Lists 7.Tetnperature.Pulse Sheet

II,Standing Order Sheets . i63 IMPACT OF COMPUTERIZED INFORMATION SYSTEMS ON NURSE STAFFING IN HOSPITALS 159

Figure 8.Team Action Sheet.

BEST COPY AVAILABLE

ACTION SHEET FLOOR 03 TEAM A 03/06/72

10 00P 03/06 ilIR I I I I IRESP

SlOREOERBE I I I I I .....1---- I

10 OOP 03/06 IDRESSINGsIippo ICLEAN IIN CNAIR I

GLOBE,HERSE IL ANKLE 113 KIN IL FooT I I ____1011_21___I______ILHISDUILI

10 OOP 03/06 lop IRESD I I I I

JOHNSON,RUS I I I I I I I I

10 OOP 03/06 'CHEST PT 16LOW6LovEI I I I

I JONES,LOCKS I IS MIN I I I --I J I t t 1 1 10 OOP 03/06 (WALK I I I I

KANE,JoHN WITH I I I I I (HELP I I 1 1 1

10 01P 03/06 16P IRESP I I I I

KTNG,CHARLE I I I I I I

I i I I I

10 OOP 03/06 IIPPe I I I I

KINO,CMARLE I I I I I I I I I 1 I

10 OOP 03/06 IAMIULATE IMANrIvENTI I I I

LAWSEN,R000MTH I I I I I .-----IASSISIABC I I

10 OOP 03/06 IIPP6 I I I I I

ROONSIROLAN 12CC NS I I I I 1 Ira5

10 OOP 03/06 19P IRESP I I I I

SIRONG,LAWR I I I I I I I I I A I

1000P 03106 I I I I I I

I I I I I 1 t t 1 1

10 OOP 03/06 I I I I I I

I I I I I 1

I i 11 OOP 03/06 1 *URINE SCAT I

OLOSE,NERBE 1 I I I I I 1 1 1111 IMIO

11 00P 03/06 I 13 CR I I I

I 6LO6E,NERIE 'GREATER I I ! I I I I

I 11 00P03/06 I I I I I

I I I I I 1 ---__----1 I I

166 160 RESEARCH ON NURSE STAFFING IN HOSPITALS ard" formats can be specified at the time the nurse thus ensuring a greater probability of accuracy of writes her orders. The combination of nursing input. This workload determination can be cal- orders and physicians' orders does not include all culated for each nursing unit, nursing shift, nurs- of the tasks done by nurses, but it does cover the ing team, and even hour of the day if necessary. great majority of the patient related variable tasks. Type of Personnel for Each Task Number of Each Task To Be Done This must be a nursing decision and will not This has been the most difficult aspect of de- be done by an information system. However, in termining nursing needs in the pastthe quanti- situations in which the workload to be done ex- fication of the workload. In the system I have de- ceeds the number of personnel available, the com- scribed, we are informing the nurses, on an hourly puter could be used to specify which tasks should basis for each nursing team, which tasks need to be done by which personnel (assuming that many be performed. Itis a simple process for the com- tasks can be done by more than one personnel puter to add up these tasks according to any cate- type) so asto maximize the amount of work gories desired. Thus we have captured in the com- actually done. More will be said about this later. puter, prior to arrival of the nursing staff, a quantitative estimate of a major part of the nursing staff Time Required for Each Task workloadindeed the part related to the most unpredictable variablethespecificneed of each This will vary from situation to situation for a patient. given task but, generally speaking, measurements The estimate is not based on patient classifica- have been made on the variety of nursing tasks tion. It is not based on patient diagnosis. Rather, so as to arrive at reasonable average tildes. Having itis based upon the most appropriate determin- defined the tasks to be-done, quantitated the tasks, ant of nursing staff needs for each patientthe associated each task with a personnel type, and specific configuration of tasks needed for that p- determined an average time to perform a task, tient regardless of diagnosis or broad classification. one then has a statement of nursing staff needs. It does not require any extra nursing time to pro- After doing these calculations for each shift on vide. The basic input is unrelated to nurse staffing. each nursing unit one can then considerstaff It is related to the actual process of patient care, assignments.

Assignmentof Personnel

One approach would he to measure these nurs- way to staff this unit would be to pick some arbi- ing staff needs daily fora long period of time trary level of preparedness; e.g., to be able to carry and determine the range and mean of nursing out all of the tasks 80 percent of the time. This staff needs for each nursing unit for each shift. is more or less the way we staff most hospitals Since we are not yet performing a nurse staffing today. By definition, we are understaffed a certain function, we have not yet writt:n the program to amount ofthe time and overstaffed a certain do this. However, just out of curiosity, we wrote amount of time. The computer system just de- a program that simply adds up the total number scribed could be used to staff in this manner, the of tasks done on our experimental nursing unit only difference would be that we would know what each day. I ran this program on a sample of rather those proportions of under and overstaffing were. randomly selected days and the summary of the out- I am not sure knowing this has much value.. put is shown in table 4. Suppose, however, that we have two or more For the sake of discussion, letus assume that similar nursing units. Let us further suppose that the number of staff needed on each of these days the day-to-day variations in nursing staff needs are is directly proportional to the number of tasks to not completely synchronous between these two or be performedan obvious over simplification, One more units. That is, although the general load of

167 IMPACT OF COMPUTERIZED INFORMATION SVsTFNIS ON NURSE STAFFING IN HOSPITALS 161

Table 1.Nursing activity study BEST COPYAVAILABLE Nursing personnel category Direct care Indirect cate Other Percent Percent Percent All 31 38 33 Nurse ... 26 44 30 Technician 45 23 32 Aide_..___ .._. 32 33 35

Table 2. Summary of repeat studies on ,Aperimental floor Traditional Experimental Observation system ...R1 Transcription error rate ____ 3.4% 1.7% Communication errors ______I9/week 3/week Failure rate in carrying out orders 14.7% 5.8% Direct care activities (all personnel) _....- 31% 38% Indirect care activities (all personnel) __. _ 36% 28%

Table 3.Sample nursing order: "NG Tube Care" Times of Order Message on action sheet message NG tube care Check patency NG tube Q4H Mouth and nasal care Q8H Change adhesiv tape QD Empty NG tube drainage Q1214 Notify house officer if bloody .rainage QI2H Record NG drainage QI2H volume

Table 4.Total number of nursing tasks on one nursing unit on selected days

Date Total tasks 7/30/71 608 10/29/71 1,109 10/30/71 1,I88 11/11/71 957 11/12/71 748 11/15/71 652 11/21/71 527 12/9/71 516 12/10/71 797 12/11/71 611 12/15/71 804 12/19/71 627 12/30/71 273 3/18/72 847 4/11/72 397 Range 273-1.188 Mean _ 677 162 RESEARCH ON NURSE STAFFING IN HOSPITALS patients will vary roughly in parallel between these the units.Ifitso happens that we have avail- units, day-to-day variations in the particular kinds able at least enough of each type personnel our of patients will lead to unequal workloads on these problem has ended. But unfortunately, the exact units. Having the ability to measure these workloads mix of personnel needed for allthe units will on a real-time basis as our system may be able not correspond exactly to what is available every to do, a flexible or dynamic staffing system based day. Thus we can utilize the information con- on this ability to monitor workload ispossible. tainedinthecoefficients C for eachtask(i) Thus, although the total number of nursing staff for each personnel type to modify personnel as- personnelisfixed,their distribution among the signments. Thus if task (i) can be done by any nursing units will vary from day to day. one of the three personnel types(j),each one In addition to providing a real-time measure having a different value of the coefficient Cii, of nursing staff needs, can the information system then I n CijniktiXijk is maximum if the person- be used to help in distributing the staff? With nel(j)performing task(i)isalways the most the help of Dr. Judith Liebtr:.n, Assistant Professor desired(i.e., with the highest value of CIO. of Public Health Administration, we developed It is possible that on a given day the values of thefollowingmodel. Supposethat we prefer nu, will be so high that there is no solution to a nursing aide to perform task (i).If an aide is the problem assuming we wish to carry out every not available,a licensed practical nurse (LPN) task. We are then forced to assume that something could also perform this task as a second choice and less than 100 percent of all tasks will be performed furthermore, a registered nurse (RN) could do it and solvethelinear programing equations for as a third choice. this arbitrarily reduced value of nil,. We can decre- Let us then define a coefficient C, such that ment all n,,, equally by the coefficient F and deter- C-3 if an aide performs task (i), C.2 if an LPN mine the highest value of F that will allow the performs it, and C. 1 if an RN performs it. Let us equation to be solved. Or we can assign priorities make these judgments for each task. Associated with to the tasks and in the event that the total work- each task (i) will be a tinting constant t, related to load on all the units together exceeds the total how long it takes to perform task (i). If we know available staff, no matter how they are distributed, how many times (n,h) task (i) must he done on a certain tasks will be eliminated first. given shift on a given nursing unit (k), then for any This in actuality is what happens every day in task (i), nikt, tells us how long it will take to do all our hospitals in an unplanned manner. In either of task (i) on nursing unit (k). Let Xuk be the case, we will always be able to find some propor- proportion of time that personnel type (j) performs tion of the total workload that can be done by task (i) on nursing unit (k). If we always assign the the available staff and we will be able to optimize most preferred personnel type (j) to perfrom task the distribution of that staff to perform that part (i), then for that (ij) combination Xiihm 1. For any of the total workload. This optimization can then other (ij) combination for the same task (i) Xiik100. he suggested to the nursing administrator prior to arrival of the staff for that shift. Then 7E nikt,Xijkisa measure ofthetotal Many of the ideas expressed here are not original. hours of each personnel type (j) needed to ideally However, the possibility that we can actually carry staff nursing unit (k). out some of these ideas because of information Summing for' all nursing units gives the total systems under development places them in a new amount of each personnel time needed forall perspective.

169 IMPACT OF COMPUTERIZED INFORMATION SYSTEMS ON NURSE STAFFING IN HOSPITALS

Summary

The impact of computerized information systems down to those that truly inns performed by on nurse staffing in hospitals occurs in the areas nursing personnel, the physicians arstillfree to described below. order as many tasks to be performed as they feel They will permit an accurate assessment of necessary. In the past, little attention has been paid nursing staff needs on inpatient nursing units. to what orders doctors write. Recently, however, They will be able to monitor these needs on people are interested in the fact that in seemingly a realtime basis so they can he used in time comparable situations, different physicians order to make judgments about staff assignments on different services. This can be reflected in the nurs- a shift-to-shift basis. ing tasks they order. When people began to audit They will provide the nursing administrators laboratory test ordering practices of physicians, it with a statement as to the ideal number of appeared that at least in some cases there was over eachpersonneltypeneededtostaffeach utilization of the laboratories. nursing unit for each shift. The nursing ad- With an information system such as the one ministrator can use this information as she described here,itis now possible to quantitate physicianutilizationof nursing pleases.' and audit the services. Perhaps over utilization can be detected They will suggest to the nursing administrator an optimal distribution of her available staff here aswell, and appropriate feedback to the to meet the needs. Furthermore, they can indi- physicians may affect these practices. cate which tasks the optimization assumes will To give an overall answer to the question. not be done and who it assumes will be doing What is the impact of computerized information systems on nursestaffinginhospitals?, 1 say thoseJasks that will be clone. this: What other impact might these systems have? They will help determine which tasks will be Let us return to table 2 in which we showed a performed for patients in various situations. reduction in nursing staff indirect care activities. In addition to helping assign the staff to carry out They will help perform sonic of the tasks, re- the tasks, an information system will reduce the lieving nurses of nonnursing functions. types of tasks to be done by the nursing staff in They will permit measurement of nursing staff the first place. needs on a timely basis and provide recom- Finally, there is one other important factor re. mendations on the appropriate distribution of lating to nurse staffing. Once we narrow the tasks available nursing staff to meet these needs. DISCUSSION OF DR. SIMBORG'S PAPER

Discussion Leader Miss Marlene Slawson Research Associate, Operations Research Unit Division of Community Psychiatry School of Medicine State University of New York Buffalo, New York

Miss Slawson

Dr. Simborg's paper has presented some thought he development of quality performance by proc- provoking ideas, and I should like at the outset essing orders and by increasing performance of to state my own position:I am not opposed to nursing care. What does the addition of a com- the philosophy of the use of computers in cer- puterized information system add that other man- tain hospital settings, but I feel we should move ual systems cannot? Does a computerized informa- with caution in certain areas. tionsystemincrease quality? Doesitdecrease I will comment on several points of the paper. cost? Does it increase efficiency? Does itincrease Dr. Simborg indicated that 58 percent of the fail- staff satisfaction?I did not find presented any ure of tasks noted in the studyresulted from kind of evidence that comparative studies have simple nonperformance. Thefirstquestionis, been done to indicate that an automated data Were these all really nursing tasks or were they processing systemas proposed by Dr. Simborg nonnursing tasks with the failure on the part of canpresent quality data to answer the above

.the laboratorypersonnel ratherthannursing? questions Second, Was the failure of the task related to re- Ifeel the system proposed here today has cer- sources such as the lack of equipment? Third, Was tain limitations. While it can simplify the presen- the nonperformance of the task relatedto the tation of a complex idea, it can also oversimplify question of the nurse, whether or notitreally it. And one of the constant problems faced by was a valid task to be done? He indicated that many researchers in this area is that in the study one of the reasons the nursing staff does not exe- of human behavior there is the necessity of trans- cute orders is that they are too busy, hence the lating complex relationships, concepts, and ideas need to staff appropriately. Maybe an alternative into some type of a simpler representation of should be to staff more efficiently with what staff form. is available. I agree very much that itis necessary to con- Dr. Simborg indicated also that the computer- tain a conceptual description of the system. This ized information system will help perform some includes taking the institutional needs and patients' of the tasks, relieving nurses of nonnursing func- needs, coupling them with the resources, both of tions. My question here is, Can another kind of staff and equipment, and then evaluating them system accomplish the same result for less overall in terms of quality execution of patient care, pa- total cost? His basic premiseisthat automated tient satisfaction and personnel satisfaction. The data processing techniques can be used to enhance system described was based upon physicians' or- DISCUSSION OF DR. SIMBORG'S PAPER 165

drib as presented by nurses themselves in terms hospital, but it may serve as a useful tool in ad- of their own description. Alternative input that ministrative and clinical care. could enter into the model are those tasks per- From 1967 to 1969 I was involved in a cost formed by the institution in terms of protocol and analysis and quality of care study, and at that policy. time I was not aware of any successful attempts, For example, patients receive baths thatare through the uses of different changes in staffing not specifically ordered by thephysician. There patterns, to reduce the cost of patient care.Maybe are tasks performed as a matter ofhospital policy, in the total overall operational costs of the hospi- and tasks related to psychosocial enhancement of tal this could be accomplished, but I have yet to the patient; these should also be included in the see a hospital that reports to its patients,"We are overall development of a model. Tasks in the going to charge you $50 less a day because you are latter group include such processes as alleviating receiving less nursing care than other patients or patient anxiety and conducting preop psycho- because you are on a restricted diet." logical preparations. These components do not At that time I proposed that we try to look for really come into the realm of specific tasks but a more equitable systemfor charging patients are important in allocating staffand resources. based on the amount of nursing care the patient It is well perhaps to have some type ofallocation received. A more equitable system could be ar- of time in the model for crisis intervention, un- rived at with the use of an equation that includes foreseen events that occur on the floor, new ad- the following components: missions, and or emergencies. (a) fixed amount for hotel costs; We should move with caution. When someof (b) cost of individual nursircare calculated on the variables for success have been met, a system tasks to be done for the patient, type of per- that will summarize the existing knowledge will sonnel required to carry out the task, salary provide an explanationfor observed events.It per hour, and equipment used; willalso predict the occurrence of future rela- (c)other items shared by all patient charges such tionships that will he most helpful to the nursing as services of ward clerks. profession. This will essentially integrate the phys- There is much to be done in this area. Whether ical system, human organization, and the informa- or not automated data processing systems arethe tion system into an efficient and consistent overall vehicle for accomplishing better patient care re- system. I do not feel that provision of an auto- mains to be investigated using criteria established mated system is the answer to every problem in a for overall success.

172 ConceptModeling inHospital Staffing

Dr. Lillian M.Pierce Professor, School of Nursing The Ohio State UniversityHospital Columbus, Ohio

structured in the form of amech- In a recent symposium,Morris described the cepts which are anism. A model iscumbersome and complex in process of modeldevelopment in the following and parsimony of a words: "The process of modeldevelopment may comparison to the purity .There is a one-to-one cor- of enrichment or form,' system.. . be usefully viewed as a process relationships developed simple models, respondence between the elaboration. One begins with very model and those of the quite distinct from reality, and attempts tomove in in the mechanism of the evolutionary fashion toward moreelaborate models formal system.... "The terms and relations ofthe formal system whichmorenearlyreflectthecomplexity were derivedfrom the concepts andmechanism of the .. ,situation. it is of some of the model, respectively,and neither the con- "This seems harmless enough, yet mechanism are part of thatformal ..The at- cepts nor the importance to point it out explicitly. . of the model is to generate tempt to beginimmediately with a rather rich system. The purpose .The formal system islimited model may become a serious sourceof frustration. relationships... things moving and thus to statements ofrelationships in the simplest pos- Startingsimplygets make them easilytestable tends to relieve some of thetension. It does, how- sible form in order to ever, require a certain amountof poise or 'guts' (4)." Three kinds of models aregenerally described from a complicated problemand hcgin to back off in the literature: analogue,iconic or typological, with a simple conceptual structure.It requires one of and symbolic. Churchman says"an analogue model to deliberatelyomit and distort certain aspects represent some knowingly commit the sins employs one set of properties to the situation and to which the system being considerations and subop- other set of properties of suppressing' difficult studied possesses (3)." Willer says"analogue mod- timizing (1)." els are constructed byallowing some set of proper. Models are abstract represent".tionspf a set of stand for successful model not ties,structure and (or)process A to empirical phenomena. The and (or) process of the only "must be isomorphic withits domain of ap- the properties, structure, phenomena being studied, XThis is most com- plication (2)" but also muchcorrespond to a for- monly done when the Aproperties are better mal system or validatedtheory for that set of known and more familiarthan X (4)." The major phenomena (3). Willer saysof models, "A model advantage of the analoguemodel is the known contains a rationale andnominally defined con-

167 168 RESEARCH ON NURSE STAFFING IN HOSPITALS

reference or starting point. However, the inher- The construction of a symbolic model begins . ent weakness of "borrowing" rather than construct- with the nominal definition of concepts and the ing the rationale and the mechanismmay render development of rationally consistent assumptions. the model ineffiective, particularly if the analogues It is not unusual for concepts to be theoretically are not isontorphic with the phenomena being as well as nominally defined. The structural and explored. functional relationships of the concepts may be Iconic or typological models are generally scalar derived from their theoretical definition. Explicit models of direct similarityto the subject they assumptions, in addition to the implicit assump- represent. The mechanism of the iconic model tions of theoretical definitions,may be introduced. depends upon the number of similarities between The meaning of the concepts and the relations the model and reality. As the model increases in between them provide the rationale for thesym- abstraction the similarities decrease, and the model bolic model. If the construction of the model be- is rendered useless. Iconic modelsare strongest at gins with the definition of concepts and the de- lower levels of abstraction. If properly constructed velopment of assumptions, the rationale will evolve. the iconic model is isomorphic with the phenom- On the other hand, it is not unusual to begin with enon. If the transformation from reality to model an explicit statement of the rationale that then is only scalar then isomorphism is assured. becomes the basic assumption for the model. As When a set of empirical phenomena isrepre- eachconceptisoperationally defined andits sented directly by a series of nominally definedcon- meaning made clear, it leads to further definition cepts,therationaleforsuchrepresentationis and clarification of related concepts. iconic. However, unless the concepts thatrepresent Although there are no formal rules for model the phenomena are structured ina mechanism of development, itis wise to keep in mind Morris' some kind, the end result can be nothing more admonition to start simply and "make haste slowly" than aconceptual scheme withlittle,ifany, toward the more elaborate model that reflects the capacity for prediction. The effectiveness crsuc- complexity of the situation. The primary requi- cess of iconic models depends upon abstracting a site in all instances is extensive knowledge of the mechanism from the phenomena and then using phenomena for which the model is being con- the mechanism to order or connect theconcepts. structed. The symbolic model is more formal in itscon- There are, of course, certain phenomena suffi- struction than either the analogue or iconic model. ciently circumscribed that expert knowledge from Additionally, the symbolic model has a strong but a single discipline is adequate. However, the phe- not rigid mechanism. As Willer says, "Symbolic models are constructed by the meaningful inter- nomena with which we are dealing is of such connection of concepts. Models of this sortare complexity that, to date at least, attempts to con- symbolic in that (1) their general rationalecon- struct models have been most successful when the sists of allowing a set of connected concepts to knowledge from several different disciplines has symbolizeasetof phenomena, and(2)their been brought to bear on the problem. There- symbols or concepts are the source of their mech- mainder of this paper discusses one particular anism (1)." model.

Development of a Patient Care Model

Introduction part of the program dealing with health systems has been directed toward the development ofa The project discussed here ispart of a long measure of patient care and a model of health range research program of the Systems Research care systems relating the measure to health sys- Group at Ohio State University. The primary ob- tem behavior. jective has been the development of methods for When one speaks of a measure of patient care, investigating complex man-machine systems. That the listener generally tends to think of a scale,a

174 169 CONCEPT MODELING IN HOSPITAL. STAFFING things checklist,or some other suchinstrument. The netics a theory of machines that treats not measure of patient caredeveloped (luringthis but ways of behaving. It does not ask, Whatis this project is not a "tool kit." It is a descriptive model thing? but, What does it do? It .is thusessentially at the tactical, or bedside,level. Itis descriptive functional and behavioristic. (11). Cybernetic con- of the interactions of the individual patientwith cepts are familiar to a number ofdisciplines such various members of the health team and reflects as medicine, physiology,psychology, and sociology. patient response to system .behavior. The cybernetic model isprimarily concerned The early phases of the work were characterized with communication and information flow in com- by multidisciplinary organization. Representatives plex systems. The model feedback, control, and of the various disciplines investigated those aspects regulation has applicability for biological and so- of the problem that were traditional withthem. cial systems as well as the engineeringproblems For example, the sociologists looked at suchthings to which it was first applied. Every systemis sub- adopt to as status relationshipsand work performance, the ject to forces that make it necessary to psychologists examined personality variables and change in the environment, internal or external. patient progress, and the operations researchers The homeostasis, dynamic equilibrium, orsteady studied problems of inventory and resourcealloca- state model is essentially a processof balancing change tion. Fruitful as these activities were, theystill did the components of a system in response to not attack the majorproblem: how wellis the in the environment. The system canadapt through hospital doing its job? The degree to whichthe feedback information to a wide variety of situa- concepts and variables were common tions (13). research developed is to the various disciplines waslimited. The patient care model we have a descriptive model atthe tacticallevel.Itis descriptive of the interactions of theindividual Conceptual Framework for the Model patient with all members of the health team.The patient is viewed as an information source com- The later phases of this study have been inter- municating with members of the health team disciplinary. Unifying concepts were selected, and through a variety of sensory input channels.In- the research has been conducted in aconceptual formation about the state of the patient isob- framework common to representatives of all the tained by the nurse or physician or otherhealth disciplines involved. Cybernetic concepts provided team member who observespatient behavior. the framework for the development of the pa- Comparisons are made between the actual state tient-caremodel (5,6,7,8).These conceptshave of the patient and the desired state.Action is taken been criticized as being mechanistic;i.e.,ruling to reduce any difference that mayexist between out the richness of variable humanbehavior. They them. The rationale for the model isthat the are, in fact, based on thebehavior of living or- patient care system is adaptive,relying ou ;,,for- ganisms (7).Beginning withBernard'sdisturb- mation feedback to cope withdisturbances from ances and responses of theinternal and external theinternal or external environment(ii), and milieu (9) and with Cannon's ideas ofhomeo- that the purpose of the system is thereestablish- stasis (10),the concepts have been revised and ment and raaintename ofhomeostasis (or dynamic extended by advances in informationtheory, com- equilibrium) for the patient (15,16,17). munications, and computer technology (11). The patient care model was developedby first The basic concepts of cybernetics areinforma- conceptualizing an integrating modelbased on tion and communication, regulation and con- cybernetic concepts (17) to serve as atheoretical trol.VonBertalanfrydescribescyberneticsas point of reference for the collection ofdata about "theory of control systems based on communica- the phenomena that the model represents, pation (transfer of information) between systemand tient-healthteam interaction. Thesedata wire environment and within the system and control used to refine and extend the model to its present (feedback) of the systems function in regard to be environment (/2)." Weiner, who coined the term, form (figure1). The system or subsystem can defines cybernetics as control and communication described as aseries of functional components in the animal and machine (6). Ashby callscyber- with the following behaviors. 170 RESEARCH ON NURSE STAFFING IN HOSPITALS

Figure 1.Patient care model.

XL Memory t Environment YL_ XLt y

Ys1CA(Ys-4A) (X ,,,X V(11,1

Information Control 1=11111w Regulate wmass.01. Source

Monitor

176 CONCEPT MODELING IN HOSPITAL. STAFFING 171

The control function consists of specifying the We considered electronic recording equipment limits within which the patient is to he maintained but found it to be impractical when the patient (symbolized as YR). The limits are based on knowl- was ambulatory. We also considered videorecord- edge about the individual patient as well as medi- ing. This method is excellent for collecting con- cal and nursing knowledge in general. Limits are tinuous data in well controlled settings such as subjecttochangeas thepatient progresses the operating and recovery rooms or the intensive through the course of his hospitalization or as new care unit where cameras can be ceilingmounted information is obtained. and remotely controlled and where the time span The .function of the memory unit isto store for observations is relatively short-4 to 8 hours. knowledge and information. The content of the But the high cost for equipment and technical memory unit is constantly updated. The patient assistance and the low reliability of the equip- represents the signal source transmitting informa- ment when used continuously over long periods, tion about patient state (symbolized as VA) to the as well as the limited range of the camera's scope monitor where tF'c information is observed and when the patient is ambulatory, made it impracti- recorded or displayed. The function of the com- cal for our purposes. parator is to compare the actual state of the pa- We finally decided that nonparticipant observa- tient as observed by the monitor to the desired tion with manual recording of data in the form state specified by the controller, and to note any of a narrative description of events would give us differences between them (symbolized as Y8 VA). what we needed. We decided to use nurses to The ...egulatoryfunction consistsofselecting collect the data, for several reasons. First, observa- anti carrying out the appropriate action [symbol- tion is considered an essential part of the nursing lized as (X,, Xr,)]necessary to reduce the dif- process. Second, the hospital would be a familiar ference between the actual and desired patient environment and nurses would be less concerned states. The patient, as a process, responds t' in- withtheirsurroundings than would observers puts from both the environment and the regula- from another discipline. Finally, nurses would be tor. able to identify various personnel interacting with the patient and the procedures being carried out. We developed a manual of data collection de- Collection of Data for the Model si3ned to orient the observers to the project and indicate what was to be observed and how the To develop a model that accurately describes observations were to be recorded (18). The entire the behavior of the system in terms of the pa- philosophy of the data collection was that of the tient's response to the allocation of hospital re- "now generation,tellitlikeitis. We used a sources, it was necessary to observe the interac- variety of techniques to prepare the observers, in- tion of the patient and various members of the cluding the use of videotape and practice sessions health team in a variety of settings throughout in various hospital settings. We also used video- the hospital. Since the systemis adaptive, itis tape and dual observations to determine interob- not possible to predict a priori when interactions server reliability. or changes will occur. The observers were formed into six member To obtain the comprehensive data needed to teams, with each member of the team responsible develop the model, patients were observed con- for the same 4-hour block of time each day dur- tinuously (24 hours a day)fromadmission ing the patient's hospital stay. Observations were through discharge. We explored several methods begun when thepatient was admittedto the for collecting the data. During the earlier work hospital and continued until discharge. Wherever on this phase of the project, we collected data in the patient went the observer also went and re- the operation room by visual observation and man- corded, as objectively as possible, a narrative de- ual recording of events supplemented with tape for, or with the recording. The time span for those observations scription of "who did what to, was relativelyshort,the number of variables patient and how the patient responded." Ten was relativelyfew, and the situation was con- patients were observed. The length of stay ranged trolled by the anesthesiologist and the surgeon. from 9 to 24 days. With observations on a 24-hour RESEARCH ON NURSE STAFFING IN HOSPITALS

basis, approximately 3,000 hoursoftime series Because tabular dataare difficult to interpret, data describing system behavior were collected. a graphic time seriesplot was developed. The Five complete sets of data are being used. plotis designed to demonstrate the concomitant The data, recorded as narrative description, had variation of patient andresource variables. The to he reduced to a format suitable tor «mpter original plot of 21 variables hasbeen expanded to analysis. Again using the conceptualIramework to allow up to 120 variablesto be plotted simultane- provide the guidelines, we developed acoiling ously. The plotscan be examined to determine formatthat allowed numberson punclicards to the state of the system when be translated into situations that re- sentences. The sentences can quire nursing intervention arise(20). be read as "at this time, in this location, this jw- The third. basic problem isa search program. son is doing this to, for or with the italic:al, and Having specified, or located,a critical patient vec- this is the patient's response" or "at this time, in tor, such as pain, the dataare searched prior to this location, this is the patient's state and this is our following the occurrence of sucha vector to being done, by this person." determine patterns ofresource allocation that follow or precedeit.Analysis of the data in this Analysis of the Data fashion makes itpossible to determine to what extent resources are allocatedon basis of pa- The analyses of the dataare described in the tientinformationandtowill;extentother final reportforthis project. The description of considerations are more important (2!). the programs for analysisconies directly from dolt The three previousprograms deal with the in- report (19). dividual patient. They describethe interaction of With the capability ofcomputers. the systems the patient with thesystem environment in time. researcher is able to deal with thecomplexity of To generalize from theindividual patient,itis Multivariate systems. However,itis necessary to necessary to sum the results across patients. This specify the final output andprepare the necessary isa function of what Ashby calls the Rx D programs to generate the specified output. So far table (11)that describes theconsequences of a wehave developed four basicprograms for data system response to a disturbance in the environ- analysis. ment (see figure 1). If the model is descriptive ofsystem behavior, The R'srepresentresource allocationinre- one of the first questions to he answered is, who sponse to the D's, disturbances suchas pain. For does what for patients, when, with whatresources, example, ifpainisa disturbance (D) and the and how does the patient respond? Aprogram administrationof Demerol aregulatoryaction was developed to cross-tabulate the following: (R). the resulting state of the patientwould be (a) resourc_ and patient variables with time, found in the cell at the Rrow, Dcolumn inter- (h) patient with resource variables, section. As more data are obtained, the cellscan he filled in. If there (c) resource with resource variables, are statistical regtdatives the probabilities of various patient (d) patient with patient variables. responscs, to regu- latory-disturbance can be determined. The yield from this program has beena series The patient care model hasnot been developed of frequency histograms that demonstratepatient- to that stage earlier identified as the ultimate resource interactions such as the frequency with goalofthemodel builderthe mathematical which various levels of personnel carry out various model. It has, however, been developedto a stage activities during each day of hospital stay. Scrutiny that allows its to examine the patient'sresponse of these functions is expected to suggest hypothe- to system behavior, to determine who is doing ses that can be tested, either inthe real world what for patients, when, with whatresources and or with data generated by the model. with what responses. 173 CONCEPT MODELING IN HOSPITAL. STAFFING References

(1) Mc..ris, William T. "On The Art of Model- (13) Pierce,Lillian."General Systems Theory: ing," In Ralph Stogfill(ed.), The Process An Overview."Paperpresentedat2nd of Model Building in the. Behavioral Sci- Nursing Theory Conference, University of ences, Columbus, Ohio. Kansas, Department of Nursing Education, (2) Black, Max. Models and Metaphors:Stud- 1969. (ed.) Self iesin Language and Philosophy.Ithaca, (14) YOvits, M. C., and Cameron, S. New York: Cornell University Press, 1962, OrganizatingSystems. New York;Perg- p. 238. amon Press, 1960. (3) Churchman. C. West;Ackoff,RussellL.; (15) Howland, D. "Approaches tothe Systems and Arnoff, E. Leonard. Introduction To Problem." Nursing Research, 12: 172-176, OperationsResearch.New York:John Fall 1963. NViley and Sons, Inc., 1957, p. 157. (16) Howland, D. "A Hospital SystemsModel." (4) Willer,David.Scientific Sociology.Engle- Nursing Research, 12: 232-236, Fall 1963. wood Cliffs, New Jersey: Prentice-Hall, Inc., (17) Howland, D., and McDowell, Wanda. "The 1967, pp. 18-19. Me. surment of Patient Care: A Conceptual (5) Beer, S. Cybernetics and Management.New. Framework."Nursii Research,L3:4-7, York: John Wiley and Sons,Inc.,1960. Winter 1964. chapter 1. (18) Pierce, Lillian;;.: r, Nancy; and Lara- (6) Weiner, N. Cybernetics. New York: John bee, Jean. Data ction Manual. Adap- Wiley and Sons, Inc., 1948. tive Systems Re ..arch Group, Ohio State University, Columbus, Ohio. (7)Ashby, W. R. An Introduction to Cyber- netics. New York: John Wiley and Sons, (19) Howland, D.; Pierce, Lillian; and Gardner, Inc., 1956. M. The Nurse-Monitor in a Patient-Care System. Report No. RF1651, Adaptive Sys- (8)Guilband,' W. T. What is Cybernetics? Lon- don: Heineman, 1959. tems Research Group, Ohio StateUniver- sity, Columbus, Ohio, (in preparation.) (9)Bernard, C. An .!ntroduction to the Study ofExperimentalMedicine.New York: (20) LaRue.R."Mediplot."RF1651Wp64, Dover, 1957. (on file) Adaptive Systems Research Group, (10) Cannon, W. B. The tf. isdom of theBody. Ohio StateUniversity, Columbus, Ohio, New York: W. N. Norton ik Company, 1932. 1969. (11) Ashby, W. R. Designfor aBrain. New (21) Reid, R, "Computer Program forBack- 'York: John Wiley and Sons, Inc., 1960. ward/Forward Search of Data." RF1651 WP. (12) Von Bertalanffy, L. General SystemTheory. 62,(onfile)Adaptive Systems Research New York: George Braziller, Inc., 1968. Group, Columbus, Ohio, \1969. DISCUSSION OF DR. PIERCE'SPAPER

Discussion Leader Dr, Wanda E. McDowell Associate Dean For Research School of Nursing University of Michigan Ann Arbor, Michigan

Dr. McDowell

Throughout thisconference,concernshave variables that must be attendedto and studied. been expressed regarding the lackof model de- The modelisespecially helpful in selecting velopment, the need fora theoretical framework critical variables fornurse action or for nurse to direct inquiry,, the disappointingprogress in intervention. I thinkwe have to consider what the development ofmeasures of care, and the the nurse attendsto, what the physician attends failure to make the utility of nursing explicit. to, and certainly what the patient attendsto be- Implicit in all theseare further concerns regard- cause he, himself, is a regulator of his condition. ing decision making, patientoutcomes, and allo- We have to complete this cycle cation of resources. and consider consequences for the patient.I Dr. Pierce's paper chose the recovery represents an approach to all room as a source of data formy doctoral dis- these concerns. The key elementsin the adaptive sertation. In that setting, thereis a circumscribed system model are communication and control,As period of time in which thepatient itis is captive, presently composed,itisdescriptive, not so to speak, and the classes of variables that predictive. We do need one to know the outcomes attends to are fairly limited. But ifyou think of and consequences of variousregulatory strategies the physiological and psychological in the management of variables of patient condition, andwe concern there, then as the patientreturns to the do need to be ableto predict these outcomes. unit following recovery (from anesthesia)other Otherwise the approach (the model) is not going variables come into focus. Weare concerned with to be utilized by practitioners, by administrators, social variables and economic variables,for ex- or others in the health care system. It should be ample. useful to researchers, also. Thesevariables mange overtime,too,as There has been further criticism that themodel do the goals of the subsystem. has been used only in hospital settings.But the One thing that attempt originallyand I think it still pertains should be added to figAre 1 ofDr. Pierce's paper is that time is a critical variable that has been to move out of the hospitalsetting thus illus- into the community and thehome, wherever there trates the dynamic component of the model. An- is a consumer ofcare. In a series of three papers other very crucial element isthat the focusis on nursing research and the development ofthe on the patient. Of primary concern, too,is the system model produced in 1964, therewas con- interdisciplinaryeffortrequiredtotacklethis cern for the various settings in which thecon- question, this problem. There hasbeen criticism sumer of care finds himself and the classes of of the model anda lot of argument about it, and

174 140 175 DISCUSSION OF DR. PIERCE'S PAPER models in the real world we someone made the commentearlier, Why isit we test available kind that each one has to develop his ownmodel? Why will delay inquiry and interaction and any handle can't we use a model, redesign it,refineit, and of professional, scholarly attempt to get a be critical in a constructive kind of way?Unless on the concerns we have beendiscussing. Summary, Synthesis, and Future Direction

Dr. Mary K. Mullane Professor, College of Nursing University of Illinois Chicago, Illinois

Our conference started with a discussion of thatitcan measure only whatis being done staffing studies. based on an engineering model overtly. It cannot measure what nurses have not of nursing performance. In the studies reported, defined. It cannot Net ideals or standards of ex- nursing activities were measured along production cellence. Measurement means some type of defini- industry modes: standards, standard times, work tion and quantification of what can be physically sampling, etc. These' are very useful to the degree discerned. that nursing isa sequence of tasks to 'beper- Another criticism of the engineering model de- formed within given limits of time. rives from the construction of discrete,specific Criticismoflimitationsoftheengineering time per activity from work and time studies; i.e., model was evident in the discussions following a mean, a point estimatorwithout indication of the formal presentations. Some suggested the en- levels of confidence. Without confidence intervals gineering model isfaultyinthatthetasks ap- being-made explicit, nurse staffing based solely on proach does not reflect nursing responsibility for an engineering model will be suspect notonly by emotionalsupport,forpatient instruction,for nurses but by administrators andeconomists pre- coping with nonmedical and nonhospital patient cisely because the flexibility of confidence intervals conceths, and for family management. is not built into staffing estimates. If the notion Illustrationsofnursing performanceinthe of confidence interval in nurse staffing estimates "psychosocial" facets of patient care and support dues not apply, can engineers and statisticians make were frequently cited in papers and discussion at us understand why it does not? this conference. Nurses in hospitals have not yet Patient classification schemes/were presented and explicitly defined these facets of rare, and I think ctiticized during the confere4e, but time is per- we nurses are fully cognizant that suchdefinition haps more similarity in patient schemes than we must be ours. It cannot be delegated. Definition, arc prepared to admit. Questions of their validity description, and then measurement cannot be and reliability have been raisedfrequently, yet hoped for except through our own efforts. several speakers reported coming to comparable One criticism of the engineering modalityis conclusions.Irefer you to the Canadian series

177 178 RESEARCH ON NURSE STAFFING IN HOSPITALS that was reported for us by Phyllis Giovannetti. It performance. The bargaining model makescon- seems evident that patient classification schemeas structive advocates out of each, placing them in a base measure of nursing workload is generally opposition to negotiate apportionment ofhospital applicabledespitetheconcernsexpressedin- resources to achieve both hospital efficiency and ability or unwillingness or impropriety of general- professional effectiveness; i.e., quality patientcare izing among hospitals that oftenseem comparable at a tenable cost. from the outside. Gerrit Wolf's peer reviewreport reinforces the Hospitals seem comparable from theusually re need to attend to administrative relationships and ported data size, whether or not theyare volun- the bargaining skill of all partiesto it. Negotiation tary, etc. But they vary widely internally.They goes on constantly between nursing administration vary as to style and capability of their general and other departments, and perhaps theprocess administration. They vary about medicaltherapeu- and content of such negotiations would shedlight tic capabilities and ambitions, theadoption of fash- on nursing load and its changes, nursing utilization, ions, whether theyare cobalt bombs, organ im- and so on. plants, or transplants. They vary as to internal The bargaining conceptleadsinescapablyto organization. They varyas to allocation of re advocacy. Improving patientcare throughthis sponsibility and of costs to nursing. Eachof these mode requires that both administratorsand profes- variances has a major effectupon nursing care and sionals accept negotiationas good, healthy, and nursing administration. Dr. Gerrit Wolf'sreport of useful. Attitudes will needto be changed and his peer review study pointed thisout. sophistication in the modes andprocesses of bar- The fact that the engineering modelalone is gaining must be developed. Successful bargaining presently incomplete does notsuggest to me it rests upon negotiation from strength, mature un- should be replaced with another model.Rather, derstanding of power, and the development and progress reported in the literatute suggests itscon- constructive use of it. tinuing usefulness;fengineers, nurses, hospital One wonders what would be revealed in the administrators, and trusteesare sophisticated about study of the amount of nursingpower, its sources, what it can and what it cannot do andare pre- present and potential, and its tun, present and pared to apply it with full understanding ofthe potential. Can patientcare in hospitals really be proper concerns and reservations of each of the either efficacious or efficient withnursing unable parties. or unwilling to become skilled advocates for it? Human services have all of the problems, human Application of the bargaining model raisesmany problems, that production industry has. Butthese useful questions, power being onlyone of the more are compounded by a product that in itselfis obvious. human and therefore must be described andmeas- Dr. Flagle and Mrs. Giovannetti asked in dif- ured in human terms. Industrial design is much ferent ways whether it is not now time to approach easier to deal with than nursingcare design. Nurs- nursing staffing and its host of related issues through ing supply and performance and procuctionare systems analysis. Mrs. Giovannetti's paper contains characterized both by human inputs and human a scheme for looking at the entire health care outcomes,Describing and measur;ng themin system, nursing included. Dr. Flagle suggests fun- human terms (and engineering modescan be damental questions about the objectives of the humanized) is the only real way, however difficult health care system. Specification of objectives of this may be. the health care system and the measurement of John Griffith introduced the bargaining model. their achievement are difficult. Perhaps thatap- He did not suggest substituting the bargaining proach for some researchers could help ration- model for the engineering model; theycan com- alize the nursing subsystem in hospitals. plement one another. A bargaining model would The Pierce report adds both weight and portent make explicit the administrative relationships be- to that approach. Often in the discussion, factors tween trustees and administrators concernedprop- in the climate in which nursing is practiced erly with costs, and the professionals, nursingin- mentioned, and their effect upon nursing staffing, cluded, concerned with quality of professional nursing utilization, and nursing effectivenesswere

183 SUMMARY, SYNTHESIS, AND FUTURE DIRECTION 179 alluded to, but their identification and systematic Our habit is to describe the ideal. Although the in depth investigation lie ahead. idealiscertainly important, experience suggests Dr. .14,:de Hone spoke of the effect of the variety one can work toward it only by delineating and of medical styles upon nursing staffing.Experi- pursuing first what is feasible and rational. One CObSerVerS of the hospital scene know -one of our speakers introduced the notion of -moderate cannot completely study nursing or get definitive utility,"the conceptI am advancing here.Isit answers except within the framework of the given possibletoidentifycriticalindicesofquality hospital's medical style, medical capability, medi- rather than concluding that we have no quality at cal ambition if you will, medical modes of practice. all unless we measure everything imaginable? The ranges of therapeutic: capability present or Another point of future directionliesin the grasped for has a critical effect upon nursing staff- refinement of patient classification schemes. ()nes- ing. Medical milieu simply must be studied to afford tions of validity and reliability remain, but the a complete understanding ofthe problem. The generalapplicabilityofpatientclassification same istrue for thehospital's style of general schemes suggest their usefulness. As Mrs. Giovan- administration. Research into these relationships is netti pointed out, the question is not whether they urgently needed. are good or had; the question is how do we refine The most frequent question raised in discussion them so asto make them more widely useful. periods at this conference has been that of quality What other measures of patient load are there? of nursing care and its measurement. Though be Surely the future will bring greater study of ginning attempts at definition and measurement organizationaland operationalmanagementof are under way, the major work lies ahead and its nursing services. Systems analysis and design hold priority after this conference would certainly seem great promise, in my view. So too does the concept to be very high. Perhaps ways can be found to of management by objectives. Horizontal and veni- document and synthesize what work has been clone cal relationships and controls exercised by nursing and subsequently to convene persons knowledge- and upon nursing need to be studied. Long over- able in investigation of the question and in ap- due are studies of the effect of hospital adminis- plication of modes of documenting qualityin tration and of the effect of medical practice upon nursing care given. the' organization, objectives, and character of nurs- My assignment for this summary included the ing load and nursing care outcomes. charge to make explicit some of the "straws in We gave patient satisfaction no attention in this. the wind" of the future. Clearly, greater atten- conference. It may be that the, fashion of studying tion to the economics of healthcareis made patient satisfaction has passed. It may be that we mandatory by the growing resistance to continuing know all we need to know aboutii.I am not rise in costs, John Griffith referred to the inclusion familiar with what we know about it, butI was of health care costs under the Federal Price Com- struck by the absence of discussion of patient satis- mission; references to the cost benefit ratio are faction. Indeed, I drew an inference from discus- increasingly common in literature and professional sionthatpatients might be thoughtincapable conversation. of judging the quality of care. 1 do warn against The emphasis of this conference on the urgency this inference, because we are moving more and of identification and measurement of quality in more to the notion of satisfied customers ina health care deliveredisuncles .cored by develop- population becoming more sophisticated. and more ments likethe Commission on the Quality of vocal about health care. Health Care as proposed in the Kennedy bill, and Study is needed of the porposes and processes the refinement of standards and criteria of the of surveillance of nursing performance: of super- Joint Commission on Accreditation of Hospitals. vision, the supervisory process in nursing: of con- Professionals who are making health care the field trols upon or within nursing; of the skill and of their career endeavorsphysicians, nurses, ad- sophistication of nursing administration and the ministrators, engineers, social scientistsmust come consequences of their presence or absence. to grips with an operational, measurable definition Almost 20 years ago I studied nursing adminis- of quality of health care, specifically nursing care. tration in hospitals in Nlichigan.I was startled to .04 180 RESEARCH ON NURSE STAFFING IN HOSPITALS

learn then that the nursing department budget ing, and justifying nursing care and servicere- exceeded the budget of almost 40 percent of the quirements we are developing now be applicable industries of the State of Michigan listed in the outside of hospitals? What has beenor will be Manufacturers'Directory.Fewrecognizethe learned in hospitals that, if appropriately applied, magnitude of the operation nursing administrators will prevent public health nurses from "reinvent- are responsible for or the educational and personal ing the wheel?" qualifications crucial to adequate management of If the already discernible movementaway from present day nursing services. Directors of nursing individual fee for service medical practice toward are often appointed because of tenure in other HealthMaintenanceOrganizationsandother posts in the hospital or acceptability to the medical group or corporate modes or practice actually pro- staff, future will be stormy indeed wherever such practice continues. ceeds, then I fail to see how medical servicescan I recognize that our conference was directedto- escape analysis and insistence upon quality assess- ward staffing in hospitals, and hospitals willcon- ment and staffing justification, as nursing pres- tinue to hold a crucial place in the healthcare ently is experiencing. What a public-gain it would delivery system. Rut the future, inmy view, will be if the modes developed inresponse to appro- see greater development outside of hospitals than priate pressures on the nursing profeisionwere in them. Will the modes for rationalizing,measur- found to hue fransfors.Wlity zo.medicinel

4 Appendix ,

BEST COPY AVAILABLE Task Force Reports

TASK FORCE TOPIC: Future Directions for Research on Nurse Staffing in Hospitals

In this context, theask Forces are asked to 2.funding, and appropriate sources, A.Identi(v areas of needed research: 3.other. 1.identify factors that may facilitate research C.Determine priorities for research. in these areas, D.Explore means of coordinating work of re- 2.identifyspecificobstacleshinderingre- searchers and utilizers: ways to improve com- search in these areas.. munications. B.Determine needed resources: E.Recommend nature of further conference on , 1.skills, and how they may be developed, nurse staffing research.

The Task Forces Task Force No. 1 Task Force No. 3 Miss Elinor D. Stanford, Leader Dr. NOTH poulin, Leader Miss Deane B. Taylor, Reporter Pr. Ber044 AO Pt. Dr. Virginia S. Cleland 11!ss Marie R.ennedy Dr. Charles D. Flag le Pr. 1-,4140 Dr. Richard C. Jelinek Dr. Reginald It evans Dr. E. tartly Jaco Miss Jessie M. Scott Dr. Stanley Jacobs Mr. PlarcP, Jr. Miss Angie Kammeraad Dr. FranA. Sloan Mr. Stanley E. Siegel Dr, 1141-Al A. Wandelt Miss Marjorie Simpson Dr. Harvey Wolfe

Task Force No. 2 Task Force No. 4 Dr. Maria Phaneuf. Leader Mrs. Margaret Sheehan, Leader Miss Marylou NkAthie, Reporter Miss Pamela Poole, Reporter Dr. Alexander Barr Dr. Mario Bognanno Mr. Richard G. Gardiner Mrs. Margaret Ellsworth Mrs. Phyllis Giovannetti Miss Verna Grovert Mr. John Griffith Dr. Doris Bloch Miss Dolores M. LeHoty 'Dr. Elizabeth Hagen Miss Ruby M. Martin Dr. Richard Howland Dr. Donald Siniborg Dr. Eugene Levine Miss Bernice Szukalla Dr. Wanda E. McDowell Dr. Ger-it Wolf Miss Marlene R, Slawson

181 186 182 RESEARCH ON NURSE STAFFING IN HOSPITALS REPORT OF TASK FORCE NO, 1

Leader Reporter Miss Elinor D. Stanford Miss Deane B. Taylor Director, Nursing 1 esearch Field Center Associate Professor Division Jf Nursing State University of New York National lmitutes of Health Buffalo, New York San Francisco, California

Recommendaticy is

A.Future Research Studies team of interdisciplinary consultants to really go in 1. A study to establish a theoretical base of and work with a group in a hospital or university to knowledgrof the .4r5nrAragrelationship." identifytltp problem, prepAre The proposal, the What do we mean by "care" and "caring?" methodology, etc. Perhaps the nurse would need Based on the above applied research to then some financial support to help her get started for a determine, year or so until services become known to nurses a. Where should care be given or provided? who would like to get into research but have not b.. Who should provide various levels of had doctoral preparation.) care? (responsibility and accountability) b.Obstacles to research: c.Careful delineation between the:tarious (1) fragmented roles of auxiliary staff caring roles in the delivery of health care being locked in by union contracts, systems. (2) ethnic demand for career opportuni- 2.Various disciplines need to join force,. in ties in the health service field, designing studies including economists, in- (3)rigidity of present professional and dustrial psychologists, sociologists, statisti- administrative people to studies of cians, industrial engineers, and well quali- their functions and roles, fied nurse researchers. (4) vested interest groups who are threat- 3.Studiestolookatvarious. staffing mix ened by possible findings of studies patterns. The nurse clinician(s) role and fear of loss of jobs, incompetency to will/anion should definitely be in these mix do job, job descriptions not clear so patterns. Their contributions to the quality uncertainty as to role and functions, of patient care need to he identified and the (5)inadequate preparation of those who cost analyzed within inpatient and ambula- will be involved in the study in any tory settings. way results in false data and thwart- 4.Studies of nursing practice in other coun- ing by personnel. tries, more cross cultural studies. What can B.Skills Needed and How These May Be De- we learn from other health care systems? veloped 5. What are the factors that make for delega- 1.Doctoral preparation of adequate numbers tion or assignment of tasks in multipatient of rttrse researchers. Most now arc in ad- settings: i.e., the question of specialization ministration and teaching wit?, no time to and division of labor vs. versatility? 'become really involved in research or even (Personalsuggest ionfromNlissTaylorskilled toassistnurses in other than graduate hilt .eresearchers shouldform an independent programs (master level), which really is not consultant firm, like private practice, and organizea enough preparation.

1S'; APPENDIX: TASK FORCE REPORTS 183

2. Nurses who are interested in research should E. A Working Conference to Develop a Master work with an ongoing research study per- Plan on Strategy for Research in Nursing haps as apprentices. This would have to 1.Participants should do "homework" before include writing a propos.ii, etc. getting together. 3.Interpersonal skills, interviewing and data 2. The conference should follow soon after collecting experience, preparing data for this (present) conference analysis and writing reports. 3.Papers should be prepared and sentto 4. Funding to do needed research was not participants early, and authors should at- discussed by the group. tend but only to discuss salient points raised C. & D. These topics were not discussed by the by participants that would be pertinent to group. the development of the master plan. 184 RESEARCH ON NURSE STAFFING IN HOSPITALS BEST COPYAVAILABLE REPORT OF TASK FORCE NO. 2

Leader Reporter Dr. Maria Phaneuf Miss Marylou McAthie Professor of Nursing Regional Nursing Consultant College of Nursing Nursing Research Field Center Wayne State University Division of Nursing Detroit, Michigan National Institutes of Health San Francisco, California

Recommendations

A. & B. These topics were not discussed by the multidisciplinary and should include, as a group. minimum, ,indostrial engineers, psychologists and nurses. C.Determine Priorities for Research 1.Research should concentrate on clinical I).Explore Means of Coordinating Work of Re- practice with emphasis on the need to searchers and Utilizers develop quality control. 1. Any approach to researchers and utilizers a. patient oriented Develop a study of should be multidisciplinary. structurecuttingacrossprofessional 2. The providers of care must be involved. stratifications, b. Examine the careprocess relatedto 3.Organizations such as hospitals, associations, patient r,luirements as opposed to tra- insurance carriers, should be included. ditional examination of the care process 4. Consultants and resource people must be as related to professional categorization. trained. 2. This process can be implemented by es- 5. An educational process should he estab- tablishing research teams throughout the lished that would prepare potential im- country. The research teams should be plementers to use results of research,

1N9 APPENDIX: TASK FORCE REPORTS 185 REPORT OF TASK FORCE NO. 3 BEST COPY AVAILABLE Leader Reporter Dr. Muriel Poulin Dr. Bernard Ferber Boston University School of Nursing Director, Bureau of Research Services Boston, Massachusetts American Hospital Association Chicago, Illinois

Recommendations

A. Identify Areas of Needed Research have greater capability(groupings of 1.General considerations (guidelines) smaller hospitals). a.Nurses must assume the initiative for p. Team approach maximizes implementa- research with ,the -:mistance of other tion possibilities. disciplines. 2. Areas of needed -research h. Decisionstoinitiate and implement a.Recruitment and retention of nurses research are made by the power struc- require satisfaction on the job. ture. b. Why do staffing levels vary, both within c.Costs must be considered. a hospital and between hospitals? d. There isa gap between nursing staff c. How can nursing be improved? (providers of nursing service) and upper d.Identify criterion measures of outcomes echelons. How do nurses break into the of health care. Then identify input of coalition of power? nurses to achieve desired outcomes at e.Allrelevantstaffmust beinvolved various points of time. in research endeavors. Involvement is eMeasure and test nursing inputs. What needed to obtain the necessary coopera- is it that is going to make a difference to tion and to satisfy various needs. the patient? f.What do we want to accomplish by f.Study staffsubstitutability. (How can research activiti? staff be substituted?) g.Nursing is presently functioning in an g.%hat are the boundaries (responsibili- acceptable manner. ties) of providers (staff)of care? What h. How do we learn new thin is? is the interaction between staff providing i. In studies to improve mining care, can care? we offer improvements? h. Study organizational structure in which j.Are studies getting to the right people? care is given. How do people act within k. Do we have to be purists? the system? 1. teally good work (research)isnot published. Applied research is not pub- i. / Study system/organizational constraints lished. How is this rectified? that preclude giving patient the treat- rn. Allinputs (disciplines) must he con- ment required. sidered, as all affect output. Study and clarify the expanded role of n. Who will do research? Where do we get the nurse; make a clearer definition of nurses to do research? Individuals on the nurse input. unit must be involved. S.Factors that may facilitate research o. What are the capabilities of hospital a. T. am approach: interdisciplinary with staffto do research. Larger hospitals strong nurse input.

190 REST NWY AVAILABLE 186 RESEARCH ON NURSE STAFFING IN HOSPITALS

b.Utilization of individuals in setting to he to Federal grants. There must be greater studied. Individuals involved in collect- efforts at joint funding. joint funding ing data should be aware of aims, pur- could be between Federal agencies, Fed- poses, and objectives of the research. eral and private sources, etc. c.Increase in number of faunally and well b.Research activity should be included in trained nurses who can do research. the hospital budget. Hospitals should d. The creation of an awareness of the include research in their budget as well need for research. as look for supplement funding. e. The value system of the nurse educator c. The Division of Nursing should make (related to item 4). research in staffing one of its program 4.Obstacles that hinder research a.Insufficient funding, limited sources of d.Health system should put dollars aside funding. for research. b. The hospital administration (adminis- e. There should be Federal funding to trator)does not fully appreciate the support new researchers(recent grad- research nurse. Education is indicated. uates). This could be accomplished by c. Nurse attitudes: the immediate demands facilitating the number of grants to such of an operational system have first (top) priority, pressures of an operating sys- f.Some help should he available to aid new tem. researchers (those with limited research d. The self-recording mechanism. experience) to obtain research grants. e. Value system of nurse educator. C. & D. These topics were not discussed by this 1.Separation of nurse educator, adminis- group. trator, practitioner. g.Less than full commitment of people E.RecommendationsFuture Conferences who are involved in the research activity. 1.More nursing administrators should attend fut,tre conferences. R.Needed Resources 2.Nat.-ow the focus of future conferences 1.Skills (fewer topics). a.Develop skills of nurses to be research 3.Future conferences shouldpresent more coordinators, liaison persons, also studies with nursing input and data, more charged with interpretation of the re- emphasis on studies done by nurses. search effoi t. 4. Should the current meeting have only in- b.Develop skills in assessment and process volved nurses? There was a feeling that of nurses at practicing level. nurses must coordinate and articulate their c.Nurses must know organizational struc- own ideas before a meeting involving multi- ture of hospital: how hospital works; disciplines. Nurses must decide their own what prevents action; what solicits re- function(role)before having a meeting sponses. such as this conference. d. Continuing program to develop new 5.Staffing is not the p,oblem. Effectke utiliza- researchers. tion of people should be the focus as it 2.Funding relates to outcomes (identify nursing out- a.Sources should be developed in addition comes). APPENDIX: TASK FORCE REPORTS BEST COPY AVAILABLE 187 REPORT OF TASK FORCE NO. 4

Leader Reporter Mrs. Margaret Sheehan Miss Pamela Poole Nurse Consultant Nursing Consultant Division of ,Nursing Department of National Health and Welfare National Institutes of Health Ottawa, Ontario, Canada Bethesda, Maryland

Recommendations

A. & B. These topics were not discussed by this b.Analysis ofconflictrelationships be- group. tween the director of nursing and ad- mieditration and trustees on the basis of C.Priorities for Research issues; i.e., what issues produce conflict? 1.Directly related to nurse staffing in hospitals c.Internationalcomparativestudiesof a.Further pursuit of Miller-Bryant studies nurse staffing in hospitals. on the mix of personnel with particular D.Coordination of Work Researchers and Uti- emphasis on cost-benefitanalysis. lizers b. Can we demonstrate the unique contribution of the RN; e.g., the psychosocial 1. Communication components of nursing as opposed to the a.Clearing house for studies; e.g., Division task analysis approach? of Nursing, AN F. b. Encouragereportinginprofessional c.Studies which foCus on the patient; i.e., journals. what is he generating that we have to c.Establish a list of people interested in attend to? What kinds of decisions need nurse staffing, such as this group, and to be made? What kinds of appraisals keep in touch on an annual basis with need to be made? what they are doing. Circulate this list of d. How can we improve staff productivity? current activities. e. How much is spent, regarding effort and d. Regional meetings of researchers and cost and time, on the control of quality utilizers on the. problem of nurse staffing. of care? f. What are the costs of on-the-job training E.Nature of Further Conferences where turnover is high? 1.Narrow the focus. g,... Development and study of models of 2.Provide lead time to prepare oneself for how th' .43 could be. more free discussion and reflection on prob- ii. The effect of technology on nurse staff- lems to be discussed. Distribute papers in ing. advance.

i.More multidisciplinary approaches to 3.Maintain the multidisciplinary approach. . staffing studies. 4.Maintain the Task Force format. 2.Indirectly related to nurse staffing 5. Small interest groups might be organized in a.Studies of director of nursing; i.e., inno- the evenings on a voluntary attendance vative, styles of leadership, etc. basis.

*U.$. GOVERNMENT PRINTING OFFICE: 1913 0-4119-134