Customizing practice : constructing a practice milieu

Understanding how nurses remain in hospital staff nurse positions to practice nursing provides information that can be used in developing programs and practices to increase the stability of the inpatient registered nurse (RN) work force. The purpose of this study was to analyze RN experiences of inpatient nursing and to propose a theory that explained the process of remaining on inpatient medical/surgical units to practice. The major research question was: 'How do RNs manage staying in staff nurse practice positions on inpatient medical/surgical units to practice nursing for periods of 2 years or longer?' This exploratory, descriptive study used grounded theory to guide the research process. Purposive and theoretical sampling, semistructured interviews, and constant comparative analysis provided the mechanisms for participant selection, data generation, and analysis. Participants included 21 RNs from four inpatient medical/surgical units. 'Customizing Practice: Constructing a Practice Milieu' is the substantive theory that emerged as the relationships and dimensions of the core variable (customizing practice), and two major theoretical concepts (negotiating boundaries and positioning) were delimited. This theory is conceptualized as an active, dynamic process in which the nurse designs and constructs a practice based on her personal life needs, professional goals, and perceived professional strengths and limitations. The theory suggests that nurses are concerned with providing care to others and with providing care for themselves as they engage in their nursing work. The theory provides additional insight for increasing staff nurse stability and decreasing staff nurse turnover on inpatient medical/surgical units.

CUSTOMIZING PRACTICE: CONSTRUCTING A PRACTICE MILIEU by Patsy Lea Stapleton A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Nursing The University of Utah December 1995 Copyright C Patsy Lea Stapleton 1995 All Rights Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL SUPERVISORY COMMITTEE APPROVAL of a dissertation submitted by Patsy Lea Stapleton This dissertation has been read by each member of the following supervisory committee and by majority vote has been found to be satisfactory. ~1L1 (" (99-5 Susan L. Beck Charles H. Gregg S. Kay Hart J. Wolfer THE UNIVERSITY OF UTAH GRADUATE SCHOOL FINAL READING APPROVAL To the Graduate Council of the University of Utah: I have read the dissertation of patsy Lea Stapl etOD in its final fonn and have found that (1) its fonnat, citations and bibliographic style are consistent and acceptable; (2) its illustrative materials including figures, tables, and charts are in place; and (3) the final manuscript is satisfactory to the supervisory committee and is ready for submission to The Graduate School. ()t~u If 1995 Date ' Approved for the Major Department ~i~ Linda K. Amos ChairlDean Approved for the Graduate Council Ann W. Hart Dean of The Graduate School ABSTRACT Understanding how nurses remain in hospital staff nurse positions to practice nursing provides information that can be used in developing programs and practices to increase the stability of the inpatient registered nurse (RN) work force. The purpose of this study was to analyze RN experiences of inpatient nursing and to propose a theory that explained the process of remaining on inpatient medical/surgical units to practice. The major research question was: "How do RNs manage staying in staff nurse practice positions on inpatient medical/surgical units to practice nursing for periods of 2 years or longer?" This exploratory, descriptive study used grounded theory to guide the research process. Purposive and theoretical sampling, semistructured interviews, and constant comparative analysis provided the mechanisms for participant selection, data generation, and analysis. Participants included 21 RNs from four inpatient medical/surgical units. "Customizing Practice: constructing a Practice Milieu" is the sUbstantive theory that emerged as the relationships and dimensions of the core variable (customizing practice), and two major theoretical concepts (negotiating boundaries and positioning) were delimited. This theory is conceptualized as an active, dynamic process in which the nurse designs and constructs a practice based on her personal life needs, professional goals, and perceived professional strengths and limitations. The theory suggests that nurses are concerned with providing care to others and with providing care for themselves as they engage in their nursing work. The theory provides additional insight for increasing staff nurse stability and decreasing staff nurse turnover on inpatient medical/surgical units. v I dedicate this work to my mother, a role model and inspiration to me throughout my life; to my father, who instilled the value of doing your best; and to my sisters, brothers, and aunt, who have always provided support, encouragement, and love. TABLE OF CONTENTS Page ABSTRACT . . . . . . . . . . . . . . . . . . . . . . • iv Chapter I. THE STUDY PROBLEM Background and Significance of Statement of the Problem • • Study Purpose and Question • Conceptual Framework • Assumptions • • • • II. LITERATURE REVIEW the Study 1 . . . 1 • • • • 7 7 8 • • • • 12 • • 14 Turnover Research • • • • • • • • •• •••• 15 Nature of Turnover • • • • • • • • • • • •• • 16 Factors Influencing Turnover • • • • • • • • • • • 20 Causal Models of Nursing Turnover • •• • 30 III. RESEARCH METHODS • 35 Grounded Theory Method • • •• •••• 35 Processes of Grounded Theory • • • • • • • • • • • 36 Study Method • • • • • • •• • •• •••• 40 Trustworthiness of Data • • • • • • • • • • • • • 55 Human Rights • •• ••••••••• • 58 IV. THE STUDY FINDINGS • • • • • 60 Participant Characteristics • • • • • • • 60 Theoretical Context • • • • • • • • • • • 65 Theoretical Concepts . . . . . . . 80 Summary • • • • • • • • • • V. DISCUSSION. Research Perspectives • • • • Limitations of the Study • • Implications for Research, Education, and Practice • • • • • • • • • • •• 115 121 122 127 Summary • • • • • • • • • • • • • 128 131 Appendices A. RECRUITMENT FLYER B. CONSENT FORM • • • • C. DEMOGRAPHIC QUESTIONNAIRE D. INTERVIEW GUIDE SELECTED BIBLIOGRAPHY viii . . . Page 133 135 138 142 144 CHAPTER I THE STUDY PROBLEM Background and significance of the Study The primary goal of a nurse administrator is to create a practice environment that promotes and facilitates excellence in nursing practice (Wolf, 1990). One factor contributing to the achievement of this goal is adequate registered nurse (RN) resources. Maintaining adequate nursing resources has been difficult in many hospitals because of increases in patient acuity, changes in hospital reimbursement, changes in nursing care delivery systems, and persistent problems with high nursing turnover (Roberts, Minnick, Ginzberg, & Curran, 1989). currently, two thirds of practicing nurses are employed in acute care settings (Aiken & Mullinix, 1987). Conditions affecting the inpatient nursing practice setting will, therefore, have an impact on a sUbstantial portion of working nurses. Inpatient hospital practice conditions influence the nature of nursing practice, as well as the decisions of RNs regarding continued employment in that setting. Approximately 95% of care in acute care settings is provided by nurses (Curtin, 1986) and is, in many cases, positively or negatively affected by the stability or turnover among nursing staff. 2 Nursing turnover is defined as "the process whereby nursing staff leave or transfer within the hospital environment" (Jones, 1990, p. 20). Nursing turnover rates range from 20% to 200% (Duxbury & Armstrong, 1982; Helmer & McKnight, 1989; Hinshaw, Smeltzer, & Atwood, 1987; Kiely, 1989) and historically have been higher in nursing than in other female-dominated professions (Price & Mueller, 1981a). Turnover costs per hospital vary but are estimated to be as high as $20,000 per nurse for an annual cost to the industry of $3.1 billion (DeCrosta, 1989; Droste, 1987; Hinshawet al., 1987; Jolma, 1990; Jones, 1990). In addition to increasing health care costs, excessive nursing turnover jeopardizes the quality and quantity of patient care; contributes to decreased staff satisfaction, productivity, and morale; and leads to further staff turnover and instability (Alexander, 1988; Curran & Minnick, 1989; Jolma, 1990; Kerfoot, 1988; Mobley, 1982; Moritz, Hinshaw, & Heinrich, 1989; Wolf, 1981). The consequences of turnover have provided the impetus for numerous studies on nursing staff turnover, the majority of which have been conducted using traditional quantitative research methods. These studies have identified organizational and personal variables 3 contributing to turnover and, in many cases, have provided and tested linear models for predicting turnover (Curry, Wakefield, Price, Mueller, & McCloskey, 1985; Hinshaw et al., 1987; Parasuraman, 1989; Price & Mueller, 1981a; Seybolt, Pavett, & Walker, 1978; Weisman, Alexander, & Chase, 1981). Even though these studies have made valuable contributions to the development of the body of knowledge relating to staff nurse turnover, in general, the variables identified as predictors of turnover have explained only 10% to 20% of the variance in nursing turnover (Alexander, 1988; Curry et al., 1985; Hinshawet al., 1987; Price & Mueller, 1981b). Despite the extensive research focus, turnover is still neither well-explained nor understood. The low­explained variance in turnover in these studies may suggest that unidentified variables exist that make significant contributions to nursing turnover; it also is possible that the models of turnover, which have been developed primarily using male subjects (Price & Mueller, 1981a) and expanded on in additional studies, do not comprehensively account for female behavior. Finally, the traditional research methods generally used in turnover studies may not provide the most comprehensive approach for studying this issue. The major focus of previous studies has been on why nurses leave the inpatient setting. A few studies have 4 been conducted that examine nurse stability or why nurses stay or remain within institutions to work (Hart & Moore, 1989; McCloskey, 1974; McClure, Poulin, Sovie, & Wandelt, 1983; Prestholdt, Lane, & Mathews, 1988). Only one study was found that examined nurse stability from an emic perspective (Hutchinson, 1984). This perspective is important if the researcher believes that sUbjective experience and personal meaning are the mechanisms through which phenomena are understood and knowledge is discovered and created. No studies have been found that specifically examine stability in relation to the long-term staff nurse employee, that is, those remaining in the institution for a period of 2 years or longer. This dimension of study can enhance one's understanding of inpatient RN staff retention by uncovering and articulating the perceptions, beliefs, and values that RNs have about remaining on inpatient nursing units to practice. The studies and surveys that have been conducted in relation to nurse stability provide information that highlights the need for additional study in this area. Even though it might be assumed that nurses remaining in the hospital setting generally are satisfied with that practice environment, the literature suggests otherwise. Nurses employed in hospitals have expressed generalized dissatisfaction with nursing (Huey & Hartley, 1988) and rate, as inadequate, many of the same factors that result in the resignation of other staff nurses (Prescott & Bowen, 1987). In general, nurses are feeling overworked, underappreciated, and dissatisfied (Aiken, 1990; Mauksch, 1990). 5 Even though they may be dissatisfied, the employment patterns of nurses seem to indicate that they are leaving their nursing jobs, not the nursing profession. Nurses have one of the highest work participation rates of all female-dominated professions (Aiken & Mullinix, 1987). Only 6% of nurses work outside of nursing (McKibbin, 1990). A recent study of nursing turnover in Louisiana hospitals reported that 72% of nurses who resigned nursing positions sought reemployment in similar clinical positions-many in institutions in the same community (Prestholdt et al., 1988). Similar results were reported in other studies (Misuse of RNs, 1989; Weisman et al., 1981). McCloskey (1974) reported that 59% of the nurses who left hospital staff nurse positions could have been induced to stay in their positions if additional incentives and rewards had been provided by the employing institution. These statistics seem to indicate that nurses are committed to working and remaining in the nursing profession but that they are seeking inpatient practice environments that more successfully meet their personal and professional needs. 6 Nurse stability and nurse turnover are intricately linked to each other and to the quality of nursing care delivered (Kerfoot, 1988; Wolf, 1981). A body of knowledge related to increasing RN stability and retention in the hospital care setting is needed. Understanding the experiences of nurses remaining on inpatient units to practice nursing will contribute to the discipline and practice of nursing by articulating the nature of inpatient nursing practice and providing information that can be used in developing programs and practices that increase the stability of the inpatient staff nurse work force. Exploration and examination of the experiences of nurses remaining on inpatient units to practice nursing provide a new perspective for understanding inpatient nursing practice. The unique perspectives of the present study (the naturalistic design and the emphasis on stability and process) provide additional information about the realities of retaining nurses on inpatient units to practice nursing. This perspective provides a means for articulating, from the perspective of the RN, the experience of remaining on inpatient units to practice nursing, the conditions that facilitate remaining on inpatient units to practice nursing, and the meaning and value attributed to this behavior. 7 statement of the Problem Even though many nurses resign from staff nurse positions on inpatient nursing units after a brief period of time, approximately two thirds remain employed as staff nurses in the inpatient setting for periods of 2 years or longer (Aiken & Mullinix, 1987; National League for Nursing, 1990). These nurses remain despite evidence that the inpatient nursing practice environment is not always a satisfying environment in which to work. The literature is replete with studies on nursing turnover, but few researchers examine and articulate the process of remaining on an inpatient nursing unit to practice nursing. In order for nurse administrators to decrease turnover and to continue to provide high-quality, cost-effective nursing care, programs and practices that increase staff stability must be designed and implemented. This process requires additional research to supplement the existing knowledge base related to nurse turnover. study Purpose and Question The primary purpose of the present study was to explore and describe the work experiences of RNs who had practiced nursing on inpatient medical/surgical units for 2 years or longer. Constant comparative analysis was used to propose a beginning sUbstantive middle-range theory explaining the characteristics and dimensions of the process of remaining on inpatient medical/surgical units to practice nursing. The major research question addressed was: "How do RNs manage staying in staff nurse practice positions on inpatient medical/surgical units to practice nursing for periods of 2 years or longer?" Additional research questions were: 8 1. What are the key factors that affect (facilitate and/or inhibit) remaining on inpatient medical/surgical units to practice nursing? 2. What meaning do RNs attach to remaining on inpatient medical/surgical units to practice nursing? 3. What are the predominant concerns of RNs remaining on inpatient medical/surgical units to practice nursing? 4. What are the major stressors and supports associated with remaining on inpatient nursing units to practice nursing? conceptual Framework Two conceptual frameworks guided this research process. These frameworks are Watson's (1988) theory of nursing as human science and symbolic interactionism, which is the theoretical foundation for grounded theory research. These frameworks provide perspectives relative to the nature of knowledge development and of nursing research that have implications for the present study. Watson (1988) suggested that a human science 9 perspective allows nurse researchers to explore nursing issues in a manner that is more consistent with the nature of nursing. Nursing, as a human science, is concerned with meaning and human values and is based upon A philosophy of human freedom, choice, responsibility. A biology and psychology of holism (nonreducible persons interconnected with others and nature). An epistemology that allows not only for empirics but for advancement of esthetics, ethical values, intuition, and process discovery. An ontology of time and space. A context of interhuman events, processes, and relationships. A scientific view of the world that is open. (p. 16) From a human science perspective, meaning and context are not only valued but are central to understanding. The inner, subjective experiences of individuals rather than the outer, context-free observations provide the focus for human science research. The second framework reviewed was symbolic interactionism. Symbolic interactionism is a down-to-earth approach to the scientific study of human group life and group conduct. Its empirical world is the natural world of such group life and conduct. It lodges its problems in this natural world, conducts its studies in it, and derives its interpretations from such naturalistic studies. (Blumer, 1969, p. 47) Symbolic interactionism was named by Blumer (1969), but the basic foundations of its framework were developed and articulated by Mead (1934). The three basic premises providing the foundation for this framework are that the 10 (a) meaning is the basis for human action, (b) meaning is determined in and through the process of social interaction, and (c) meaning is used and modified through an interpretive process (Blumer, 1969). Traditionally, human action has been viewed as a product of various psychological and social factors that produce it. Meaning has not been identified as having a central role in human action (Blumer, 1969). In contrast, symbolic interactionism posits that it is meaning that is central to human action; therefore, human action will be understood not by examining predisposing factors but by identifying the meaning that an object has for an actor. Mead (1934) suggested that the world of human beings is socially constructed through the process of interacting with the self and others. Human interaction occurs through the use of symbols that represents certain objects in the world. Even though a symbol represents an object, it is not the object itself. Therefore, individuals must interpret how the symbol is being used and what the desired individual and joint actions are in response to the use of the symbol. Through this interactive process individuals are constantly defining and interpreting, redefining and reinterpreting their world. Through this interaction the "negotiation of reality including attributing both identity and cause to action" occurs (Charon, 1991, p. 151). Rather than responding to actions of others without interpretation (nonsymbolic interaction), individuals interpret action, assign meaning to the action, and act based on the definition and interpretation (symbolic interaction) (Blumer, 1969; Charon, 1991). From this perspective, meaning and, consequently, reality are seen as contextual in nature and as constantly evolving and changing. The individual is seen not as one who merely reacts to the environment but as a goal-directed being whose action influences and is influenced by others (Charon, 1991). 11 These frameworks provide an alternative direction for pursuing knowledge development. From this perspective, meaning and, consequently, behavior are a result of the interaction of humans and the psychosocial context of their environment. The elements that are most important to the construction of meaning and behavior can be determined only through examination of the subjective experience. In the present study, the meaning and value attributed to the phenomenon of remaining in the hospital to practice nursing would be expected to influence the actions of nurses and of their colleagues. The meaning and value attributed to a phenomenon can be determined only by exploring the world of the subjects from their perspective (Blumer, 1969; Charon, 1991). The present 12 study explored, from the perspective of the RN, the inpatient practice world in order to understand better the meaning and realities of remaining on inpatient medical/surgical units to practice nursing. Assumptions The first assumption of this study was that hospitals will continue to be important components of the health care delivery system. It was further assumed that because of the essential nature of nursing in the delivery of quality patient care, advancements in medical technology, and the aging of the general population, hospitals will continue to need sufficient numbers of appropriately skilled and experienced RN practitioners to provide quality patient care. The second assumption of the study was that, in order to understand behavior, the meaning behind the behavior must be identified. Behavior is viewed not as a product of specific initiating factors but as purposive action resulting from the meaning that a particular situation has for the participant. The final assumption of this study was that the nurse participants in this investigation are experts about their lives as nurses. The phenomenon of remaining on inpatient medical/surgical units to practice nursing can, therefore, be understood better from their perspective. It was assumed that these nurse participants are knowledgeable 13 about their inpatient nursing practice world and events that influence their decision to remain there. It was further assumed that the participants would be able to articulate and communicate beliefs and meanings engendered in this world when asked to do so. CHAPTER II LITERATURE REVIEW The grounded theory method articulated by Glaser and strauss (1967) was used to develop a new framework from which to study and understand the process of remaining in the hospital setting to practice nursing. The present study was designed to examine turnover but from the perspective of those who remain in the organization to practice. The grounded theory method was used because a paucity of information is available about nurses who remain in hospitals to practice nursing. stern (1985) suggested that a pre study literature review may be a disadvantage for the researcher because it may (a) lead the researcher in the wrong direction of inquiry, (b) lead to premature closure of the study, and (c) provide inaccurate information. However, with these caveats in mind, an initial literature review was conducted to develop the initial study questions and to enhance the theoretical sensitivity of the researcher (Glaser, 1978), that is, the "attribute of having insight, the ability to give meaning to the data, the capacity to understand, and [the] capability to separate the pertinent from that which isn't" (strauss & Corbin, 1990, p. 42). 15 This review of the literature provides a general overview of the turnover research, a discussion of the major categories of variables identified as contributing to turnover, and a synopsis of the findings of five extant models of nursing turnover. Turnover Research Employee turnover has been of interest since the early part of the 20th century and has resulted in the development of a sUbstantial knowledge base relating to turnover (Price, 1977). The comprehensiveness of the turnover data and scientific rigor with which the research has been conducted has been evaluated in several qualitative and quantitative literature reviews (Cotton & Tuttle, 1986; Mobley, 1982; Mobley, Griffeth, Hand, & Meglino, 1979; Muchinsky & Morrow, 1980; Muchinsky & Tuttle, 1979; Porter & steers, 1973; Price, 1977). The majority of turnover research is descriptive in nature and has been conducted using the logical-positivist paradigm. The literature has been reviewed extensively by nurse and nonnurse researchers, with concepts and potential relationships in the turnover process derived from that review. Finally, the turnover theories have been applied to the life setting for support and verification. These efforts have resulted in the identification of a plethora of variables related to turnover and to the development of several causal models 16 that attempt to explain turnover (Alexander, 1988; Curry et al., 1985; Hinshaw et al., 1987; McCloskey, 1974; Mobley, 1982; Muchinsky & Morrow, 1980; Parasuraman, 1989; Pooyan, Eberhardt, & Szigeti, 1990; Prescott, 1986; Prestholdt et al., 1988; Price, 1977; Price & Mueller, 1981a, 1981b; Seybolt, 1986; Weisman et al., 1981). Major weaknesses in the turnover research relate to the use of cross-sectional rather than longitudinal study designs, lack of replication of studies, inconsistent use of standardized measures across studies (with the exception of job satisfaction), and low explained variance in the turnover models. Even though turnover has been the focus of a significant amount of study, there is still much to learn about this phenomenon. Nature of Turnover Turnover can be understood better through an examination of the dimensions of turnover. The following section addresses the definitions of turnover and the types and consequences of turnover. What Is Turnover? Turnover is defined in different ways, depending on the perspective from which it is being examined. Turnover can be viewed from an organizational or an individual perspective. From an organizational perspective, turnover can be 17 defined as "the proportion of organizational membership that voluntarily leaves an organization during a specified time period" (Alexander, 1988, p. 61). From this perspective, turnover is viewed as a characteristic of an organization. Aggregate data are used to establish a turnover rate for an organizational unit. The organizational unit could be the organization, as a whole, or specific subunits within the organization. Recent studies support the notion of significant within-hospital variance in turnover; therefore, the organizational subunit, not the organization as a whole, should be used as the unit of analysis for employee turnover (Alexander, 1988; Hart & Moore, 1989; Prescott, 1986). Turnover also can be viewed from an individual perspective. An individual perspective looks specifically at the individual within the context of the organization and is defined as the "the degree of individual movement across the membership boundary of a social system" (Price, 1977, p. 4). The individual perspective is aimed at determining the factors that influence the individual to withdraw from an organization. Individual turnover has been the focus of the majority of turnover research. Types of Turnover Traditionally, turnover has been viewed as voluntary or involuntary. Voluntary turnover is the "individual movement across the membership boundary of a social system 18 which is initiated by the individual" (Price, 1977, p. 9). In other words, the person quits! Voluntary turnover has been the focus of most turnover research. Involuntary turnover is turnover that is not initiated by the employee (Mobley, 1982; Price, 1977). Most often involuntary turnover is initiated by the orqanization, but it also can occur as a result of events not under the control of the orqanization or the individual such as death, retirement, or other life demands. Turnover can be examined in terms of whether it is avoidable or unavoidable. Dalton, Krackhardt, and Porter (1981) suggested that those who leave for avoidable reasons are distinct from those who stay and from those who leave for unavoidable reasons. Even though minimal, some empirical data support this taxonomy (Abelson, 1987). Consequences of Turnover Turnover most often is viewed in terms of its negative impact on individuals and orqanizations. However, there are also positive consequences of turnover. Some of the positive benefits of turnover to the organization include the (a) infusion of new ideas and methods into the organization, (b) displacement of nonproductive workers, and (c) potential cost reductions through job restructuring and possible automation (Mobley, 1982; Muchinsky & Morrow, 1980). The benefits to an individual might include (a) new motivations and challenges in a new job, (b) increased self-esteem resulting from seeking and finding a new job, (c) decreased stress, and (d) the potential for new opportunities for professional advancement and skill development (Chandler, 1990; Mobley, 1982). 19 One of the most negative consequences of turnover for the organization is cost. Replacement costs for RNs can reach $50,000 (Kerfoot, 1988). Equally important is the potential for decreased staff productivity and efficiency, as well as decreased morale of the remaining staff (Mobley, 1982; Muchinsky & Morrow, 1980). Individual costs of turnover include loss of seniority and benefits and increased stress because of changing jobs. Chandler (1990) suggested that excessive tenure in an organization may lead an individual to a content or structural plateau or both. A person who is "content plateaued" is no longer challenged by the job, and a person who has reached a "structural plateau" has reached the top of the professional ladder. In nursing, moving up the professional ladder generally has required staff nurses to move away from the bedside into management or clinical specialist positions. However, this movement has changed with the development of career ladders for staff nurses. Chandler also suggested that increased individual tenure in an organization results in an increased comfort 20 in maintaining the status quo and increased difficulty in accepting new challenges and directions within or outside the organization. Factors Influencing Turnover Factors related to turnover have been categorized in different ways. For this discussion, however, the factors were categorized as contextual or individual variables. The contextual variables reflect the organizational/ individual work context, whereas the individual variables reflect personal characteristics. contextual Variables For discussion purposes, the contextual variables have been subdivided into organizational and integrative categories. The organizational variables focus on the structural and descriptive characteristics of the organizational work setting, whereas the integrative variables reflect the interaction between the environment and the individual. organizational variables. Organizational variables generally are considered to be conditions and factors within the control of the organization and, therefore, amenable to change by the organization. The consistency with which the organizational variables are identified as contributing to turnover suggests that work conditions have been important factors to consider in turnover 21 (Cotton & Tuttle, 1986; Prescott, 1986; Prestholdt et al., 1988; Weisman, 1982). The organizational variables cited in the literature related to turnover are pay; supervisory characteristics; routinization; autonomy and/or centralization; responsibility; integration or peer group interaction; communication; recognition and feedback; RN staff/patient ratios; overtime worked; work load; professional development; staff relationships such as peers, supervisory, and physician; stress; and work schedules (Alexander, 1988; Hinshaw & Atwood, 1983; Jolma, 1990; McCloskey, 1974; Mobley, 1982; Muchinsky & Morrow, 1980; Muchinsky & Tuttle, 1979; Pierce, Freund, Luikart, & Fondren, 1991; Pooyan et al., 1990; Porter & steers, 1973; Prescott & Bowen, 1987; Weisman et al., 1981). In relation to nursing, the organizational variables seem to address the structure of nursing work within the hospital setting. Despite suggestions that changes in these conditions could reduce turnover, hospital and nurse administrators have been slow to initiate significant change in work conditions (Aiken, 1989). Major changes in work conditions would seem to require a general reevaluation of the basic beliefs and values about the role of the nurse in the hospital and the traditional relationships between hospital administration and nurses. Integrative variables. These variables provide a 22 means through which individual differences in perception of the organization and the environment can be expressed. This category of variables includes perception of job satisfaction, job opportunity, and organizational commitment. Job satisfaction, the "extent of positive affective orientation to a job" (Blegan & Mueller, 1987, p. 230), has been studied extensively as a means of decreasing nursing turnover. Early studies suggested that job satisfaction directly affected employee turnover (Porter & steers, 1973). However, subsequent studies have suggested that job satisfaction is neither the best nor the only predictor of nursing turnover (Curry et al., 1985; Mobley, 1982; Prescott, 1986). Numerous factors and conditions contributing to nurse job dissatisfaction have been identified in the nursing literature. Nursing surveys suggest that the barriers to job satisfaction have changed little ·over the years. Nurses reported being dissatisfied with salary, amount of paperwork, support from hospital and nursing administration, opportunity for continuing education, availability of child care, lack of recognition, and excessive work demands (Huey & Hartley, 1988; Wandelt, Pierce, & Widdowson, 1981). A recent qualitative study identified structural problems such as scheduling, staff shortages, inadequate 23 support services, work overload, lack of support from employers, and stress; inadequate rewards such as pay/benefits insufficient for education and work performance and career advancement; and social and gender issues such as lack of respect/recognition, lack of mutual support, and devaluing of nursing in society as major dissatisfiers for nurses (Seymour & Buscherhof, 1991). Interestingly, some of these variables such as lack of respect, work overload, and insufficient pay were identified as primary reasons for nurse dissatisfaction and high nurse turnover in the early hospital environment (MacEachern, 1932). The literature supports a consistent correlation between job satisfaction and turnover (Mobley, 1982; Mobley et al., 1979; Muchinsky & Tuttle, 1979). The higher the level of job dissatisfaction, the more likely employees are to quit their job (Porter, Steers, Mowday, & Boulin, 1974; Prestholdt et al., 1988; Seybolt et al., 1978; Weisman et al., 1981). In a meta-analysis of 120 studies, cotton and Tuttle (1986) found overall job satisfaction, satisfaction with work, pay, and supervision to be highly reliable correlates of turnover (R<.0005). The relationship to turnover is consistent, but the exact mechanism by which job satisfaction affects turnover is not clear. The predictive models of nursing turnover hypothesize and provide support for job satisfaction as an 24 intervening variable between contextual and individual variables and job termination. Several studies show that job satisfaction has no direct effect on turnover or explains only a small portion of the variance in turnover (Curry et al., 1985; Hinshaw et al., 1987; Parasuraman, 1989; Price & Mueller, 1981a, 1981b; Weisman et al., 1981). The assumption of the literature seems to be that individual employees attempt to balance job satisfactions and dissatisfactions. As long as the job satisfactions outweigh the dissatisfactions, the employee will remain on the job. However, the specific factors required to maintain or change that balance continue to be elusive. In a meta-analysis of 48 studies of overall nurse job satisfaction, Blegan (1993) found strong correlations between job satisfaction and stress (-.609) and commitment (.526). Moderate and small relationships were reported between job satisfaction and communication with supervisor, autonomy, recognition, routinization, communication with peers, locus of control, age, years of experience, education, and professionalism. Interestingly, of the high and moderate correlates of job satisfaction, only poor communication with supervisor and lack of recognition were identified in the survey data as being major dissatisfiers for nurses. The causal models of satisfaction and turnover also 25 have been fairly successful in explaining the variance in job satisfaction, although the models do not consistently use the same variables as predictors of job satisfaction. Approximately 43% of the variance in job satisfaction were explained by nonrepetitive task assignments, opportunities for promotion, age, fair rewards, working the day shift, and appropriate work load (Blegan & Mueller, 1987); 35% by felt stress, individual motivating score, and role conflict (Parasuraman, 1989); and 26% by opportunity, routinization, participation, instrumental communication, and promotional opportunity (Price & Mueller, 1981b). Opportunity refers to the availability of alternative employment options. The nonnursing turnover literature supports a consistent positive correlation between perception of the availability of alternative roles in the environment and employee turnover (Mobley, 1982; Muchinsky & Morrow, 1980; Price, 1977). As the perception of alternate opportunity decreases, so does turnover. Conversely, employee turnover increases as the perception of job opportunity increases. Cotton and Tuttle (1986) found this correlation to be significant (g<.0005). In an analysis of unemployment levels (job satisfaction and turnover), Carsten and Spector (1987) found that as job opportunities decrease that employees tend to remain in their jobs in order to avoid unemployment, even if they are dissatisfied with their current job. 26 opportunity has been less extensively studied in the nursing literature and has received weaker support in terms of ability to predict turnover. Price and Mueller (1981b) and Curry et al. (1985) found that opportunity had only a small positive net effect on turnover. Prestholdt et al. (1988) found, however, that 70% of nurses who resigned their hospital nursing positions sought and found new employment. In addition, those nurses who remained believed that it would not be difficult to obtain another nursing position if they resigned their present position. Nurses traditionally have had numerous job opportunities; this has been especially true in the past few years because hospitals have experienced a severe nursing shortage. Organizational commitment is an attitudinal variable that is conceptually distinct from job satisfaction. organizational commitment is defined as the strength of an individual's identification with and involvement in a particular organization [and is characterized by] (a) a strong belief in and acceptance of the organization's goals and values; (b) a willingness to exert considerable effort on behalf of the organization; and (c) a definite desire to maintain organizational membership. (Porter et al., 1974, p. 604) Porter et al. hypothesized that organizational commitment is a more global construct than satisfaction and that commitment to the organization may override transitory or short-term dissatisfactions with the work environment 27 (Mobley et al., 1979; Porter et al., 1974). Donovan (1980) did not look specifically at organizational commitment but seemed to suggest that many nurses have limited commitment to their employing institutions. She suggested, in contrast to other professionals, that many nurses readily quit their jobs, leaving behind colleagues, seniority, benefits, and work settings. position changes usually are made even though there is no advancement in staff level position and there is marginal or no increase in pay. Therefore, the benefits associated with leaving a position remain unclear. Prescott and Bowen (1987) and Prestholdt et ale (1988) found that nurses who feel committed to their employing institution are less likely to resign their positions, even if working conditions are inadequate. Bateman and Strasser (1984) found that organizational commitment was one of several predictors of job satisfaction and that, as a result of organizational commitment, employees were significantly more satisfied with promotion opportunities, work, coworkers, and supervision. Organizational commitment was evaluated as a component of the turnover process in two nursing turnover models. Both models supported a direct effect of commitment on intent to leave (Curry et al., 1985; Parasuraman, 1989). Even though one model found commitment to have a direct effect on turnover, this effect was interpreted as nonmeaningful (Curry et al., 1985). Individual Variables 28 Individual variables are those that relate to personal characteristics of the employee and are important because of the belief that individual characteristics often influence behavior. This category of variables includes characteristics such as educational preparation, age, tenure within an institution, behavioral intentions, and kinship responsibilities. Kinship responsibilities include characteristics such as marital status, family size and family, work-family conflicts, and alternative sources of income (Mobley, 1982; Mobley et al., 1979; Muchinsky & Morrow, 1980; Muchinsky & Tuttle, 1979; Porter & steers, 1973; Rosen & Korabik, 1991). with the exception of behavioral intentions, it generally has been acknowledged that personal factors do not explain a large percentage of the variance in turnover. However, they consistently contribute to the explanation of the variance in turnover (Cotton & Tuttle, 1986; Mobley, 1982; Muchinsky & Morrow, 1980). Behavioral intention is addressed independently because of the significant contribution to the turnover process. Nonnursing and some nursing literature have supported 29 a negative correlation between tenure within the institution and turnover and employee age and turnover. These studies suggest that as the length of employment or tenure within an institution increases, turnover decreases; and as age increases, turnover decreases (Cotton & Tuttle, 1986; McCloskey, 1974; Mobley, 1982; Muchinsky & Morrow, 1980; Porter & steers, 1973; Prescott, 1986; Price, 1977; Price & Mueller, 1981a, 1981b; Seybolt, 1986; Weisman et al., 1981). Other nursing studies have reported no relationship between age and turnover (Hinshaw et al., 1987; Lucas, Atwood, & Hagaman, 1993) or only indirect effects (Parasuraman, 1989). It has been suggested that age alone does not produce variations in turnover but that other variables, highly correlated with age such as work characteristics and position level, actually produce the variations in turnover (Price & Mueller, 1981a; Weisman et al., 1981). Educational level and kinship responsibility have received mixed support in the literature in terms of their contribution to turnover (Cotton & Tuttle, 1986; Curry et al., 1985; Hinshaw et al., 1987; Lucas et al., 1993; McCloskey, 1974; Mobley, 1982; Price & Mueller, 1981a; Rosen & Korabik, 1991; Weisman et al., 1981). It has been suggested that kinship responsibility will become increasingly important as the number of dual-career families increases and as nonwork values become more 30 important in personal life (Mobley, 1982). The use of behavioral intentions, as the basis for prediction of individual behavior, resulted from the work of Ajzen and Fishbein (1980). They postulated that the immediate precursor to actual behavior is the intention to act. In turnover, the intent to act can be intent to leave or intent to stay. Intent is influenced directly by individual values and normative pressure to act. The intention variable has received consistent support as the mechanism linking individual attitudes and individual turnover response and as being superior to affective variables as predictors of turnover (Cotton & Tuttle, 1986; Curry et al., 1985; Mobley et al., 1979; Mobley, Horner, & Hollingworth, 1978; Steel & Ovalle, 1984). Again, however, even though the relationship to turnover is consistent, the ability to explain employee turnover by intention and antecedent variables remains limited (Curry et al., 1985; Hinshaw et al., 1987; Mobley et al., 1979; Parasuraman, 1989; Price & Mueller, 1981b). Causal Models of Nursing Turnover The results of the studies on the revised Causal Model of Turnover for Nurses (Curry et al., 1985), the Anticipated Nursing Turnover Model (Hinshaw et al., 1987), the Integrated Model of Nursing Turnover (Parasuraman, 1989), the Causal Model of Turnover for Nurses (price & Mueller, 1981b), and the Hospital Staff Nurse Turnover 31 Model (Weisman et al., 1981) are summarized in Table 1. These models of nursing turnover have made a valuable contribution to the understanding of the nursing turnover process but do not provide an adequate picture of the process. The models depict turnover as a linear sequential process in which contextual and individual variables contribute to the development of affective and intentional responses that, in turn, affect actual turnover behavior. The models provide no mechanism for feedback or reinterpretation or redefinition of the event. These models are a representation of the enormous amount of work that has been conducted in relation to turnover. However, the fact that only small percentages of the variance in turnover are explained suggests the need to revisit turnover from an alternative point of view. In summary, examination of the literature revealed that, despite being the focus of extensive study, turnover was neither well-explained nor well-understood. The literature review indicated that turnover is viewed as a product of individual characteristics and job and work attributes. The primary research focus of previous research has been on why individuals leave their positions, not on why they remain in their positions to work. Also, the descriptive, causal models of turnover developed to define turnover consistently have explained Table 1 Causal Models of Nursing Turnover Nursing turnover model Curry, Wakefield, Price, Mueller, and McCloskey (1985) Hinshaw, Smeltzer, and Atwood (1981) (medical/surgical model) Parasuramao (1989) Variables tested Opportunity, routinization, centralization, instrumental communication, integration, pay, distributive justice, promotional opportunity, role overload, professionalism, general training, kinship responsibility, work unit size, job satisfaction, commitment, and intent to leave Tenure expectation, mobility factors, group cohesion, job stress, control over nursing practice, autonomy, professional job satisfaction, organizational job satisfaction, and anticipated turnover Organizational and job experience variables, personal/demographic variables, felt stress, organizational commitment, job satisfaction, and intent to leave Variables with significant direct effects on turnover- Intent to leave (.38) Professionalism (-.01) Commitment (.081 Anticipated turnover (.11)b Intent to leave (.22) 6 months (.14) 12 monthsh Variables with significant total effects on turnover Intent to leave (.38) Job satisfaction (-.10) Kinship responsibility (.01) Not reported Not reported Nursing turnover 13% 6% (medical/ surgical) 6% eN I\J Table 1 (continued) Nursing turnover model Price and Mueller (l981b) Weisman, Alexander, and Chase (1981) Variables tested Opportunity, routinization, participation, instrumental communication, integration, pay, distributive justice, promotional opportunity, professionalism, general training, kinship responsibility, job satisfaction, and intent to stay Personal attributes, job-related attributes, autonomy, job satisfaction, and intent to leave ·Path coefficients. bSeta coefficients. CTotal effects >.05. Variables with significant direct effects on turnover- Intent to stay (-.37) Opportunity (.11) General training (.06)- Hospital A and Hospital B length of employment (-.17, -.12) Intent to leave (.12, .14)- Variables with significant total effects on turnover Intent to stay (-.37) Job satisfaction (-.09) Opportunity (.15) General training (.11) Length of service (-.09) Hospital A and Hospital B length of employment (-.18, -.17) Position level (NS, -.11) Communication (NS, -.12) Autonomy (NS, .16) Job satisfaction (-.09, -.11) Intent to leave (.12, .12) Nursing turnover 18% 8%, 10% w w 34 less than 20% of the variance. This review suggests the need for continued study concerning the employee turnover process. CHAPTER III RESEARCH METHODS The paucity of information available concerning the experiences of nurses remaining in hospitals to practice nursing and the desire to utilize a humanistic approach in the study of nursing phenomenon provided the impetus for the exploratory naturalistic design used for the present study. The grounded theory research method was specifically selected to answer the following research question: "How do RNs manage staying in staff nurse practice positions on inpatient medical/surgical units to practice nursing for periods of 2 years or longer?" The first portion of this chapter describes the grounded theory research approach. The remainder of the chapter addresses sampling techniques, study participant selection criteria, data collection and analysis, and study trustworthiness. Grounded Theory Method Grounded theory is a research method that is used appropriately when little is known about a phenomenon of study or when a different perspective will provide new insight into a well-researched subject (stern, 1980). Grounded theory is an inductive research method based in 36 the principles of symbolic interactionism. This method is used to examine complex social situations and human interactions in order to develop sUbstantive or formal theory that explains the basic social and psychological problems and processes occurring in these contexts (Hutchinson, 1986; stern, 1980). The emphasis on process illustrates the dynamic, nonstatic nature of human interaction and allows for developing theory that transcends time, specific incidents, and specific participants (stern, 1980). The flexibility in the grounded theory method allows the researcher to refocus the research study and to seek additional participants in the study as the data are collected and analyzed and potential relationships in the data are identified. The inductive nature of the grounded theory approach allows the researcher to generate theory from subjective individual experiences. The individual experience provides the empirical data from which concepts and theoretical categories are abstracted and conceptualized. As such, the theory generated in this research process is "grounded" in the data and is not from a priori theoretical frameworks. Processes of Grounded Theory The grounding of a theory in the data manifests a theory that is well-integrated and relevant, and it explains the phenomenon under study (Glaser & strauss, 37 1967). This integrated theory is a result of the creativity of the researcher and use of the fundamental operations of the grounded theory method: (a) sUbstantive and theoretical coding, (b) constant comparative analysis, and (c) theoretical sampling. These operations provide for the simultaneous collection and analysis of data throughout the course of a study. Coding strauss (1987) suggested that the researcher should conceptualize empirical data through the "coding" process. Coding is the process through which data are compared continuously. A code is the product of data analysis. Substantive coding is the process of examining the data line-by-line and establishing codes that name and describe the data (Stern & Pyles, 1985). These codes provisionally describe the experience under study. Substantive codes generally are expressed in terms of the participant's own words and denote the process or action in the study context (Hutchinson, 1986; Stern, 1980). As themes, patterns, relationships, and similarities in substance emerge during the coding process, the data are clustered together into theoretical categories (Stern, 1985). These categories are abstractions of the sUbstantive codes that contribute to their formation. These theoretical codes result from hypothesizing about and interpreting the research data (Wilson & Hutchinson, 38 1991) and move the research from description to theory (stern, 1985). The questions asked during this process include the following: What are these data pertinent to? Is there a category or a property of a category these data relate to? What is the basic problem here (Glaser, 1978)? The dimensions, conditions, consequences, and properties of the categories are proposed through the coding process (strauss, 1987). As each new category emerges, it is compared with all other data for similar or dissimilar properties. If similar properties or dimensions of properties are shared by the categories and relationships among the categories are identified, the categories can be reduced to a smaller number of higher­level concepts until the core variable is identified. The core variable explains a large portion of the variation in behavior, occurs frequently in the data, and has the function of integrating the theory (Glaser, 1978; strauss, 1987). The core variable guides further theoretical sampling and data collection. saturation of the core category provides for a dense and theoretically complete theory. constant Comparative Analysis Constant comparative analysis is the continuous comparison of each datum obtained in the study to each other datum obtained in the study (stern, 1980) and is concerned with generating and plausibly suggesting as many categories, properties, and hypotheses about the study problem (Glaser & strauss, 1967). The researcher generates the conceptual categories used in the theory, the conceptual properties of the categories, and the generalized relations among the categories and their properties through the constant comparative process (Simms, 1981). 39 Glaser and Strauss (1967) suggested four stages to this process: (a) comparing incidents applicable to each category, (b) integrating categories and their properties, (c) delimiting the theory, and (d) writing the theory. Comparing incidents, integrating categories, and delimiting the theory occur as the data are collected, coded, and analyzed. Stern and Pyles (1985) suggested that the process can be divided into the procedural components of (a) collecting empirical data, (b) forming concepts, (c) developing concepts, (d) modifying and integrating concepts, and (e) producing the research report. It should be reemphasized that even though constant comparative analysis, as outlined, seems to indicate a linear research process, it is a process that requires simultaneous data collection, analysis, and interpretation. 40 Theoretical sampling Theoretical sampling is the process of continually tailoring collection of the research data to the emerging theory (Glaser & strauss, 1967). In this process, data are selected intentionally throughout the course of the study in order to provide relevant information about the emerging categories and concepts. Theoretical sampling assures that the emerging categories are grounded in the data and will "fit" the data, thus making the practical application of the theory possible (Glaser & strauss, 1967). In theoretical sampling, completeness of the data, as opposed to obtaining a preselected number of participants for the study, serves as the final determinate for terminating data collection. study Method Design This investigation was an exploratory, descriptive study that used the grounded theory method to explore the processes used by staff nurses staying in the hospital to practice nursing. study Setting The study setting was the medical/surgical units in a 520-bed, not-for-profit, university-affiliated hospital in the intermountain region of the western united states. The hospital was selected because it is a large referral 41 center for the intermountain region and because it had five designated medical/surgical units from which to recruit study participants. The diversity among nursing staff in terms of age, educational preparation, and longevity within the institution was expected to be enhanced by the number of medical/surgical units at the hospital. Data source triangulation was accomplished by selecting participants from different units within the hospital (with different educational backgrounds, years of nursing experience, and age groups) and by using process and theoretical memos. Medical/surgical units were selected as the units for study for a number of reasons. In the literature, it has been reported that nurse turnover rates are higher on medical/surgical units than on specialty units, that job satisfaction rates are lower (Aiken, 1990), and that nursing workloads and intent to leave are higher than for other units (Jolma, 1990). McCloskey (1990) reported that nurses on medical/surgical units also tend to have fewer weekends off and fewer social groups with which to interact. In the past, recruitment to these units has been a problem, with 70% of medical/surgical units requiring 60 days or longer to recruit replacement RNs (National League for Nursing, 1990). It also has been suggested that nurses may see medical/surgical units as units on which "time" must be spent before moving into 42 more specialized nursing units (Curran & Miller, 1990). The increased acuity and age of patients admitted to these units have enhanced the difficulty in staffing these units (CUrran & Miller, 1990). rarticipant Selection Purposive and theoretical sampling were the techniques used to recruit participants. utilization of these sampling techniques precluded predetermining the sample size prior to initiation of the study. Purposive sampling is used in grounded theory research because the context of the experience is considered to be a critical element of the study. The object of a purposive sample "is not to focus on the similarities that can be developed into generalizations but to detail the many specifics that give context its unique flavor" (Lincoln & Guba, 1985, p. 201). By purposefully selecting participants who can illuminate the experience by reflecting upon their personal experiences, maximum amounts of information about the phenomenon can be obtained and understood from the perspective of the participants. Initially, participants were selected intentionally for the study because of their personal knowledge, experience, and willingness to share (with this researcher) their perspective on the process of remaining on an inpatient medical/surgical unit for 2 years or 43 longer to practice nursing. staff nurses eligible for inclusion in this study were RNs who (a) had worked continuously on the same inpatient medical/surgical unit on a full- or part-time basis for a period of 2 years (24 months) or longer, (b) were female, (c) were not under disciplinary review on the unit at the time of the study, (d) were not planning to leave the institution within the next 6 months, and (e) were willing to participate in the research study. Because the majority of nurses (97%) are female (McKibbin, 1990), this study focused on the female experience of remaining in the hospital to practice nursing (Hall & stevens, 1991). Two notes were received from staff members during the study period, suggesting that male RNs also might have experiences that could contribute to this study. This researcher left notes near the recruitment flyer expressing appreciation for this input and the reason for using only female RNs at this particular time. Theoretical sampling guided subsequent data collection as categories and themes about the experience of remaining on inpatient medical/surgical units began to emerge. Data collection continued until data saturation occurred and no new information was being elicited. saturation refers to the completeness of the category and involves the process of expanding, dimensionalizing, and limiting the identified category (Hutchinson, 1986; stern, 44 1980). Procedures consent to conduct the research study was obtained from the Institutional Review Board and the Human Subjects committee of the University of Utah and the participating hospital. Potential nursing units were identified by determining, through the nursing research division of the participating hospital, which hospital units were designated medical/surgical units. A copy of the research proposal was given to the nurse managers on the nursing units being considered for inclusion in the study prior to initiating the study. This information was requested by the participating hospital review board. In addition, at the request of the nurse researcher at the participating hospital, efforts were made to contact verbally the nurse managers of the potential study units regarding the proposed research study. This researcher successfully contacted and made appointments to review the study with four of the five nurse managers of the five designated medical/surgical units. During a 5-week period, repeated unsuccessful efforts were made to contact the nurse manager of the fifth medical/surgical unit. The inability to access the fifth unit precluded the use of the unit for the research study. Two of the four nurse managers contacted suggested 45 that the nurse researcher attend scheduled unit staff meetings and present the proposed research study to the staff at that time. The researcher attended one staff meeting on one unit and two staff meetings on another unit and briefly explained the study purpose, data collection techniques, method for enrolling in the study, and estimated time commitment. The staff was asked to participate during off-duty time. The staff was assured that participation in the study would be confidential. Questions related to the study were answered during that time. The nurse manager was present during the presentation on one unit. The other nurse manager was present during one of the two staff meetings on the other unit. The remaining two managers did not invite the nurse researcher to attend scheduled staff meetings. This researcher briefly met with these managers at a convenient time to review the proposed study. The managers received a copy of the research protocol prior to the meeting and pursuant to the request of the Institutional Review Board. However, during the meeting, the specific inclusion criteria for the study were not reviewed in an effort to maintain nurse participant confidentiality. The recruitment flyer, which had been designed for the study, was posted in various prominent locations on each of the nursing units, including the staff bathroom, 46 the main nursing station, and the staff nurse lounge (see Appendix A). After 10 weeks, the initial recruitment flyer was revised, indicating that the study was continuing and that additional participants were still needed. A flyer remained posted on the unit for the duration of the study. On one unit the flyer and collection envelope were replaced on two separate occasions because the material had disappeared from the bulletin board. RNs willing to participate in the study filled out an information card that requested the participant's name, telephone number, tenure on the unit, and best time for the researcher to call. The informant then placed the information card in a sealed information collection envelope located next to the flyer. The researcher checked the collection envelope and picked up the information cards every other day. The potential nurse participants, following identification, were contacted by phone by the investigator to determine if they continued to be interested in participating in the study. This researcher determined at initiation of the study that three attempts would be made to contact the RN staff volunteering to participate in the study. If, after three attempts, no contact was made, the participant would be excluded from the study. All participants were contacted within three 47 attempts for the initial interview. During the telephone call, the purpose of the study was reviewed and reassurances were given regarding confidentiality. The participant was asked to schedule an interview lasting approximately 60 minutes at a place and time convenient for them. The participant was reminded that the interview would be audiotaped. The researcher also explained that some notes probably would be taken during the interview. The participants were asked during the initial phone call if they would participate in a shorter, follow-up phone interview at a later time. The researcher explained that this interview would be conducted in order to clarify findings and to obtain additional information that might be needed as the study progressed. The participants determined where the interviews would take place. Eight selected their home; 11 selected the hospital in an area on or off the unit; 1 selected the researcher's home; and 1 selected another place. Interviews scheduled at the hospital were conducted prior to or after the work shift or during off-duty lunch time. The researcher had an initial concern about participant confidentiality for interviews conducted on the hospital unit. However, when questioned, the participants expressed a lack of concern about this issue. One participant clarified that specific responses would be 48 confidential. Because the participants selected quiet, infrequently used rooms on the unit for the interviews, minimal to no interruptions occurred during the interview sessions. At the time of the interview, the purpose of the study was reviewed and any questions and/or concerns about the study or participation in the study were addressed by the investigator. All participants were reminded that the interview would be audiotaped and that notes would be taken during the interview. All participants agreed to this process. The participants were assured that participation in the study was confidential and that any information that might be quoted directly would not be linked to them in an identifiable manner. The participants were assured that withdrawal from the study was possible at any time and that it would occur without negative consequences. Prior to beginning the interview, the written consent to participate in the study was reviewed and (see Appendix B) obtained from each nurse participant. Each participant received a personal copy of the consent form. Data Generation Data generation for this study occurred over a 6- month period. The primary data sources were the participants, and the primary instrument for data collection was the researcher. The RNs shared insights into their personal experiences as inpatient medical/surgical staff nurses. 49 Initial interviews were conducted with 21 participants. Of the 21 participants, 15 were reinterviewed for verification and/or clarification of previously received information. Thirty-six interviews were conducted. The researcher was unable to reinterview 6 of the original participants. The researcher discontinued attempts to contact 5 of the original participants after three unsuccessful telephone attempts were made. The researcher was unable to contact the remaining participant because of the inability to obtain a work telephone number. The initial interviews lasted between 45 minutes and 1 hour and 50 minutes. The initial interviews were conducted in person because the researcher believed that the person-to-person contact would be helpful in establishing rapport with the study participants. The second interviews were conducted via the telephone, lasting from 7 minutes to 35 minutes. All interviews were audiotaped and transcribed within a 2-week time frame following conclusion of the interview session. A structured format was used to obtain pertinent demographic data about the study participants (see Appendix C). A semistructured interview or interview guide (see Appendix D) was the principal technique used in 50 data collection (Patton, 1987). This researcher began the interview by informing the participants that it was important to hear about inpatient nursing from their point of view and that this researcher was there to learn from them. Each participant was asked to describe in detail, as possible, her experience of inpatient nursing. Each participant also was asked how she felt about working on her unit and what had facilitated and hindered her remaining on her unit to work. The participant also was asked to share any experiences she had had with thinking about leaving and what had been involved in that experience. During the course of the study as categories and themes began to emerge from the data, additional avenues and directions were explored as they became apparent. The interview guide provided the flexibility needed for clarification and redirection of the inquiry based on the responses of the participants (Marshall & Rossman, 1989). For example, after the initial interviews, it became apparent that stress was one issue mentioned repeatedly by the participants. As a result, a question relating to the participant's perception of stress was added to the interview guide. A semistructured interview guide was used because it provided the participants and the researcher with (a) a general focus for the interviews while still revealing the 51 definition and structure of the study problem from the perspective of the participant (Bernard, 1988; Marshall & Rossman, 1989) and (b) a mechanism for helping to assure that the same general topics were covered when addressing different study participants (Patton, 1987). The initial interview questions were constructed by the researcher and conceptualized using the format established by Patton (1987). The questions are classified as experience/behavior questions, opinion/belief questions, and feeling questions. Experience/behavior questions seek to determine the experiences, activities, or behaviors of the participants, whereas opinion/belief questions seek to illuminate what participants think about the world and what their desires and values are. Feeling questions attempt to elicit the emotional response of participants to what is occurring around them. The interview guide contained the general questions used to guide this study. The interview guide and demographic questionnaire were pilot tested on two staff nurses prior to initiation of the study. The demographic questionnaire was revised slightly based on the review of that information. No changes were made in the interview guide. The interview questions had to be clear and focused in order to make the participant comfortable and to facilitate the exchange of information. Specific probes 52 were used, as needed, during the interviews to facilitate the interview process. silence was used to allow the participants to reflect and fully articulate their thoughts. Affirmative gestures or noises also were employed to encourage the participants to continue with their descriptions. Directive and clarification probes were used to direct the interaction toward a specific area of interest and to provide more contextual information (Bernard, 1988; Patton, 1987). The participants talked freely during the interviews, and it seemed the nurses wanted and were eager to share their stories. The interviews, despite some initial nervousness on the part of the researcher, seemed comfortable and open, and it seemed there was a free and honest exchange of information. The participants were free to ask questions if they chose; most did not. Some of the informants were less expansive in their explanations than others, thus needing more probes and direction during the interview. Data Management and Analysis The Ethnograph computer program was used to facilitate management of the large amounts of qualitative data obtained during the study. Data collection and analysis were conducted simultaneously using the processes of coding, constant comparative analysis, theoretical sampling, and memoing. 53 The recorded interviews and field notes were transcribed, with the exception of two sessions, within 3 days following conclusion of the interviews. Each transcript, before coding, was read in its entirety to obtain an overall feel or perception of the interview. General thoughts and ideas about the interview were written on the transcript at that time. After the initial reading, the transcript was examined line-by-line and the sUbstantive coding process began. Substantive codes, or codes describing the activity or process occurring in the data, were identified. Many of these codes were derived directly from the words of the participants. Some of the initial sUbstantive codes included "guiding others," "advancing," and "feeling stuck." As new codes were identified, they were compared and contrasted to the developed codes. As data collection and analysis continued, general categories or themes began to emerge. These categories were developed as the similarities and potential relationships occurring among the sUbstantive codes were identified; then the codes were collapsed into a larger category. The category of "moving better" was developed from the substantive codes of "doing other things," "making it better," "learning," and "developing." Hemoing. Process memos or descriptive field notes were maintained throughout the course of the study. 54 Process memos are used by the researcher to record observations about the context of the interview, as well as the behaviors and interactions of the participants and the researcher (Sandelowski, Davis, & Harris, 1989). For the present study, memos were written or tape-recorded immediately following the interview, as suggested by Patton (1987). Memos provided the opportunity for the researcher to examine personal thoughts about the meaning and significance of the interview and to monitor key ideas and phrases that emerged from the interview. The notes also provided a mechanism for monitoring potential questions and information that should be sought during the next interview. stern and Pyles (1985) and Hutchinson (1986) suggested that theoretical memos help to document the thinking and decision-making process of the researcher, to capture ideas that might otherwise be lost, and to provide the conceptual groundwork for generating the theory. Theoretical memos were written in the margins of transcripts throughout the coding process to retain thoughts about the emerging categories, properties, and potential relationships in the data. A large wall chart helped to explore visually the categories, potential relationships, dimensions, and properties evolving from the data. Hutchinson (1986) suggested that by using the basic 55 operations of the grounded theory method and personal clinical and academic knowledge that a researcher transcends the empirical nature of the data and thinks in theoretical terms. Thus, the researcher produces "a parsimonious theory, one that is comprehensive without being wieldy, consists of a few theoretical codes, a greater number of categorical codes and a majority of in vivo or substantive codes" (p. 122). continuous return to the field during the study helps to ensure the theoretical completeness and appropriateness of the theory. The intent was to produce a substantive theory that would explain comprehensively the processes involved in remaining in the hospital to practice nursing. Trustworthiness of Data Because the goals and purposes of qualitative and quantitative research are different, the criteria used to judge the reliability, validity, and resultant rigor of the study also may need to be different (Leininger, 1985; Lincoln & Guba, 1985). The criteria suggested by Lincoln and Guba (1985) and further explicated by Sandelowski (1986) as appropriate measures of rigor in qualitative research are credibility, applicability or fittingness, auditability, and confirmability. 56 credibility The credibility or truth value (internal validity) of a grounded theory approach is measured by the accuracy with which the study results depict the experiences described by the study informants. study results are considered credible if the informants recognize the experience as their own from the reconstructed description and interpretation of the experience presented by the researcher (Lincoln & Guba, 1985; Sandelowski, 1986). credibility of the findings results from the researcher continually reviewing the emerging constructions with the study informants (Lincoln & Guba, 1985) and examining personal "behaviors and experiences as researchers in relation to the behavior and experiences of subjects" (Sandelowski, 1986, p. 30). The ability to separate or bracket personal experiences or preconceptions from those of the subjects helps to avoid "going native," which is a condition in which the researcher is unable to distinguish his or her experiences from those of the subjects (Miles & Huberman, 1985; Sandelowski, 1986). Process memos maintained by the researcher, member checks or confirmation of emerging theory with the study participants, peer debriefing, sharing the data, and analysis with colleagues were techniques used to minimize the threats to credibility as the data collection and analysis progressed. 57 Fittingness Fittingness refers to how meaningful and applicable the results of the study are to the audience and how well the study results "fit" into the "context outside the study situation" (Sandelowski, 1986, p. 32). The purpose of qualitative research is to describe phenomena that occur in a particular time and context. If a researcher attempts to transfer the information to another context, as much information as possible must be obtained about the similarities of the contexts (Lincoln & Guba, 1985). This researcher attempted to provide a rich, dense description of the phenomenon under study in order to allow others to determine if this information could be used in other contexts. Au4itability Reliability is established in quantitative studies through consistent and stable findings demonstrated by replication of the study. Reliability in qualitative research cannot be determined through the same processes because of the unique nature of the investigation (Leininger, 1985). consistency and stability in qualitative studies are addressed more appropriately by auditability criteria. Auditability is the ability of a researcher to follow the decision trail of the principal researcher and to arrive at similar conclusions given the "researcher's 58 data, perspective, and situation" (Sandelowski, 1986, p. 33). Theoretical and process memos were the mechanisms used to document decisions made by the researcher regarding "hunches" about the data, what new information or direction of inquiry was required, and which informants were needed to provide the necessary data. These memos and the theory that emerged from the study document a logical, comprehensive explanation of the processes used by nurses staying in the acute care setting to practice nursing. Confirmability Confirmability refers to the ability to determine that the study results are grounded in the data, were arrived at logically with limited researcher bias and use of a priori theoretical concepts, and have sufficient explanatory power (Lincoln & Guba, 1985). Confirmability is possible by using the audit trail and descriptive memos, and it is achieved when the criteria of fittingness, auditability, and credibility have been met. Human Rights The study proposal was submitted to the University of Utah, participating institution, and Institutional Review Board for approval. A consent form was given to each participant to read and sign prior to participating in the study. The researcher answered any questions the 59 informant had about participating in the study. An explanation was given to the participants that they were free to discontinue participation in the study at any time without negative consequence, if desired. Confidentiality was maintained through the use of participant code numbers. study data were referenced through code numbers-not individual participant names. Names of the study participants were available only to the researcher and were maintained in a secure environment. CHAPTER IV THB STUDY PINDINGS This study was designed to examine the experiences of inpatient staff nurses and to develop a sUbstantive theory that provided a tentative explanation for the process of remaining in the hospital to practice nursing. This chapter delineates the characteristics of the study participants, the contextual data relevant to the SUbstantive theory, and the concepts and relationships of the theory itself. participant Characteristics Twenty-one RNs participated in this study. Three volunteer participants were excluded. Two participants failed to meet the inclusion criteria, and 1 failed to keep a scheduled interview appointment despite attempts to follow-up. Educational preparation A majority of the informants received their basic RN educational preparation from an associate-degree (43%) or baccalaureate-degree program in nursing (43%). The remainder of the informants (14%) were diploma-prepared RNs (see Table 2). Approximately 19% of the participants Table 2 Participant Basic Educational Preparation, Employment status, Marital status, Religion, and Income Source 61 Number Percent Basic educational preparation Associate degree Diploma Baccalaureate degree (nursing) Employment status Full time Part time Marital status Not married Married Living with significant other Religion LDS Non-LDS Income source* Sole income source Primary income source 9 3 9 15 6 4 16 1 17 4 6 9 43 14 43 71 29 19 76 5 81 19 29 43 *Does not equal 100%, not mutually exclusive categories. 62 received a baccalaureate degree in nursing in addition to their basic education. One participant was advancing toward a master's degree in nursing (4%), and 1 participant had received a nursing specialty certification (4%). Employment status At the time of the study, 71% of the participants were employed full time and 29% were employed part time (see Table 2). Further analysis revealed that 47% of the nurses had been employed full time for their full tenure on the unit, whereas 14% had been employed part time for their full tenure on the unit. Several of the participants (39%) had changed their work status at some time during their tenure on the unit. Marital status/Religion/ Income Source A majority of the participants was married (76%) and members of the Church of Jesus Christ of Latter-day Saints (LOS) (81%). six of the participants were the sole income source for their household, and 9 were the primary income source for their household (see Table 2). Personal and Household Income The net personal income for one third of the participants (7) was between $20,001 to $30,000 per year. The net household income for the majority of the participants (12) was greater than $50,001 per year (see Table 3). Age and Years of Registered Nurse ExPerience 63 The mean age of the study participants was 39.7 years, with 14% of the informants between the ages of 22 and 30, 48% between the ages of 31 and 38, 14% between the ages of 42 and 50, and 24% between the ages of 51 and 58. The age range was 36 years, with a minimum participant age of 22 years and a maximum participant age of 58. The mean of nursing experience was 13.2 years. The range was 34 years, with the maximum number of years of nursing experience at 36 years (see Table 4). Years Employed On unitl In Institution The range for years of employment on the unit was 22 years, with a minimum of 2 years and a maximum of 24 years' continuous employment. The mean length of employment on the unit was 8.2 years. Approximately 71% of the participants had been employed on the unit for 10 years or less. The remaining 29% had been on the unit for 11 years or longer. The mean length of employment in the institution was 11.5 years. The range was 31 years (see Table 4). 64 Table 3 Personal and Household Income Personal Household Income Number Percent Number $10,001 to $20,000 2 10 1 $20,001 to $30,000 4 19 3 $30,001 to $40,000 7 33 2 $40,001 to $50,000 4 19 3 $50,001 to $60,000 4 19 6 $60,001 or > 0 0 6 Ii = 21. Table 4 Registered Nursing Experience, Age, and Length of Employment on the unit and in the Institution Mean years Minimum Age 39.7 22 Years of registered 13.2 2 nursing experience Years on unit 8.2 2 Years in institution 11.5 2 Percent 5 14 9 14 29 29 Maximum 58 36 24 33 study participants and National Averaqes 65 The goal of a qualitative study is not to generalize data, but comparing statistics of the study participants to national norms provides additional information to consider as the results of the study are reviewed. A majority of RNs in the United states is female (97%) and Anglo American, with an average age of 39 years (McKibbin, 1990). A majority of RNs (70%) is married with children at home and works full time (68%), which is the same as the participants in this study. Twenty-seven percent of RNs have basic preparation at the baccalaureate-degree level, 25% at the associate-degree level, and 40% at the diploma level (McKibbin, 1990). More participants in this study had attained a baccalaureate degree in nursing than the national average. with the exception of educational preparation, as a group, the participants in this study were close to the national averages for RN demographics. Theoretical context An important component of the grounded theory process is the construction of the context in which behaviors associated with the emerging theory occur. context refers to the "environment or setting where the behavior occurs" (Hutchinson, 1986, p. 121). In this study, three distinct contextual elements emerged: (a) change, (b) instability, and (c) challenge. 66 change A key contextual element is change. As the nurses shared their experiences of nursing within the inpatient setting, it became apparent that their world is a world of constant and continuous change. One nurse explained: I mean, anybody who's been in nursing knows that your experience this year is totally different than it was last year. The way we, the way we treat patients, the medications we give. I would be afraid to go to a hospital where the nurses didn't stay in nursing. In this context of constant change, their professional and personal behaviors exist and are acted out. Some of the change is planned; much is not. Planned change can be thought of as change that occurs over time as a direct result of a selected course of action. Unplanned change can be thought of as change that is more immediate in nature and requires reconfiguring or readjustment of a proposed course of action. For the present study, personal change was defined as change that occurred outside of the professional work setting. Professional change was change that had a direct impact on the work setting or nursing work within the work setting. The planned personal changes that occurred seemed to result from changes in the family structure such as marrying, divorcing, or separating from a husband or significant other and having children; physical location such as buying a home or moving to a new location; or 67 priorities in life such as time spent with family or meeting personal needs and need for continuing nursing education. The unplanned changes that occurred in their personal lives primarily involved temporary illnesses of self or family members and unexpected financial demands. The participants, while recognizing the connection and integration of their personal and professional life, attempted to separate these two lives as much as possible. Many of the planned structural and functional changes occurring in the practice world of the nurse participants were the direct results of present-day social and economic concerns related to health care cost and access. In addition to the planned changes, they also faced hourly unplanned changes in work, patient, and personal life conditions. within the time frame of this study, several changes were occurring in health care, which had a direct impact on the work experiences of the nurse participants. Even though significant changes were occurring in health care outside the institution and unit level, most relevant to the participants in the present study was the change that directly affected the unit and them as individuals. At the community level, new and continued efforts were being made to curb the cost of health care while maintaining quality and improving access. These efforts were being played out in decreased lengths of patient 68 hospital stay, the continued shifting of health care delivery from the inpatient to the outpatient setting, the emphasis on managed and capitated care, and fierce competition among local health care facilities for their market share of patients. Economic constraints and a prevailing philosophy of doing more with less resulted in hospitals downsizing, changing staffing mix levels, and restructuring nursing care delivery models. The changes contributed to the diminished demand for RNs in the hospital setting. In contrast to the nursing shortage that had persisted for several years, the "now-adequate" supply of nurses has been exemplified by the sUbstantial decrease in inpatient RN nursing opportunities within the intermountain region. One nurse explained: Well gosh, if I was going to quit now, I think I'd have to think about it pretty hard because there's not all that many jobs out there. It's not that I don't think I could get a job. I guess I could if I wanted to, but one of the things right now is the-you know, the end of the nursing shortage. The institution in which the present study was conducted was implementing and participating in a number of the local change strategies. During the study time period, the institution made the decision to downsize in terms of numbers of management staff. Two of the study units found this strategy to be of questionable relevance to them, whereas two of the units were anticipating upcoming vacancies in their nurse manager positions. Even though these management positions initially were not targeted in the downsizing plan, elimination of the positions would help reach the overall goal of decreased numbers of management staff within the institution. The RN participants were unclear if the management positions would be refilled. A possibility existed that a major reorganization would occur on the units with vacant manager positions. 69 This reorganization might include incorporating the units with vacant manager positions into the management scope of the remaining nurse managers. This possibility was a relevant potential change for nurse participants in the two affected units. One nurse shared her thoughts about the potential change: Um, there is a manager that works in this hospital and she put forth her preference for this unit; but from hearing her staff talk about things that she has done and how she manages, I think, um-this is bad. • • • So if she ever got, I mean I wouldn't like-quit the day she got here, I hope that I have enough open-mindedness to try and give her a chance, but I think if I got a manager that didn't give support-once you've had that I don't know that I could go back to having someone who doesn't give you the support. Another nurse shared similar feelings: If we get a new head nurse who I feel is too manipulative, who doesn't support shared governance or differentiated practice or doesn't support the staff • • • if it was an administrative, I don't know, puppy • • . if it was, then I would quit. I could see that happening very easily. In fact, that's what I'm afraid of. These views illustrate the uncertainty that accompanies change. 70 Prior to initiation of the present study, the head nurses were asked to decrease the nursing unit budgets by 3% and were expecting that further reductions would be required. The request for budget reductions had several major implications for the nursing staff. In an attempt to reduce budget and to maintain patient care quality, the nursing units were implementing new patient care delivery models. One of the participating nursing units used a dyad system, that is, RNs and nursing assistants (NAs) working together to deliver care, as a trial nursing care delivery method. The remainder of the units used primary care nursing and total patient care nursing as the primary modes of patient care delivery. On these units, new NAs were being hired and trained, and the RN staff was being asked to reorganize and reevaluate their care delivery role and to incorporate the new members into their health care team. Change included delegating patient care activities to the NA that previously had been the responsibility of the RN. Nurses had to learn and utilize new skills relating to appropriate delegation, supervision, and feedback. Many of the nurses had not worked with NAs before, which constituted a major change for many of the nurse participants. Another important change was revamping the "patient 71 assignment grid." Based on patient numbers and acuities, the grid was the mechanism used by the staff to plan and make nursing assignments. A specified number of patients was assigned to each nurse based on the complexity of care required by those patients. Revamping the grid resulted in an increase in the number of patients assigned to each nurse, thus increasing the workload for each nurse. This change was relevant on all the participating units because it directly affected nursing workload and the perceived ability of the nurse to complete required patient care activities. Another change that occurred just prior to the beginning of the study was the discontinuation of the nursing career ladder for the nursing staff. The career ladder was a mechanism to offer the staff nurse the opportunity for professional advancement without moving from the patient care role. The career ladder offered an opportunity for recognition as a nurse with a higher competency level but also was associated with a significant wage increase. The relevance of this change for each nurse seemed to be related to how deeply the RN associated with this advancement mechanism. The hospital also had decided to eliminate the education benefit that had been utilized by several of the RNs. 72 Instability Another contextual element that emerged as the nurses shared their work experiences was instability. Instability could be considered related to change, but it also is unique in adding the dimensions of suddenness and unpredictability to the environment. Instability was reflected in individual work assignments, patient conditions, and turnover of patients on the units. One participant explained: I never know from 1 minute or one night to the next night what I'm going to be faced with. Um, it's always something different. It never stays the same. Another participant shared her experience: I have this 40- or 50-year-old woman in the bed who is now a 12-year-old girl and was hurt and crying and venting you know, and so I just had to sit there and help her and all the other little things had to wait because she was hurting you know and sort of try and put her back together for the night because it can happen very quickly. A contributing factor to the instability experienced at the unit and individual level was the planned budget reductions. These reductions had a heavy impact on scheduling and overtime, and they contributed to unplanned scheduling changes. Maximum efforts were made within the hospital to decrease overtime and to utilize available nursing resources in the best possible way. Decreased overtime and efficient resource utilization were experienced by nurses being requested to float to other units within the hospital and even units in other hospitals within the consortium that were staffed inadequately. RNs also were requested to use vacation 73 time and/or take a day off without pay if the unit census was low and the unit was overstaffed. Notification of these changes usually occurred 1 to 2 hours prior to beginning the shift or at the time of arrival on the unit for the assigned shift. This instability in staffing resulted in the RNs being unable to predict when they would be working, what type of work they would be doing, and where they would be working. These requests increased the RNs' anxieties about receiving their contracted number of work hours during the week and also about their abilities to plan for and work in an environment of their choosing. As 1 RN explained: The last shift I went to [X Hospital] 2 days, uh, came back here; the next day I got sent home after 4 hours, came back the next day and got sent to cardiac, cardiology unit for 4 hours, and then they sent me home again. So, I'm not getting to spend time with the patients that I feel most comfortable with • • • where I feel like my expertise is. People are losing their hours, and people are getting worried about their benefits. Another RN expressed: So, if I'm budgeted for 24 hours and I got sent home 8 hours last time, last week, and if I can't make those up, then I just lose my benefits. Plus, I have to use my vacation to fill in the hours that I miss so I have no vacation. 74 sick calls also exacerbate the instability in the work environment. Sometimes it is possible to find an alternative person to replace the ill caller; sometimes it is not. In either case, the result of the ill call is a redistribution of workload with a reconfiguration of assignments and adjustment to a new configuration of the nursing work group. Instability also was expressed in relation to patient turnover and patient conditions. The instability in relation to patient turnover was reflected by this RN who stated: We have an average of about 10 to 15 turnovers on the floor a day and on some days it has been in the 40s. We have had 2 days in the last few months with a 48 turnover on a 46-bed unit. The whole patient load. Patient turnover results in nurses redesigning and reordering their work in order to incorporate the newly assessed personal and physical patient care needs and demands into their work assignment. In respect to the effect of patient turnover, 1 nurse explained: Now the days when I have my best day is when I have no discharges and I have the same three to four patients all day long. Because, even if they are difficult, I can get into a routine and plan what they need and go throughout the day. Whereas, if I have four patients and even if two or three of them go home, you get a transfer out of one unit and a transfer out of another unit, and you've got to get to know an entirely new patient and what they might need, and it is just kind of hard to do. 75 The instability of patient conditions also affected how nurses negotiated practice in the inpatient environment. Nearly all the nurse participants believed that patients are more sick now than they have been in the past. Because of the increased intensity of their illness, there is less time between when observations are made and when the patient's condition might deteriorate. Nurses know that because of this instability in the patient's condition, they need to be particularly vigilant in their patient observations. Unpredictability and instability were associated with this environment, but there was a certain degree of "predictability" associated with the unpredictability. Even though the nurses could not predict when or what type of event might occur, they were confident that some event would occur. The events that might occur included alterations in patient conditions, assignments, staffing, personal interactions with patients and/or other staff, or a change occurring at the unit or organizational level. One nurse explained: If you have an empty room, plan on a transfer. I mean that's just the deal with it. • • • Here, if you have somebody go home, somebody's coming in 90% of the time. That's just how it is. The ability to predict in an unpredictable situation seemed to add some stability to the work environment. Instability occurred in conjunction with challenge, the 76 third contextual element that emerged from the data. Challenge Challenge is the remaining contextual element that emerged from the experiences of the nurse participants in this study. The challenges experienced by the participants ranged from easy to difficult, from basic to complex. Overall, the context of challenge, as described by the participants, provided a sense of forward movement, energy, achievement, growth, and productivity. Challenge seemed to provide stimulation for the participants and required them to utilize their abilities, energies, and resources efficiently and effectively. The challenge provided an opportunity for the participants to "stretch" themselves in an attempt to move beyond their present level of practice. Challenge seemed to be viewed as a positive influence; that is, it presented a means for growing and developing by identifying and addressing new areas of skill development. The study participants were challenged with meeting the physical, emotional, and cognitive demands of their work. The physical demands included working 8- to 12-hour shifts, consecutive and nonconsecutive days, shift rotations, and nights and weekends when others and members of their family were at home engaging in "normal" off-duty activities. The physical demands also included performing "hard" physical labor. This labor included physically moving, lifting, and transporting patients; walking the hallways and moving from room-to-room in the process of providing care; and "battling" with equipment attached to and necessary for the care of the patient. For participants on the night shift, the physical challenge also included getting sufficient sleep as they adjusted to sleeping during daytime hours. One nurse shared her experience with the physical demands of the job: Oh, the lifting. I think that is probably the biggest problem when you have real heavy patients, and you have to do a lot of lifting. It's hard. It's hard. You can get a lot of people in there, but it still doesn't make it easy. I mean you can distribute it around, but it's still hard. Um, sometimes, just due to the nature of our floor, there's a lot of coming and going • . • a lot of running around, patients coming back and forth, and, so sometimes, you know you get tired. You get tired of being on your feet constantly and sometimes not having a break. Emotional challenge also came from working with "difficult patients" and families. The term "difficult 77 patients" did not necessarily carry with it a negative connotation. Difficult patients were identified as those patients who required significant investment from the nurse in terms of physical, emotional, and/or time. These patients included those who needed additional emotional and physical support in dealing with, and working through, their diagnosis and treatment; those who were "on their lights" all the time with small, seemingly inconsequential, demands; and those who were "emotionally unstable." One participant explained: Lately, it is putting up with difficult patients-patients who are not emotionally stable. Patients that are demanding. Finding myself with more patience toward them is challenging to me, so that is what I am working on lately. Another RN suggested: What is difficult for me may not be difficult to you or someone else. Say a patient who has to get up to the commode every 15 minutes-pick her up and get her to the commode, but try and find a way to solve the problem-to work it better. You know, make the problem not a problem but a solution. You can use that information for the next patient and say, "This works this time, let's see if it works this way. If not, let's do it a different way." So, I think that is the challenge. 78 Rapidly advancing medical technology and changes in the care delivery system presented additional challenges for the participants. Not only were new technical skills constantly being required, but these skills were accompanied by the development of additional cognitive skills that allowed for the integration of new technical knowledge into the delivery of care to the patient. The nurses were challenged to know what to look for, what to do, and when to take action in their daily practice arena. One participant explained: Of course we have the LVAD, the left ventricular assist device. We have a team of nurses. • We are gradually working toward teaching everybody to take care of that machine, and we have the third heart patient on the floor now that had one of those and is just waiting for their own heart. You have to learn how to troubleshoot for the machine and that is very interesting. Because of the changes in the care delivery system, many nurses were challenged with effectively negotiating practice through others. This participant shared her experience with working with the new NAs on the unit: We are getting more aides to work with us. It is a challenge to get them to follow through, and we have to keep checking up on them right now. Just getting them to work together as a team is a bit of a challenge right now. But it works out. Some days it's not bad; there is just the occasional day when you feel like you have just been hither and yon, but that is normal. In summary, individual nurse behaviors related to remaining in the hospital to practice nursing exist and 79 are constructed within the context of change, instability, and challenge. The change, while constant, varies in terms of its nature and relevance to nurses who perceive the change. Instability in the workplace is enhanced by increased patient acuities, decreasing lengths of hospital stay, and greater emphasis on tighter budgetary control. Nurse behaviors are, therefore, influenced by an environment that is unpredictable and has the potential for sudden and rapid alterations in patient and work conditions. Likewise, nurse behaviors also are influenced by the physical, emotional, and cognitive challenges of the workplace. The personal mosaics shared by the nurses in this study reflect an effective negotiation of practice within this changing work world and provide the basis for the sUbstantive theory that emerged. 80 Theoretical concepts Two interrelated categories emerged from the data through constant comparative analysis. These categories emerged from the personal experiences of the study participants and, thus, are "grounded" in the study data. These categories (negotiating boundaries and positioning), when examined in relationship to one another, led to the discovery of a core variable or core category (customizing practice). Together these integrated categories and their properties provide the conceptual elements of the sUbstantive theory, "Customizing Practice: Constructing a Practice Milieu," which seems to describe appropriately and accurately the process used by RNs to remain in the hospital to practice nursing (see the Figure). customizing practice customizing practice is the core variable that emerged from the data and provided the unifying theme that described how nurses remain in the hospital to practice nursing. The core variable integrates the developing theory and describes the substance of what is occurring in the data (Glaser, 1978). The "6 C's" (causes, conditions, contexts, contingencies, covariances, and consequences), suggested by Glaser (1978), were used as guidelines for conceptualizing and delimiting the core variable. customizing practice is the process of defining, planning, and constructing a practice milieu that meets Customizing Practice: Constructing a Practice Milieu / NEGOTIATING BOUNDARIES Delimiting work involvement Delimiting performance expectations 1 CUSTOMIZING PRACTICE II POSITIONING Focusing Moving better Receiving rewards Substantive Theory, customizing Practice. 81 identified personal and professional practice needs. Customizing practice occurs over time and requires individuals to attend continually to events occurring in their personal and professional life, and it requires self-awareness, introspection, and flexibility. This concept is abstracted from two interrelated categories: (a) negotiating boundaries and (b) positioning. Negotiating Boundarie. 82 Negotiating boundaries is the process of establishing and delimiting work involvement and work performance boundaries for individual nursing practice. This concept involves recognizing, thinking about, and talking to self and others about the realities, limitations, and possibilities for nursing work within the inpatient environment; it also involves defining boundaries in relation to self and others. Negative boundaries require that nurses explore and reconcile differing perceptions of nursing: (a) nursing as taught and (b) nursing as practiced. A discovery during this study was the tendency and desire of the participants to separate their personal and professional life arenas. The participants shared that it was important for them to have "another life," a life outside and separate from that of the workplace. Therefore, establishing limits, through the process of negotiating boundaries, was viewed as a means of controlling and owning one's life path; it is a way of taking care of the personal and professional self. The thoughts of this nurse, who had been on her unit for 3 years, reflect the thoughts of many of the nurse participants about home and work boundaries: I like being a staff nurse. I don't see myself as somebody who is so involved that every, I mean, I don't like to take my work home. I like to go home and have my life. I like to go to work and have my work life. There are nurses who are just, their life revolves around work, and there are those people at work that that's their life. I don't want it, to be one of those people. Negotiating boundaries is product and process. The 83 product is the boundaries that are established through the negotiation process. The process is the mechanism for establishing and negotiating professional practice limits. The process requires the participant to look and "relook" continually to think and "rethink" about the personal and professional boundaries that have been established and the relationship of each to the other. Boundaries are redefined as the needs of the individual change. Negotiating describes the flexibility and dynamic nature inherent in the boundary-setting process. It denotes that boundaries, once established, can be reflected upon and renegotiated based on the ever-changing personal and professional needs and capabilities of the individual. The nurse participants explored and delimited boundaries relating to physical and emotional involvement 84 in work and performance expectations for self and others. Delimiting work involvement. The physical work involvement boundary was negotiated by the nurses' working status (full- or part-time employment), working additional shifts, and scheduling work. Even though this was a process owned and controlled by the participants, it was a process influenced by financial issues; the personal desire to work (self-development, keeping license, professional development, and satisfaction); feelings of pressure to work; and nurse manager support. The nurses described varying degrees of flexibility in negotiating this boundary. Participants who were the sole or primary source of income for their household expressed less flexibility in terms of altering their work status. Their full-time work status provided a necessary income level, as well as vacation, sick time, insurance coverage, and retirement benefits. participants who were not the sole or primary source of income believed they had more flexibility in terms of "cutting back." participants understood and recognized that a change in work status (more or less) could result in personal and financial sacrifices that would affect their home and personal life. However, if the benefits associated with the change in work status outweighed the benefits of remaining in their current status, they would make that choice. Several participants had exercised this option at 85 some point during their nursing work. Participation in additional shift work was influenced by the desire and/or need for additional income, the desire to help the unit and the other nurses when the unit was "working short," a feeling of guilt associated with not working additional shifts, and personal life priorities. The guilt feelings were self-induced and "other" induced. Participants dealt with this guilt primarily through the strategies of weighing personal needs at the time of the request and then explaining. One nurse, who had been on her unit for 17 years, explained: In fact, sometimes it makes me feel really guilty when they are needing more people. It is not really a priority for me. I love it when I am there, but it is not a priority for me; and when they talk about the levels of nursing, I don't care about it. I.don't care. Explaining was used as a means of justifying their refusal and of trying to make others understand their need to rest, to have time off, and to take care of themselves or their family at that particular time. Work involvement also included participation in nondirect patient care activities on the unit. The nondirect care activities included staff meetings, committee involvement, in-service and continuing education, and participation in other learning or teaching events available to the staff. How much, how often, and the type of activity involvement were influenced by the nurse's personal and professional priorities in life, shift work, practice focus, and perceived impact of the activity on her practice environment. One nurse stated: Well, you know, the nurses are always wanting you to get involved and so on and I just, you know, really haven't had the time nor that much interest, which I think is probably part of nursing's problem that people don't • • • you know, when you have a family and you only have so much time, and my family is more important to me. Another nurse, who had been on her unit for 15 years, stated: It's kind of hard working nights and being involved. Well, for many years before shared governance came along, I was on the retention and recruitment committee, which I really enjoyed; and a lot of the things that we talked about and proposed they actually followed up and did something about. I used to come in you know. I'd go home and sleep a while and then come in, and I, I didn't mind because I thought it was important. But the shared governance things, I'm not too much involved in. Through the years I have been, but right now I'm tired. Several nurses suggested that "being involved" in nondirect care activities provided a means for "having a say" in the way their practice environment was constructed. One nurse, who had been on her unit for 8 years, explained: I feel like I have a lot of say on the floor and I like that. I mean, I'm on a lot of committees. I'm um, just involved in a lot things that I really enjoy, and I feel like of what I say is taken for what it's worth. like that part. Another nurse shared a similar perception: of most I I get involved in different aspects like, you know, work redesign and help implementing that. It just, it made me feel like I had some say in 86 the decisions that were being made and not just management telling me what to do, and I could see different avenues. Even before shared governance, there were other committees and things you could get involved with. You know teaching, doing um, maybe being a skills instructor, working with the hospital education department, and helping them; so, there were things that you could do to get involved, and I've always liked to do that. The physical integration of the personal and 87 professional life is symbolized by "scheduling work." Scheduling work dictated the days and the shifts that the nurses must physically be present in the work environment. Scheduling work was influenced by structural constraints of the work setting, as well as personal preference of the staff. unit and patient care requirements dictated that all shifts be "covered" with the appropriate mix and number of personnel, parameters that were established by nursing management. Work scheduling also was influenced by seniority. Nurses with longer tenure on the unit were assigned fewer numbers of weekend shifts and received preference in terms of shift assignment. The process and degree of input into scheduling varied according to the unit. Scheduling work was conducted through self-scheduling, a scheduling committee, or the nurse manager or designee. Special days off and vacation time had to be requested by the nurse. The requests were honored, if possible. If scheduling conflicts occurred, they were resolved between individual nurses, by the staffing committee, or between the nurse 88 and manager. As described by the participants, scheduling work generally resulted from a compromise between the predetermined staffing needs of the unit and the needs of the staff nurse. One nurse explained: When I started, I only wanted to do 2 days a week and that was fine with them, and I only wanted to do evenings and that was okay, and they've always been real good about pretty much letting me work what I wanted to work within their requirements. Scheduling work was necessary in order to plan for activities in both life arenas. Unplanned change or unpredictability in the work schedule made living their "other life" difficult. One nurse, who had been on her unit for 13 years, explained: But, you know, it's sort of like, I consider it my schedule, but um, the hospital really doesn't, and it can change at any time depending on staff needs. For the most part, I consider it my schedule. It is my schedule, it is my daughter's schedule, it is my husband's schedule, my mother-in-Iaw's and my father-in­law's schedule. It affects all of us. We can all work around it. We can plan, make appointments. It is a very significant part. Another nurse, with 2 years on her unit, added: But, I really think it's nice to know what your schedule is going to be month-to-month, and it doesn't change you know; it's easier to have your other life, you know-your life outside the hospital. It really makes that work a lot smoother knowing when I'm going to work, and I think most people feel that way. They like having a set schedule. You can make long-term plans around that because you know what days you have to work. You know what days you don't have to work. If the kids are out of school one day, you know you have to make changes 'cause, down in March, my kids are out of school; and now I know what day I need to get day care even though our schedules are not out because I'm pretty sure my schedule is not going to change because it hasn't in a year and a half. Another nurse explained: I work 12-hour shifts and right now 5 out of 6 days are 12-hour shifts and then I have 4 or 5 days off so that works out well. I can do whatever I want. I work 3 days in a row, have 1 day off, then work 2 in a row, then I have 4 or 5 days off. Work 1 then have 3 days off, so it works out nicely that way. I enjoy the shifts because of what it leaves me in my personal life to do. 89 When the work schedule was finalized, changes could be made only with the consent of the individual nurse. In addition to being perceived as a way of respecting and recognizing her individuality, this "rule" allowed the nurse to retain control and decision-making authority over her schedule. The nurse had a choice in changing or not changing the schedule based on her perceived needs at a given time. Work scheduling allowed the nurse to plan and prepare for personal and professional life. Transitioning was a strategy used by the nurses as they integrated their personal and professional work life. Transitioning was an individual process that occurred as the participants moved back and forth between their home and work; it involved "preparing for work" and "debriefing." Transitioning occurred across tenure and work-status levels; it changed over time. Preparing for work occurred as nurses prepared for 90 "going" to work. It involved psychologically leaving the personal life behind and moving into the realm of the professional work arena. Preparing for work was a calming influence for some, whereas it produced more anxiety for others. Some of the mechanisms described in the process of preparing for work included having quiet time, centering, questioning, wondering, worrying, and going over things. One nurse, with 4~ years of experience, explained: I'm driving down the road to go to work you know, what's going to happen there today? I • • . like I've worked there 4~ years, and it's a cardiac floor, and everything and • . • but there's never been a code while I've been there and that bothers me and so I • . • I have stuff I read and go over lots of times just to go over things so that I have it in my head and that. Another nurse, with 24 years of experience, explained: You know, you think of things that, you worry about what's going to happen, what could go wrong. Things like that. I still do that a little bit. It's not as bad as it used to be years ago. Another nurse who had been on her unit for almost 13 years described her experience with preparing for work: I like having, okay, these are my work days, that is all of it. That is my work day. I don't let anything else interfere with that. I even kind of, not even consciously prerest, "pre" some kind of quiet time to prepare. See I need that. We are, all of us, have pieces of our lives. You know, we are not nurses all the time. We are not mothers all the time. Debriefing was a learned, conscious process used by the participants as they attempted to leave the worries, concerns, and joys of the work life behind. Debriefing required skill, time, and effort in order to be used effectively. Debriefing varied along the dimension of degree of success. This nurse described the effect of time on the debriefing process: I tried to, to when I went home from work, I would try to cut work off and leave it at work, but sometimes I would find myself dreaming about it at night. I started to grind my teeth you know, so it was just kind of living through the first year. Another RN shared her experience: I leave my problems with work at work, and the 20 minutes from work to home is a kind of release time. I try not to bring it home. It had to be a conscious decision that I was not going to take my work home. I don't like to bring it home. It is something that needs to stay at the hospital. Another RN explained: I never take any of this place home with me. It stays here because I can't. I started this little thing several years ago where I take the same road home on 1-15 and, when I get to a certain sign, after that point, I trained myself not to think about the hospital. And then on the way in, when I get to that same point, I tell myself, okay, now you can start thinking about work again. But I've trained myself that I don't take these things; you know sometimes I will tell my husband something interesting that happened, but I don't worry about it. I won't take the patients home with me. I leave them at that point. I have trained myself that work is literally behind me. 91 Debriefing was necessary because the demands of work life interfered with the ability of the participants to have free time at home without thinking about work and to concentrate on the activities of their home life. One nurse shared her experience: sometimes I will do things to take care of myself before I go home because it's real hard to be a caregiver here and then go home to be a caregiver. sometimes I'll qo get a drink. Sometimes, if it's nice weather, I'll go to the park and read; um, just varies what I might do. When there's those times that I can't stop and take care of myself, then I usually try and do it internally and talk to me inside my brain. The participants also described negotiating an emotional component to work involvement. The emotional 92 boundaries change and are renegotiated as the nurses qain nursing experience and knowledge and as they beqin to recognize and understand the emotional demands of nursing practice. Emotionally the nurses attempted to "stay in touch" with the patients while maintaining a certain emotional boundary that allowed for a therapeutic and helping relationship with the patient and family. One participant, with 24 years on her unit, shared her experience: When I first started out, I had a difficult time because I got too involved personally. I took on their life as part of mine and, um, as I progressed, I was able to not get quite so involved in their lives as I had been. Another RN, with 2~ years on her unit, explained: It has been hard to learn to adjust to dealing with terminal and uh, difficult illness. It's hard for me to differentiate. • •. I feel like I have lost about 15 friends in the last 2 years and I don't know, we really don't have anything to help us work through that. Another nurse shared her feelings about her emotional work boundaries: I don't go to funerals, UDl, of patients, even though I have been very close to them and their families because that would be too much for me to handle. We've had some patients for a long while. Ub, their death has been a welcome release but I still, I can't. . •• I just make it a rule for ME that I leave them at the hospital and I don't go to the funeral. I don't visit them at home, and it isn't that I don't like them, its's just that some things are hard to deal with, so I just leave them. Leave them here. Negotiating and practicing within these boundaries allowed nurses to bring with them to practice those 93 elements necessary to "be there" in the practice setting physically, emotionally, spiritually, and cognitively for the patients, for themselves, and for others. Being there required energy, enthusia