Despite the shortage of nurses in

The Relationships Between Nurses Perceptions of the Hemodialysis Unit Work Environment and Nurse Turnover, Patient Satisfaction, and Hospitalizations Jane K. Gardner Charlotte Thomas-Hawkins Louis Fogg Carolyn E. Latham Continuing Nursing Education Despite the shortage of nurses in nephrology settings, it is anticipated that the demand for nursing care will increase as the population of individuals with Stage 5 chronic kidney disease (CKD) who require dialysis continues to grow at an annual rate of approximately 3% (United States Renal Data System [USRDS], 2006). Research in hospitals has indicated that one factor that contributes to the nursing shortage is nurses negative perceptions of the work environments. Moreover, research has also shown that hospital work environments with organizational attributes that support professional nursing practice are important Jane K. Gardner, DNP, RN, was Director of Operations and Nurse Development, Renal Care Group, Nashville, TN at the time of this study. She is a member of ANNA s Windy City Chapter. For more information on this article, contact her at jane_gardner@comcast.net Charlotte Thomas-Hawkins, PhD, RN, is Assistant Professor, College of Nursing, Rutgers, The State University of New Jersey, Newark, NJ. She is a member of ANNA s Garden State Chapter. Louis Fogg, PhD, RN, is Assistant Professor, College of Nursing, Rush University, Chicago, IL. Carolyn E. Latham, MSN, MBA, RN, CNN, was Senior Vice President, Clinical Operations, Renal Care Group, Nashville, TN at the time of this study. She is a member of ANNA s Music City Chapter and a Past President of ANNA. Acknowledgment: The authors would like to thank the Registered Nurses who participated in this research. Dr. Gardner would also like to acknowledge Dr. Raymond Hakim for his support of this research. Note: The research was funded by Renal Care Group and was done in partial fulfillment of the requirements for the Doctorate of Nursing Practice Degree awarded to Jane K. Gardner on August 28, 2004 from Rush University, Chicago, IL. The authors reported no actual or potential conflict of interest in relation to this continuing nursing education article. While the nephrology nursing shortage persists despite the continued growth of the population of individuals with Stage 5 chronic kidney disease, there is a paucity of empirical data regarding nephrology nurses perceptions of their work environments. Moreover, there are no studies that have examined the relationship of work environment attributes to patient and nurse outcomes in dialysis settings. The purpose of this study was to examine the relationships between staff nurses perceptions of dialysis work environments, nurses intentions to leave their current jobs, nurse turnover, patient satisfaction, and patient hospitalization rates. A descriptive, correlational design was used. Nurse level and facility level data were obtained. The sample for nurse-level data consisted of 199 registered nurses in staff nurse roles in 56 dialysis facilities of a national dialysis company. The sample for facility-level analysis consisted of 46 dialysis facilities, and nurse-level data were aggregated for facility-level analysis. The Practice Environment Scale-Nursing Work Index (PES-NWI) was used to measure nurses perceptions of the dialysis work environment. Nurses intention to leave their jobs and facility-level turnover rates were the nurse outcomes examined in this study. Facility- level patient satisfaction and hospitalization rates were the patient outcomes examined. Correlation coefficients were computed to measure the relationships between study variables, and independent t-tests were performed to examine subgroup differences in work environment perceptions. Overall, nurses rated the work environment somewhat favorably. Nurses who expressed intention to leave their jobs rated the work environment more negatively compared to nurses who intended to stay. Significant correlations were found between nurses perceptions of the dialysis work environment, nurses intention to leave their jobs, nurse turnover rates, and patient hospitalizations. Study findings suggest that nurses perceptions of the dialysis work environment are important for nurse and patient outcomes in dialysis settings. Further research is needed to explore the predictive ability of the work environment for nurse and patient outcomes in hemodialysis units. Goal Discuss the results of this study examining the importance of nurses perceptions of the work environment and nurse and patient outcomes. Objectives 1. List the magnet characteristics defining attributes of the work environment that promote professional nursing practice. 2. Describe the relationship between the work environment and nurse turnover and intention to leave their jobs. 3. Explain the impact of dialysis work environment on patient satisfaction and hospitalization. This offering for 1.5 contact hours is being provided by the American Nephrology Nurses Association (ANNA). ANNA is accredited as a provider of continuing nursing education (CNE) by the American Nurses Credentialing Center s Commission on Accreditation. ANNA is a provider approved by the California Board of Registered Nursing, provider number CEP 00910. This CNE article may be applied to the required recertification contact hours in nephrology nursing. NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3 271

The Relationships Between Nurses Perceptions of the Hemodialysis Unit Work Environment and Nurse Turnover, Patient Satisfaction, and Hospitalizations predictors of positive nurse and patient outcomes. However, there is a paucity of empirical data regarding nephrology nurses perceptions of their work environment, and there are no studies that have examined the relationship of work environment attributes to patient and nurse outcomes in dialysis settings. Research aimed at exploring nephrology nurses perceptions of their work environment and examining the relationship of their perceptions to patient and nurse outcomes is important for developing and testing strategies to address modifiable factors that contribute to the shortage of nurses and negative nurse and patient outcomes in nephrology settings. Nurses perceptions of the work environment in hospitals and the relationship of their perceptions to patient and nurse outcomes have been popular topics of research inquiry since the early 1980s. In 1983, research conducted by the American Academy of Nursing found that hospitals that did not experience problems with recruitment and retention of nurses shared a common set of work environment characteristics that promoted a professional practice environment for nurses. (McClure, Poulin, Sovie, & Wandelt, 1983). These work environment characteristics were referred to as magnet attributes because of their ability to attract and retain nurses. Examples of magnet attributes in hospitals include nursing involvement in decentralized decision making at the highest levels of hospital management; strong unit level nurse leadership that involves staff nurses in decision making; adequate staffing and resources; collaborative nurse-physician relations; and nursing foundations for quality patient care. Today, magnet characteristics endure as defining attributes of work environments that promote professional nursing practice as well as positive patient and nurse outcomes. The empirical literature indicates that magnet attributes are important predictors of positive patient outcomes in hospitals such as higher lev- els of quality patient care, lower mortality rates, lower rates of adverse events, and higher rates of satisfaction with care (Aiken, Clarke, Sloane, Sochalski, & Silber, 2002; Aiken & Sloane, 1997; Aiken, Smith, & Lake, 1994; Clarke, Sloane, & Aiken, 2002; Flynn & Aiken, 2002; Laschinger, Almost, & Tuer-Hodes, 2003; O Brien-Pallas, et. al., 2004; Tigert & Laschinger, 2004). For example, one study compared mortality rates in magnet hospitals to mortality rates in non-magnet hospitals that were matched on non-nursing organizational characteristics (Aiken et al., 1994). Findings from this study revealed a 7.7% lower mortality rate and higher levels of patient satisfaction in the magnet hospitals compared to the non-magnet hospitals. Similarly, research has shown that patients experience greater levels of satisfaction with nursing care on inpatient nursing units where magnet organizational attributes enable professional nursing care compared to satisfaction with nursing care on units that lack the presence of these organizational attributes (Aiken, Sloane, & Lake, 1996). At the dialysis facility level, many patient outcomes are routinely measured and monitored. For example, patient hospitalization rates are measured regularly. The Center for Medicare and Medicaid Services (CMS) requires that dialysis facilities report hospitalization and mortality data to them 90 days after a patient begins dialysis treatment. Moreover, since there is evidence that patients receiving dialysis treatments for less than 90 days are at increased risk for poor outcomes (e.g., hospitalization and mortality) (Wingard, Pupim, Ikizler, Thomas, & Hakim, 2005), many dialysis providers monitor patient hospitalization rates in the first 90 days of treatment as well. Patient hospitalizations are considered important indicators of the general health status of patients with Stage 5 CKD who are on dialysis and may be considered a reasonable nurse-sensitive quality indicator in dialysis settings. Patient satisfaction with care is another patient outcome that is routinely measured in outpatient dialysis facilities. The American Nurses Association (ANA) regards patient satisfaction as a nurse-sensitive quality indicator in hospitals (ANA, 1999), and it is likely that patient satisfaction also is an outcome that is sensitive to nursing care in dialysis settings. While the relationship between magnet attributes and outcomes for patients with Stage 5 CKD has not been examined, the strength of empirical findings for important relationships between magnet attributes and patient outcomes in hospitals underscores the need to determine these relationships in dialysis settings. Research also continues to elucidate the importance of magnet attributes for positive nurse outcomes in hospital settings such as higher levels of job satisfaction, lower levels of burnout, and lower rates of workrelated disability (Aiken et al., 2002; Aiken & Sloane, 1997; Clarke et al., 2002; Flynn & Aiken, 2002; Heath, Johanson, & Blake, 2004; Laschinger et al., 2003; O Brien-Pallas et. al., 2004; Tigert & Laschinger, 2004). For example, in a focus group study that explored nurses perceptions of healthy and unhealthy work environments, focus group participants overwhelmingly described magnet attributes as essential solutions for unhealthy work environments (Heath et al., 2004). Since positive patient and nurse outcomes are features of magnet hospitals, it is important to examine the relationship of work environment attributes on patient and nurse outcomes in nephrology settings. Organizational attributes that support professional nursing practice have been linked to important nurse outcomes in hospitals (Aiken et al., 2002; Aiken & Sloane, 1997; Sochalski, 2002). For example nurses working in magnet hospital environments have reported significantly higher levels of job satisfaction (Aiken et al., 2002), lower levels of emotional exhaustion and burnout (Aiken et al., 2002; Aiken & Sloane, 1997), and fewer nurse nurse-reported needle 272 NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3

stick injuries (Aiken, Sloane, & Klocinski, 1997) compared to nurses working in non-magnet hospital environments. There is a body of empirical support that the work environment is an important factor that contributes to hospital nurses intentions to leave their jobs and nurse turnover in hemodialysis facilities in hospitals (Aiken et al., 2002; Larrabee et al., 2003; Shader, Broome, Broome, West, & Nash, 2001; Strachota, Normadin, O Brien, Clary, & Krukow, 2003). While these relationships have not been explored in hemodialysis settings, it is intuitive that they exist in dialysis work environments. In fact, in a study that explored nurses perceptions of hemodialysis work environments (Thomas-Hawkins, Denno, Currier, & Wick, 2003), nearly 20% of nurses reported they intended to leave their jobs in the next year. Equally as important, nurses who intended to leave their jobs rated the presence of magnet attributes in their dialysis work environments as significantly lower compared to nurses who did not plan to leave. The retention of nurses is critical for the provision of quality nursing care in dialysis facilities. Research funded by The Agency for Healthcare Research and Quality (AHRQ) that examined nurse staffing levels and adverse occurrences in hospitals indicates that lower nurseto-patient ratios were associated with higher rates of nonfatal patient outcomes including nosocomial infections, pressure ulcers, and patient falls (Hickman, Severance, Feldstein, 2003). These findings emphasize the importance of determining relationships between dialysis work environments and nurse outcomes in these facilities. There is no research that has examined relationships between magnet attributes and patient and nurse outcomes in dialysis settings. However, the seminal study (Thomas- Hawkins et al., 2003) that examined 383 staff nurses perceptions of the extent to which magnet attributes were present in their freestanding hemodialysis facility work environ- ments, as measured by the Revised Nursing Work Index (NWI-R) (Aiken & Patrician, 2000), provides some insight into nephrology nurses perceptions of the dialysis work environment. Findings from this study indicate that staff nurses who work in freestanding hemodialysis units identify several notable features of magnet hospitals in their work environments such as the expectation by administration for high standards of nursing care and collegial nurse-physician relationships. On the other hand, a majority of nurses in this study disagreed that many attributes of magnet hospitals are present in their work environments. For example, survey responses from the study by Thomas- Hawkins et al. (2003) study indicated that nurses in dialysis units feel that they are not involved in decision making at the organizational level, that administration does not listen and respond to employee concerns, and that there are inadequate staffing and resources for patient care. Furthermore, nurses who intended to leave their jobs in the next year had significantly lower mean ratings of the environment (i.e., higher level of disagreement that magnet characteristics were present) compared to nurses who did not plan to leave their jobs. Based on these findings, Thomas- Hawkins and colleagues recommended further work in this area and indicated an urgent need to examine the relationship between nephrology nurses perceptions of their work environments and patient and nurse outcomes. Survey Questions This study was conducted to examine the relationships between nurses perceptions of their work environments, nurses intentions to leave their jobs, nurse turnover, patient satisfaction with care, and patient hospitalization. Nurse-level and facility-level data were used to address the following research questions: 1. To what extent do nurses rate the presence of magnet attributes in their hemodialysis work environment? 2. What percentage of nurses affirm the importance of magnet attributes for their work environments? 3. What is the relationship between nurses ratings of their work environments and nurse outcomes (i.e., intention to leave their job and nurse turnover)? 4. Are there differences in work environment ratings between nurses who intend to leave their job and those who do not? 5. What is the relationship between nurses ratings of their work environment and patient outcomes (i.e., patient satisfaction with care and patient hospitalizations)? Findings from this study can help to identify potentially modifiable factors in the work environment that may be significantly related to negative nurse and patient outcomes. Methodology Sample Hemodialysis facility managers in a national dialysis company were contacted by email and asked to extend an invitation to all full-time and part-time registered staff nurses to participate in a voluntary survey. The principal investigator (PI) visited 56 of the dialysis company s facilities within a 3-month time period. Survey questionnaires were coded to allow aggregation of nurse-level data to each participant s facility. The PI explained the purpose of the study to participants. They were informed that their participation was voluntary and confidential. The participants completion of the questionnaire served as their willingness to participate. Participants completed the questionnaires in their dialysis facility conference room with the researcher present. Two-hundred and ten nurses completed the survey; 11 of the surveys had incomplete data and were NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3 273

The Relationships Between Nurses Perceptions of the Hemodialysis Unit Work Environment and Nurse Turnover, Patient Satisfaction, and Hospitalizations not included in the analysis. The final sample for nurse-level analysis consisted of 199 staff nurses in 56 dialysis company facilities. At least 50% of staff nurses in each dialysis company facility completed the questionnaire, with a range of two to ten nurses per facility. The sample for facility-level data included the 46 dialysis company facilities for which nurse turnover rates, patient satisfaction ratings, and patient hospitalizations were made available to the PI. Data from nurses in each of the 46 facilities were aggregated to obtain facility-level scores. Sample demographics are presented in Table 1. The Institutional Review Boards of Rush University Medical Center, Chicago, IL, and the national dialysis company approved the study. Survey Instruments Demographic data were collected using a demographic questionnaire developed for this study. Items included on the questionnaire were age, gender, years in nursing, years in nephrology nursing, years in current position, highest nursing degree and non-nursing degree completed, specialty certifications in nephrology nursing, and membership in the American Nephrology Nurses Association (yes/no). The Practice Environment Scale Nursing Work Index (PES-NWI) (Lake, 2002) was used to assess participants ratings of the presence of magnet attributes in their current job and the importance of each item to participants. The PES-NWI is a 31- item instrument, derived from the Nursing Work Index (Aiken & Patrician, 2000), that measures the extent to which nurses rate the presence of magnet attributes in their current work environment. Nurses are asked to rate each item on a scale of 1 (strongly agree) to 4 (strongly disagree). PES-NWI items are reverse coded before data analysis. PES-NWI items are arranged in 5 subscales that provide a profile of key domains in the nursing practice environment of the original magnet hospitals. The items in the Nurse Table 1 Sample Characteristics (n = 199) Mean (SD) Age 42.8 (9.4) Gender n (%) Female Male 174 (86.6) 25 (12.4) Education & Certification n (%) BSN BS (non-nursing) AD Diploma Certified Nephrology Nurse (CNN) 88 (44.2) 1 (0.5) 68 (34.2) 42 (21.1) 35 (17.5) Outcomes n/n (%) Intention to leave job 16/169 (9.5) Nurse turnover Year prior to survey First quarter of survey year Number of hospitalizations/facility Patients on dialysis < 90 days Patient on dialysis > 90 days Patient satisfaction Overall care Nursing care Years in nursing Less than 1 year 1-4 years 5-9 years 10-14 years 15-19 years 20-24 years 25+ years Years in nephrology nursing Less than 1 year 1-4 years 5-9 years 10-14 years 15-19 years 20-24 years 25+ years Years in current position Less than 1 year 1-4 years 5-9 years Mean (SD) 33.5 (38.1) 9.0 (14.5) Mean (range) 19.9 (4.6-36.9) 12.7 (4.6-20.4) Mean (SD) 48.0% (13.9%) 95.8% (2.9%) 1.5% 7.5% 18.0% 23.0% 8.0% 20.5% 21.5% 7.5% 31.0% 17.0% 17.0% 12.5% 6.5% 8.5% 22.0% 49.0% 28.5% 274 NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3

Participation in Hospital Affairs subscale reflect a participatory role and valued status of nurses in a broad organizational context (e.g., internal governance, decision making). Similar to Thomas-Hawkins et al. (2003), this subscale was referred to as the Nurse Participation in Dialysis Provider Affairs in this study to reflect the dialysis provider organization s support of a participatory role and valued status of nurses. Items in the Nursing Foundations for Quality of Care subscale emphasize a high standard of patient care. Together, these two subscales reflect nurses perceptions of the presence of magnet attributes in the broader organizational environment. The Nurse Manager Ability, Leadership, and Support of Nurses subscale focuses on the critical role of the nurse manager. Items in the Staffing and Resource Adequacy subscale describe having adequate staff and support resources to provide quality patient care, and the Collegial Nurse-Physician Relations subscale reflects positive working relationships between nurses and physicians. These three subscales reflect nurses perceptions of the presence of magnet attributes in the unit environment. Subscale scores are calculated as mean scores, with a range of 1 to 4. A total PES-NWI score is calculated as a mean of subscale scores. PES-NWI subscale scores below 2.5 represent disagreement and scores above 2.5 represent agreement that the subscale items are present in the current work environment. Alpha reliabilities for PES-NWI subscales have been reported at 0.71-0.82 in non-nephrology nurse samples (Lake, 2002). Thomas-Hawkins and colleagues (2003) reported PES-NWI subscale alpha reliabilities of 0.81 0.87 in their nephrology nurse sample. For this study, alpha reliabilities ranged from 0.85 0.96. The predictive ability of the PES-NWI has been established in hospital and dialysis settings. PES-NWI scores predicted needlestick injuries in a sample of nurses in a hospital setting (Clark et al., 2002). PES-NWI scores also predicted nurses intentions to leave their job in a sample of nurses who worked in outpatient hemodialysis settings (Thomas- Hawkins et al., 2003). Moreover, the National Quality Forum, a private notfor-profit group of public and private health care organizations affirmed the utility of the PES-NWI as a useful measure of the nurse practice environment, and they selected the PES-NWI as a nursing care performance measure in their voluntary set of National Voluntary Consensus Standards for Nursing Sensitive Care (National Quality Forum, 2003). Nurse turnover rates for each facility were provided to the PI by the Human Resource department of the dialysis company for the 3-month data collection of the first quarter of 2004 and a total year turnover rate for 2003. The facilities calculated nurse turnover rates by dividing the number of nurse terminations by the number of average actively employed nurses within a year. Intention to leave the job was measured with a single item, Do you plan to leave your job in the next year? Nurses reported their intent to leave by responding either yes or no to this item. Facility-level patient satisfaction with care data in the dialysis facilities for the first quarter of 2004 were obtained by the PI from the medical office of the dialysis company. The dialysis facility patient satisfaction survey is comprised of 35 items that are arranged into 7 scales: nurses, technicians, dieticians, social workers, nephrologists, the dialysis clinic, and patient education. Patients are asked to rate their satisfaction with items in the 7 scale domains on a 5-point Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree). Two facility-level patient satisfaction scores were used for this study: the percent of patients who reported satisfaction with overall care at their dialysis facility and the percent of patients who were satisfied with nursing care. The patient satisfaction survey was developed and tested by Data Management & Research (DMR) (DMR, 2001). Evaluation of the psychometric properties of the patient satisfaction survey was performed on a sample of 13,616 patients on hemodialysis. Alpha reliability of the dialysis company patient satisfaction survey is acceptable with scale reliability coefficients exceeding.80. Psychometric testing indicates the instrument has good convergent and discriminant validity, and a factor analysis supports the instrument s scale structure. The number of patient hospitalizations per facility-level for patients on dialysis less than 90 days for the first quarter of the survey year, for patients on dialysis greater than 90 days for the same time period, and a total number of hospitalizations for the year prior to the survey were obtained by the PI from the dialysis company medical office. Data Analysis Data were analyzed using the Statistical Package for the Social Sciences (SPSS). To examine nurses ratings of the presence and importance of magnet attributes in their current job, nurse-level data were used. Frequency distributions and mean scores for PES-NWI items, subscales, and the total instrument were computed and examined. To determine PES- NWI mean differences, if any, between nurses who expressed intention to leave and those who did not, independent t-tests were performed. To examine relationships between PES-NWI scores and patient and nurse outcomes, facility-level data were used. PES-NWI subscale and total mean scores for nurses who worked in the same facility were aggregated to obtain PES-NWI scores for each facility, and bivariate correlation analyses were performed. Results Nurse Ratings Nurse ratings of the presence and importance of magnet attributes in their current job. Mean PES-NWI subscale scores ranged from 2.80 to 3.22 and the PES-NWI total score was 3.09 (see Table 2), indicating that there was overall agree- NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3 275

The Relationships Between Nurses Perceptions of the Hemodialysis Unit Work Environment and Nurse Turnover, Patient Satisfaction, and Hospitalizations Table 2 Nurse Ratings for PES-NWI Subscales and Items Subscale and Items PES-NWI Mean Score % Agreeing Item is Present % Affirming Item is Important Nurse Participation in Dialysis Provider Affairs 2.87 68 90 1. Involvement in internal governance 2.77 66 95 2. Participation in policy decisions 2.55 59 95 3. Opportunities for advancement 2.74 62 86 4. Administration listens & responds 2.78 75 99 5. Highly visible/accessible nurse executive officer 2.89 66 90 6. Career development opportunities 2.79 64 93 7. Nursing supervisors consult with staff 3.26 80 97 8. Serve on special committees 3.07 81 81 9. Nurse executive officer equal in power & authority to other organization 2.75 61 82 senior leadership Nursing Foundations for Quality of Care 3.26 83 92 1. Use of nursing diagnosis 3.04 74 78 2. Active quality assurance program 3.36 88 98 3. Preceptor program for new RN hires 3.23 78 99 4. Nursing care based on nursing model 3.05 76 91 5. Patient care assignments promote continuity 3.11 76 78 6. Philosophy of nursing pervades environment 3.17 81 94 7. Up-to-date nursing care plans 3.33 83 79 8. High standards of nursing care expected 3.58 94 99 9. Active continuing education program for nurses 3.41 84 99 10. Working with nurses who are clinically competent 3.53 90 100 Staffing and Resource Adequacy 2.80 62 99 1. Enough staff to get work done safely 2.76 59 100 2. Enough RNs to provide quality care 2.91 67 100 3. Adequate support services to allow me time to spend with patients 2.67 57 99 4. Enough time to discuss patient care problems 2.87 67 99 Manager Ability, Leadership and Support of Nurses 3.28 82 99 1. Facility manager who is good manager and leader 3.38 83 100 2. Management backs up nursing staff 3.29 80 100 3. Management uses mistakes as learning opportunities not criticism 3.25 84 99 4. Supervisory staff supportive of nurses 3.41 84 100 5. Praise & recognition for job well done 3.18 79 96 Collegial Nurse-Physician Relations 3.22 83 96 1. Teamwork between RNs and MDs 3.21 82 100 2. RNs and MDs have good relationships 3.38 89 98 3.Collaboration between RNs and MDs 3.14 79 97 Total PES-NWI Score 3.09 276 NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3

Table 3 Subscale Comparison with Study by Thomas-Hawkins et al. (2003) Variable Name 2003 Mean * 2004 Mean * t-stat p-value Subscale 1: Nurse participation 2.25 2.87 (.05) 12.4 0.001 Subscale 2: Quality Care 2.73 3.26 (.04) 13.2 0.001 Subscale 3: Manager ability 2.62 3.28 (.06) 11.0 0.001 Subscale 4: Staffing & resource 2.28 2.80 (.07) 7.3 0.001 Subscale 5: RN/MD relations 2.90 3.22 (.05) 3.22 (.05) 0.001 * 1 = Strongly disagree that item is present; 4 = Strongly agree that item is present ment among participants that PES- NWI subscale items were present in their work environments. The mean score for Nurses Participation in Dialysis Provider Affairs was 2.87, indicating that nurses agreed that their work environment was somewhat supportive of a participatory role and valued status for nurses. The mean score for Nursing Foundations for Quality of Care was 3.26 indicating that the nurses agreed that nursing foundations for a high standard of patient care were present in their work environments. The mean score for Manager Ability, Leadership, and Support of Nurses was 3.28, indicating that the nurses agreed their manager was a good leader and supported nurses. The mean score for Staffing and Resource Adequacy was 2.80 indicating that nurses agreed, somewhat, that staffing and resources for patient care were adequate. Finally, the mean score for Collegial Nurse- Physician Relations was 3.22, indicating that the nurses agreed their relations with physicians were collaborative. Nurses were also asked whether each PES-NWI item was important to them and their job (see Table 2). Affirmation of the importance nurses attributed to a magnet attribute was assessed by the percentage of nurses who answered yes to the question Is this important to you and your job? for each PES-NWI item. For the 31 PES-NWI items, 23 items were affirmed as important by more than 90% of the nurses, and 29 of the items were affirmed as important by 80% of the nurses (see Table 2). These findings indicate that staff nurses in dialysis environments affirm the importance of magnet characteristics in their work environments. Relationship of PES-NWI Ratings and Outcomes Intention to leave. Nearly 10% of nurses who responded to the question Do you plan to leave your job in the next year indicated an intention to leave. PES-NWI total scores were significantly related to nurses intentions to leave (r = -.254, p less than 0.01); that is, lower overall ratings of the work environment (i.e., disagreement that magnet characteristics were present) were significantly related to nurses intentions to leave. Moreover, lower ratings on the Nurses Participation in Dialysis Provider Affairs subscale (r =.269, p less than 0.01), Nurse Manager Ability, Leadership and Support of Nurses subscale (r =.346, p less than 0.01), and Staffing and Resource Adequacy subscale (r =.219, p less than 0.01) were significantly related to nurses intentions to leave their jobs. Independent t-tests were computed to compare differences in PES-NWI mean scores between nurses who intended to leave their jobs and those who did not (see Figure 1). Nurses who planned to leave their job in the next year reported significantly lower mean scores, that is a higher level of disagreement, on the Nurse Participation in Dialysis Provider Affairs subscale (t = -3.60, p less than.05), the Nursing Foundations for Quality Care (t = -1.96, p =.05), the Nurse Manager Ability, Leadership, and Support of Nurses subscale (t = -3.30, p less than.05), and the Staffing and Resource Adequacy subscale (t = -2.90, p less than.05) compared to nurses who did not plan to leave. The intend to leave group also reported significantly lower mean scores for the total PES-NWI (t = -3.39, p less than.05) compared to nurses who did not intend to leave. There were no significant differences between these two groups in their perceptions of nurse-physician relationships (t = -.515, p =.61). Nurse turnover. The mean nurse turnover rate (i.e., percentage of nurses who left their jobs) for the 46 facilities for the year prior to the survey was 33.5%, with a range of 0% to 150%, and the mean facility-level nurse turnover rate for first quarter of the survey year was 9% with a range of 0% to 66% (see Table 3). Mean PES-NWI scores were also examined in relation to nurse turnover in facilities. The Staffing and Resource Adequacy subscale was the only PES- NWI subscale that was significantly correlated with facility nurse turnover rates for the year prior to the survey (2003) (r =.32, p less than.05) and the first quarter of the survey year, (r =.36, p less than.05). This finding indicates that nurses who disagreed that staffing and resources in the unit were adequate for patient care are more likely to leave their job Patient hospitalizations. The mean number of hospitalizations for the 46 facilities for which data were NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3 277

The Relationships Between Nurses Perceptions of the Hemodialysis Unit Work Environment and Nurse Turnover, Patient Satisfaction, and Hospitalizations available for the year prior to the survey was 19.85, with a range of 4.6 to 36.9 for patients on dialysis less than 90 days and 12.69 with a range of 4.6 to 20.4 for patients on dialysis greater than 90 days (see Table 1). Patient hospitalization days were calculated by facilities using the number of patient risk days as the denominator in the formula based on the calculation method used by the USRDS. The use of patient risk days, measured on a daily basis, has the advantage of providing ongoing adjustment for varying numbers of patients in facilities. This measurement method is based on use of raw data rather than averages, and provides precise hospitalization data, even as patients transfer, get transplanted, change modality, or expire (USRDS, 2006). Negative overall ratings of the dialysis work environment were significantly related to hospitalizations for patients on dialysis greater than 90 days. This was true for the total PES-NWI scores (r = -.34, p less than.05), Nursing Foundations for Quality Care subscale scores (r = -.36, p less than 0.05), Staffing and Resource Adequacy subscale scores (r = -.30, p less than 0.05), and Collegial Nurse-Physician Relations subscale scores (r = -.36, p less than 0.05). In addition, PES-NWI Collegial Nurse-Physician Relations subscale scores were negatively related to the number of patient hospitalizations in a facility for patients receiving hemodialysis less than 90 days (r = -.36, p less than.05). Of interest, there also was a significant relationship between nurse turnover and patient hospitalizations; that is, as facility nurse turnover increased, the number of patient hospitalization in facilities increased as well (r =.25, p less than 0.05). Patient satisfaction. The mean percentage of patients who were satisfied with their overall care for all facilities was 48%, with a range of 25% to 94%, and the mean percentage of patients satisfied with nursing care in facilities was 96% with a range of 88% to 100%. PES-NWI scores were not significantly related to patient satisfaction scores. Nurse turnover was related to patient satisfaction as nurse turnover increased in facilities, the percentage of patients in facilities satisfied with their care decreased (r = -.53, p less than 0.01). Discussion A shortage of nephrology nurses in dialysis units persists despite a continued growth in individuals with Stage 5 CKD requiring long-term dialysis therapy. While several national organizations have acknowledged that improving nursing work environments is imperative for addressing the nursing shortage (American Hospital Association, 2002; ANA, 2002; Joint Commission on Accreditation of Healthcare Organizations, 2002; U.S. General Accounting Office, 2001), little is known of nephrology nurses perceptions of dialysis work environments and the relative importance of their perceptions for nurse and patient outcomes. The current study examined the relationship between nurses perceptions of the work environment and nurse and patient outcomes in hemodialysis facilities. Study findings indicate that nurses perceptions of the hemodialysis work environment are significantly related to their intention to leave their jobs, nurse turnover in dialysis facilities, and patient hospitalizations. In general, nurses agreed (somewhat to strongly) that attributes that support of professional nursing practice were present in their work environments. At the organizational level, nurses agreed, somewhat, that dialysis organizations value the status of nurses in the organization and supported a participatory role at the organizational level. Nurses ratings indicate relatively strong agreement that there was an emphasis on a high standard of patient care within the dialysis organization. At the dialysis unit level, both intra- and interdisciplinary relationships were viewed positively. Nurses strongly agreed that the nurse manager was a good leader and supportive of nurses. They also agreed that relationships between nurses and physicians were positive. Collegial collaboration among nurses and between nurses and other disciplines is a crucial attribute of a professional practice environment, and the positive ratings of nurse-physician relationships in dialysis settings are consistent with previous findings (Thomas- Hawkins et al., 2003). The importance of each PES-NWI item for nurses was also explored. Ninety percent of nurses affirmed the importance of 23 magnet attributes for their dialysis work environments, and 80% affirmed the importance of 29 attributes for their work settings. However, while 90% of nurses indicated that their participation in dialysis provider affairs was important, only 68%, agreed that this characteristic was present in their work environment (see Table 2). For example, 95% of nurses indicated that their participation in policy decisions was important, but only 59% agreed that they participated in policy decisions and mean rating of the presence of this characteristic in their jobs was only 2.5. These findings suggest that, while most nurses in dialysis facilities feel that it is important for them to have some level of autonomy and control over their practice at the organizational level, many nurses do not agree that dialysis organizations support or value their need for control over their practice. The findings also suggest that dialysis organizations might consider the need for infrastructures that clearly articulate and integrate practice environments that promote and support professional nursing practice, and foster decentralized decision making by staff nurses, unit selfgovernance, and participative management (instead of top-down management) that supports autonomous roles for nurses in dialysis units. Clearly, more research is needed in this area. A growing body of research has established the benefit of magnet characteristics for nurses and patients (McClure & Hinshaw, 2002). The relationship between the presence (or absence) of magnet characteristics in dialysis work environments and several nurse and patient outcomes were examined in this study. Intention to leave and nurse turnover rates were nurse outcomes examined in this study. 278 NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3

Figure 1 Mean Scores For Nurses Who Planned to Leave Job & For Nurses Who Did Not Plan to Leave Job PES-NWI Scores 4.0 3.5 3 2.5 2 1.5 1 2.66 3.17 2.33 Plan to Leave n=16 2.96 Total** Provider** Foundations Nursing Management** Plan to Stay n=153 3.42 3.31 3.26 3.17 3.03 2.90 2.57 2.19 Staffing & Resources** RN-MD Relations Total = Mean of all subscale items Provider = Subscale 1: Nurses Participation in Dialysis Provider Affairs Foundations = Subscale 2: Nursing Foundations for Quality of Care Nursing Management = Subscale 3: Manager Ability, Leadership, & Support of Nurses Staffing & Resources = Subscale 4: Staffing & Resource Adequacy RN-MD Relations = Subscale 5: Collegial Nurse-Physician Relations PES-NWI Scales 1 = Strongly disagree attribute is present 4 = Strongly agree attribute is present ** = p <.05 for associated t-test Nearly 10% of participants who responded to the question Do you plan to leave your job in the next year reported an intention to leave their jobs in the next year. While the percentage of nurses in this study who intended to leave their jobs were lower compared to other studies (Aiken et al., 2001; Thomas-Hawkins et al., 2003), a loss of 10% of nurses in dialysis units jobs can have a significant impact on staffing. It is important to note that staff nurses perceptions of the work environment were significantly related to their intention to leave their jobs. Nurses who planned to leave rated the presence of work environment characteristics that reflect dialysis organization s value and support of a participatory role for nurses and emphasis on nursing foundations for high standards of care much lower compared to nurses who did not report an intention to leave (see Figure 1). Moreover, nurses who planned to leave rated that presence of work environment characteristics that reflect manager leadership and staffing much lower compared to nurses who planned to stay. While reasons for intention to leave were not explored in this study, research has consistently demonstrated that job dissatisfaction, burnout, and supervisor management style are important factors that contribute to nurses intention to leave their jobs and nurse turnover (Aiken et al., 2002; Smith, Hood, Waldman, & Smith, 2005; Sochalski, 2002). These findings support the need for more research that explores the extent to which nurses may be dissatisfied or burned out in dialysis settings. Moreover, factors that contribute to job dissatisfaction and burnout in hemodialysis facilities should be explored and addressed by dialysis providers, and the impact of these efforts on retention should be measured and documented. The mean percentage of nurses who left their jobs in the year prior to the survey was 33.5%, and some facilities had a turnover rate as high as 150%. These data are both startling and sobering. The findings from this study suggest that staffing in dialysis units may be related to nurses decisions to leave their jobs. Nurses negative perceptions of staffing and resource adequacy in dialysis units were significantly related to facility-level nurse turnover rates. It is notable that 90% of the participants in this study affirmed the importance of staffing and resource adequacy for their jobs, yet only 62% agreed that there was adequate staffing and resources for patient care in their units. These findings indicate that staffing levels in dialysis facilities may be important predictors of nurse turnover. Nurse staffing in dialysis units is an ongoing problem, and the implications of these findings calls attention to a need for dialysis providers to employ strategies to increase the supply of new nurses and retain nurses currently present in dialysis units; to address work environment characteristics deemed important by nurses; and to consider the presence of these characteristics in dialysis work environment a necessary requisite for the successful recruitment and retention of nurses. It is also impor- NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3 279

The Relationships Between Nurses Perceptions of the Hemodialysis Unit Work Environment and Nurse Turnover, Patient Satisfaction, and Hospitalizations tant to note that research by the Voluntary Hospitals of America (VHA) (2002) underscores potential financial implications of nurse turnover in dialysis facilities. VHA research findings revealed that hospitals with nurse turnover rates of 21% or more had a 36% higher cost per discharge rate than hospitals with nurse turnover rates of 12% or less. Moreover, hospitals with low nurse turnover averaged a 23% return on assets compared to a 17% return for hospitals with high nurse turnover; low turnover hospitals had lower risk-adjusted mortality scores as well as lower severity-adjusted lengths of stay compared to hospitals with turnover rates. These findings indicate an urgent need for research that links dialysis work environment characteristics and related outcomes to potential cost savings. They also suggest a strong financial justification for investing in dialysis work environments that support professional nursing practice. Such an investment could lead to improved nurse turnover rates and support safe and high quality patient care. Findings from this study also suggest that work environment characteristics that support professional nursing practice may be important for patient hospitalization outcomes in dialysis units. Negative ratings (i.e., disagreement that magnet characteristics were present) on three PES-NWI subscales (Nursing Foundations for Quality of Care, Staffing and Resource Adequacy, and Collegial Nurse-Physician Relations) were significantly related to hospitalization rates for patients receiving dialysis greater than 90 days, and negative ratings on the Collegial Nurse-Physician Relations subscale were significantly related to hospitalization rates for patients receiving dialysis less than 90 days. It is important to note that while 92% to 100% of nurses affirmed the importance of nursing foundations for quality of care, staffing, and nurse-physician relationships as important for their job, only 62% to 82% of nurses agreed that these characteristics were present in their work environments (see Table 2). The importance of these characteristics for negative patient outcomes in hospitals have been consistently documented, and one mechanism for this relationship has been suggested. Clarke and Aiken (2003) note that nurses in hospitals are an important 24-hour surveillance system and are the first to detect early signs of patient complications and make timely rescue responses more likely. In fact, research in hospitals has revealed that work environment characteristics such as staffing and nursephysician relationships are significantly related to failure to rescue, i.e., mortality, in hospitals (Aiken et al., 2002; Needleman, Buerhaus, Matte, Stewart, & Zelevinsky, 2002; Schmid, Hoffman, Happ, Wolf, & DeVita, 2007). Since this is the first study that links work environment characteristics to patient hospitalizations in dialysis facilities, the findings from this study suggest a need to determine the nature of failure to rescue in dialysis facilities and to examine the relationship of dialysis work environment characteristics with this outcome. The significant relationship between negative ratings of the nursephysician relationship and hospitalization rates for patients who have been on hemodialysis for less than 90 days is notable and is consistent with findings from recent research, which found that the involvement of a multidisciplinary team within the first 90 days a patient is receiving hemodialysis treatments decreases patient mortality and morbidity rates (Wingard et al., 2005). Since negative perceptions of collegial nurse-physician relationships was related to patient hospitalization rates at less than 90 days and greater than 90 days, the addition of content and activities that support a collaborative nurse-physician relationship should be encouraged in dialysis programs. Limitations Nurses in this study were recruited from a single dialysis organization. Their responses to the survey questions may have been influenced by an organizational culture that is unique to the dialysis organization and could have resulted in response biases. Future research should sample nurses from multiple dialysis organizations. A second limitation is the potential for socially desirable responses by nurse participants who were recruited to participate by nurse managers of the facility and who completed the questionnaire with a member of the research team present. The use of survey distribution strategies that minimize the potential for responses that could be socially desirable (e.g., mail surveys to participants homes) is recommended for future studies. Finally, only correlations were reported in this study. Future research should examine the predictive ability of the work environment for patient and nurse outcomes. Conclusions This is the first study that has examined the relationship between staff nurses perception of the work environment and outcomes in dialysis facilities. Findings revealed significant relationships between work environment perceptions and nurses intentions to leave their jobs, facility nurse turnover rates, and the number of patient hospitalizations. These findings clearly indicate an urgent need for research that examines the predictive ability of the work environment for patient and nurse outcomes. Opportunities exist for dialysis facilities to create and maintain favorable environments for nurses. The findings from this study also underscore the need for organizational and unit level strategies targeted at facilitating and supporting nurse autonomy and control over nursing practice, improving nurse staffing, and fostering collegial relationships between nurses and physicians in dialysis facilities. Ultimately, the employment of successful structures and processes of nursing care in dialysis facilities that support professional nursing practice could help dialysis organizations recruit and retain satisfied nurses and foster the delivery of quality nursing care. 280 NEPHROLOGY NURSING JOURNAL May-June 2007 Vol. 34, No. 3