| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Nicholas G. Castle, PhD, Health Policy and Management, University of Pittsburgh, A649 Crabtree Hall, Pittsbugh, PA 15261. E-mail: castleN{at}Pitt.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Turnover • Top management • Nursing home staff
Not surprisingly, given the chronic nature of the problem, researchers have examined interventions to reduce turnover. These interventions include training (Accorinti, Gilster, & Dalessandro, 2000), preemployment screening (Kettlitz, Zbib, & Motwani, 1997), team-care processes (Chapman, 1999), job design (Teresi et al., 1993), and staff support (Riggs & Rantz, 2001). Nurses and nurse aides provide the majority of resident care, so most research and interventions are directed toward these caregivers. Few studies have examined the turnover rate of top management, which in this case is defined as the administrator and director of nursing (DON). However, this may be a significant omission from the literature, because top management turnover may affect caregiver turnover.
Levels of Staff Turnover in Nursing Homes
Recent studies addressing turnover rates of staff in nursing homes are shown in Table 1. Focusing solely on Veterans Affairs nursing homes, Brennan and Moos (1990) found the average annual turnover rate of all staff to be 46%. In their study, set in 254 nursing homes, Banaszak-Holl and Hines (1996) found an average annual nurse aide turnover rate of 32%. In 1998, an American Health Care Association (AHCA) study of 12 nursing home chains reported an annual turnover rate of 59% for RNs and 50% for LPNs (Buck Consultants, 1999). Other studies have also shown RN and LPN turnover to be high. Anderson, Issel, and McDaniel (1997) found LPN turnover to be 103% per year, and RN turnover to be 64% per year.
|
Linking Top Management Turnover and Caregiver Turnover
I propose that the turnover rates of top management can influence subsequent caregiver turnover rates in three ways. First, top management turnover will have a destabilizing influence within the facility. Second, top management turnover will influence employees' commitment to the organization. Third, top management turnover will influence how resident care and services are provided.
Singh and Schwab (1998) state that high administrative turnover may have a "destabilizing influence" (p. 310). Indeed, both the general management literature and nursing home management literature have consistently identified negative organizational performance to be associated with top management turnover (e.g., Dreher, 1982). In the management literature, Finkelstein and Hambrick (1990) show that executive turnover leads to less consistent organizational outcomes, whereas Clingermayer and Feiock (1997) identify it with increased subsequent transaction costs for the organization.
In the nursing home literature, researchers have associated longer DON tenure with better resident outcomes (Anderson et al., 2003; Zimmerman et al., 2002). One further destabilizing influence of top management turnover could be dissatisfaction of other staff. It is intuitive that if top managers are repeatedly seen as not wanting to work at a facility (as shown by their exodus), then other members of the staff may likewise come to question their own institutional loyalty. Grau, Chandler, Burton, and Kolditz (1991) have shown that nursing home top management can influence the institutional loyalty of nurse aides.
Empirical studies suggest that top managers influence employees' commitment to the organization, as well as their turnover rates (e.g., Mathieu & Zajac, 1990). Although they did not examine top management turnover rates, numerous hospital-based studies also have shown that RN turnover rates are associated with managers' leadership attributes (e.g., Boyle, Bott, Hansen, Woods, & Taunton, 1999) and supervisors' leadership attributes (Taunton, Krampitz, & Woods, 1989).
Moreover, caregivers' dissatisfaction with top management, leading to their turnover, may manifest itself in other ways. Top managers of nursing homes do not directly provide resident care, but they are responsible for the care provided by caregivers in their facilities. In this way, they can have a significant impact on the types of services provided and the quality of those services (Castle, 2001). They do, for example, have significant influence over the facility budget and can control the distribution of monies for care and services. A leadership void, resulting from top management turnover, could negatively influence care and services (Castle, 2001). For example, Castle (2001) recently determined that shorter top management tenure was associated with poor resident outcomes. In turn, because of this negative influence on care and services, resident caregivers may be more inclined to leave the facility.
Top managers, upon joining a facility, need to become accustomed to the basic practices of the new facility. While they are doing this, their attention to staff concerns, pay, and benefits may fall by the wayside (at least initially). This may be a third way that top management turnover affects resident caregiver turnover. Indeed, some recent research by Brannon, Zin, Mor, and Davis (2002) would suggest that supervision, leadership, and rewards are important influences on the turnover rates of nurse aides. The hospital nursing literature also would seem to support this view (e.g., Taunton et al., 1989).
Conceptual Framework and Research Hypothesis
For this analysis, I found no empirical studies examining the association between nursing home top management turnover rates and caregiver turnover rates. Nevertheless, the aforementioned literature suggests that the turnover rate of top management can influence subsequent caregiver turnover rates. I hypothesize that high (low) levels of top management turnover will be associated with high (low) levels of resident caregiver turnover.
To examine this hypothesis, I used a conceptual model developed by Banaszak-Holl and Hines (1996). I chose this model first because it was developed specifically in nursing homes, and second because other authors have successfully used similar models in the nursing home setting (Anderson et al., 1997; Brannon et al., 2002).
Numerous conceptual and theoretical models of turnover exist in the literature (e.g., Bluedorn, 1982). These models examine both actual turnover and intent to turnover; they also include a wide variety of variables including demographic, job, organizational, wage, and market variables. In reviewing turnover studies, Price (2001) further codified these variables as belonging to three basic factors: individual, structural, or environmental. Individual factors (e.g., demographic variables) are characteristics of the individual worker; structural factors (e.g., job and organizational variables) are characteristics of the work setting; and, environmental factors (e.g., market variables) are characteristics external to both the individual and organization.
Similar to several other turnover models, the model developed by Banaszak-Holl and Hines (1996) includes both structural and environmental factors, and it utilizes variables for job design, organizational characteristics, residents, and the market. This model is not intended to explain turnover of individual staff; rather, its utility lies in its ability to explore "very high and very low facility turnover, drawing on factors identified in prior work to be correlated with facility turnover rates" (Brannon et al., 2002, p.159). This conceptual model is germane to this investigation because I am most interested in examining the influence on aggregate (high and low) facility-level caregiver turnover of one additional structural variable, the turnover of top management.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
I collected primary data from facilities in five states: Kansas, Maine, Mississippi, South Dakota, and Texas. I chose these states because they participated in the Centers for Medicare and Medicaid Services (CMS) Multi-State Case-Mix and Quality (NHCMQ) Demonstration Project and thus have been of interest to researchers. These states have no conceptual or theoretical relevance to this study and may not be representative of all states. For example, they likely overrepresent rural areas. I used only five states because resources were limited for this initiative and consequently the sample had to be limited as well.
I stratified facilities by state and then chose a random sample of approximately 50% of facilities from each state's pool of eligible facilities. The turnover questionnaire was included as part of a larger study examining nursing home outcomes. This limited my study because hospital-based facilities and facilities that were part of a retirement center were excluded from this other nursing home outcomes initiative. Eligible facilities included 74 nursing homes from Kansas, 23 from Maine, 81 from Mississippi, 16 from South Dakota, and 632 nursing homes from Texas.
The OSCAR procedure is conducted by state licensure and certification agencies as part of the yearly Medicare–Medicaid certification process, and it includes approximately 17,000 facilities. I used data only from those facilities that participated in primary data collection. Facility information in the OSCAR data is lagged by between 6 and 18 months. Therefore, I used data from both 1997 and 1998 to identify information representing the 1997 status of the facilities that participated in primary data collection.
There are approximately 300 data elements in the OSCAR database, the majority of which are either organizational or aggregate resident data. Facility data relevant to this study are chain membership, occupancy rate, and ownership characteristics. Resident data elements relevant to this study include the number of residents who have limitations in activities of daily living (ADLs).
The OSCAR data constitute a widely used secondary source of nationally representative nursing home data. A recent Institute of Medicine (IoM) report from 2001 advocated the use of OSCAR data for research, but it also cautioned that these data do have some limitations. These limitations have been described elsewhere (Castle, 2001). Most notably, these limitations include limited observation by surveyors when they visit a facility. Resident characteristics are obtained only partially by direct observation by the surveyors. The facility provides information on resident characteristics and the surveyors select a small sample of residents to verify the information. In addition, the information the surveyors report is pertinent only for the time they make rounds in the facility, which usually occurs during the day shift; although it should be noted that Hughes, Lapane, and Mor (2000) found the facility characteristics in the OSCAR system to be similar to those reported in the 1997 National Nursing Home Survey.
I also used the 2002 Area Resource File (ARF). This is a publicly available data set summarizing a large array (several thousand variables) of census, health, and social resource information for all counties in the contiguous United States (Stambler, 1988). In this investigation, I used the ARF to measure economic conditions in the county, including the unemployment rate, per capita income, and number of nursing home beds. The 2002 data include these figures for 1997, so the OSCAR database and ARF could be matched with presumably little measurement error.
Analytic Approach
The subject of this investigation is the association of caregiver turnover with top management turnover. Brannon and colleagues (2002) have previously shown that high and low turnover rates may have different antecedents. To account for this possibility, I use multinomial logistic regression models, with one model examining nurse aide turnover and the other model examining RN and LPN turnover. In both models, the (adjusted) risk of high caregiver turnover is estimated relative to another group, and the (adjusted) risk of low caregiver turnover is estimated relative to another group. In both analyses, the referent other group is nursing homes with medium levels of caregiver turnover. In this way, the "competing" outcomes of high and low turnover are controlled for. Multinomial logistic regression is a generalization of the more commonly used dichotomous logistic regression, which may be used when there is an alternative outcome category that may occur instead of the event of interest.
Levels of caregiver turnover can be divided in many ways. To facilitate multinomial logistic regression, I used tercile scores to define high, medium, and low turnover levels of caregivers. I defined facilities with an average of 9–37% nurse aide turnover per year as nursing homes with low nurse aide turnover, facilities with an average of 38–69% turnover per year as medium turnover facilities, and facilities with an average of more than 69% turnover per year as nursing homes with high nurse aide turnover. I defined facilities with an average of 5–21% RN and LPN turnover per year as nursing homes with low RN and LPN turnover, facilities with an average of 22–45% turnover per year as medium turnover facilities, and facilities with an average of more than 45% turnover per year as nursing homes with high RN and LPN turnover. These levels are arbitrary, but in sensitivity analyses varying the cutoff values (not shown), the results presented were robust.
I examined the correlations between the variables (not reported), and, based on a threshold of.8, they showed no problems of collinearity (Kennedy, 1992). Values for regression tolerance statistics (not reported) also showed no problems of multicollinearity.
Model Specification and Operationalization
Following the turnover model developed by Banaszak-Holl and Hines (1996), I included job design, facility, resident, and market variables. The job design variables included were nurse and nurse aide staffing levels; resident variables were ADLs and dementia; facility variables were size, chain membership, ownership, private-pay census, and occupancy; and market variables were the unemployment rate, per capita income, and number of nursing home beds in the county. I included rural location, as workers in rural facilities may have fewer alternative employment opportunities (Decker et al., 2003; Harrington & Swan, 2003). I included top management turnover as the independent variable of interest.
The definitions for variables are given in Table 2. For nurse aide turnover, administrators were asked to report the turnover rate for the previous year (1998), including aides who were full time, part time, or on contract. For RNs and LPNs, a similar question asked administrators to report the turnover rate for the previous year (1998), including RNs and LPNs who were full time, part time, or on contract. Turnover was defined as the number of staff no longer employed by the facility (e.g., terminated or resigned) divided by the number of established positions. A similar definition of turnover was recently used by Decker and associates (2003).
|
In administering the questionnaire, I took a narrow definition of nursing home top management and included only the administrator and DON of record. However, the study included a broad scope of individuals that could be in these positions, including whether they were full time, part time, or on contract with the nursing home. I did not include assistant administrators and assistant DONs in the turnover rate, because in many facilities these staff perform more of a clerical role than an administrative role. In addition, assistant administrators and assistant DONs are not employed by smaller facilities.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Table 2 presents the descriptive data. Of particular interest, nurse aide turnover rates averaged 58% in 1998, and RN and LPN turnover rates averaged 32%. Top management turnover was quite frequent. Overall, top management turnover rates averaged 39% in 1997. Turnover of administrators and DONs varied slightly (not shown), with administrators having an average turnover rate of 42% per year and DONs 36% per year. However, across all 5 years of data, administrator and DON turnover rates were correlated (r =.76).
Adjusted odds ratios (AORs) and 95% confidence intervals (CIs) for the multinomial logistic regression models examining the association between caregiver turnover and top management turnover are presented in Table 3. The second and third columns of results in this table examine turnover rates for nurse aides. The results show that top management turnover is significantly associated with high nurse aide turnover. Specifically, a 10% increase in top management turnover rates is associated (p <.05) with a 21% increase in the odds that a facility will have high nurse aide turnover rates (relative to the medium turnover group). In addition, top management turnover is significantly associated with low nurse aide turnover. A 10% increase in top management turnover rates is associated (p <.05) with an 8% decrease in the odds that a facility will have low nurse aide turnover rates (relative to the medium turnover group).
|
Among the control variables, in both multinomial logistic regression analyses, ADLs, for-profit ownership, private-pay census, and number of nursing home beds in the county were consistently significantly associated with caregiver turnover. In both multinomial logistic regression analyses, full-time-equivalent RNs and LPNs, dementia, size, and chain membership were significantly associated with high caregiver turnover, but not low caregiver turnover. The AORs for chain membership were particularly noteworthy, with an AOR of 1.40 (p <.01) for high nurse aide turnover and an AOR of 1.32 (p <.1) for high RN and LPN turnover. In both multinomial logistic regression analyses, full-time equivalent nurse aides and occupancy were significantly associated with low caregiver turnover but not high caregiver turnover.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Results show that top management turnover is associated with resident caregiver turnover. With cross-sectional data I am not able to show a causal relationship; nevertheless, this result may be significant. It would seem to highlight the importance of top managers in nursing homes. The commitment that top managers show to the organization clearly influences other staff. It follows that reducing top management turnover or more appropriately managing facilities experiencing such turnover may be beneficial.
The literature search did not produce any interventions in long-term care settings designed to reduce top management turnover. More research is needed in this area. Singh and Schwab (2000) have provided an interesting start to this kind of work. They suggest profiling top managers and hiring those with a low turnover "profile." The management literature also may provide some clues to help reduce top management turnover. For example, having fair compensation practices, implementing executive development activities, and encouraging a sense of fit with the organization are cited as important components to retention (Mitchell, Hotom, & Lee, 2001). Corporate offices of chain facilities also may benefit from initiating formalized transfer policies (Dalton & Todor, 1993). These policies may promote some initial top management turnover, but in the long run they will increase organizational commitment and eventually reduce turnover (Dalton & Todor). Corporate offices and individual owners also could review their termination policies. Little is known about this involuntary form of turnover, but on the basis of anecdotal evidence we believe involuntary turnover of top management may be high.
These results provide few clues as to how or why top managers influence the turnover of other staff. Top management may have a general destabilizing influence on the organization, may influence employees' commitment to the organization, and may influence resident care and services. On the basis of recent reviews of the turnover literature (e.g., Price, 2001), we speculate that all of these factors probably play a role in staff turnover following the departure of top managers. However, incoming top managers could lessen some of these negative influences.
When top managers leave the nursing home, the general destabilization increases transaction costs for their replacements (Clingermayer & Feiock, 1997). This destabilization can be reduced if corporate offices or individual owners focus on succession planning for top management (Ocasio, 1999). In addition, research on reducing transaction costs could focus on identifying policies and procedures to be used in facilities when top managers depart (Clingermayer & Feiock, 1997).
The importance of employees' commitment to the organization is clear from empirical studies. Less committed employees are more likely to leave the organization (Grau et al., 1991). Network exchange theory suggests that commitment comes from frequent staff interactions (Lawler & Yoon, 1998). As Van Der Merwe and Miller (1971) describe, "satisfying interactions are unlikely in groups which are temporary in nature, and are constantly in a state of erosion and replacement as a result of high labor turnover" (p. 239). Clearly, one high priority of the new top manager should be to foster organizational commitment. Positive influences in this regard include encouraging employee participation in decision making, promoting teamwork (Lok & Crawford, 2001), and interacting frequently with staff (Lawler & Yoon).
Turnover of top management also may influence resident care and services (Castle, 2001). However, the job satisfaction literature tells us that caregivers can become dissatisfied when care quality declines (Irvine & Evans, 1995). Again, one high priority for the new top manager should be resident care (Irvine & Evans, 1995). Admittedly, resident care should always be a priority for top management, but for a new top manager acculturating with the facility and no doubt involved in immediate day-to-day crises and problems, resident care could become a lesser concern.
If we assume that my speculations as to how and why top managers influence the turnover of other staff are correct, then these proposed remedial actions would seem almost trivial to implement. However, Castle and Banaszak-Holl (2003) remind us that nursing home management structures are characteristically flat and generally understaffed. Therefore, many top managers may be overburdened with daily operational concerns, thus making the additional tasks a significant addition to the daily workload. Castle and Banaszak-Holl proposed that we may need both more and better top managers in nursing homes. I add that we probably also need more and better administrative protocols, similar to those already discussed.
Given the high number of significant variables and relatively high pseudo-R2 levels in both analyses, the turnover model used would appear to have some utility to the nursing home setting. Using a similar model, Banaszak-Holl and Hines (1996) found that for-profit ownership (p <.05) and resident case mix (p <.10) were associated with nurse aide turnover rates. Similarly, Brannon and associates (2002) found that for-profit ownership (p <.00) and chain membership (p <.09) were associated with high nurse aide turnover rates, and RN turnover (p <.03) was associated with low nurse aide turnover rates. I find similar robust results for for-profit ownership and resident case mix. In addition, private-pay census and number of nursing home beds in the market were consistently significant in my models.
Results of this study also provide further evidence that high and low turnover in nursing homes can be influenced by different factors, as Brannon and associates (2002) assert. For example, the results show that bed size and chain membership are associated with high caregiver turnover rates but not low ones. Analyses further show that not only are high and low turnover rates in nursing homes influenced by different factors, but these factors also differ for different staff. Both high and low nurse aide turnover rates were found to be significantly associated with top management turnover, but only high RN and LPN turnover rates were significantly associated with top management turnover. It is not entirely clear why these different relationships exist, but these results suggest that professionals (i.e., RNs and LPNs) and paraprofessionals (i.e., nurse aides) may have different expectations of top management. Professionals may expect top management stability, whereas paraprofessionals may not; however, when stability occurs, paraprofessionals' expectations are exceeded and they remain at the facility. Alternatively, using Price's (2001) model of turnover, one could speculate that top management stability influences paraprofessionals' job involvement, job stress, or promotional chances, thereby reducing their turnover.
Limitations of the Study and Suggestions for Further Research
The dependent variables, nurse aide and RN and LPN turnover rates, may benefit from further refinement. I found no consensus in the literature on an operational definition of caregiver turnover, and several options exist in the literature. Indeed, Price (1977) described the accession rate, stability rate, and wastage rate as alternative methods to measuring turnover. It also should be noted that turnover measures staff participation in the facility, but an interdependent measure of staff participation that could be examined in the future is the absence rate (Dalton & Todor, 1993).
Although I believe the questions on turnover were relatively well conceived, some measurement error is likely to exist. I do not know whether the administrators in the sample monitored caregiver turnover, or whether they simply provided a best guess. I also have no idea whether the turnover rates that administrators provided accurately matched the staff and staff characteristics provided in the questions.
Including a measure limited only to voluntary turnover for both caregivers and top management may be useful in future studies. Top management turnover would be expected to influence voluntary caregiver turnover, but will probably have little influence on involuntary turnover. Thus, using only voluntary turnover rates of caregivers would provide a more robust analytic approach. Unfortunately, in this study, information on voluntary and involuntary turnover was combined.
In the case of top managers, involuntary turnover may be perceived by staff as beneficial if administrators or DONs were terminated because of poor performance, although, using this approach, one has to be careful with regard to what constitutes "performance." Financial measures rather than measures of resident outcomes may be more important performance measures in involuntary top management turnover.
Turnover data were examined for administrators and DONs combined. Both of these top managers are responsible for the daily operation of the facility. However, DONs in general are more involved with clinical issues. For example, they determine clinical policies and protocols, and, what is more important for this study, they may be more directly involved with caregivers. This may make caregivers influenced more by DON turnover than administrator turnover. In sensitivity analyses (not shown), no such relationship was found. Nurse aide and RN and LPN turnover seemed to show the same relationship with top management turnover, irrespective of whether the top management turnover came from the administrator or DON. Similarly, sensitivity analyses using an interaction term for administrator and DON turnover were not noteworthy (not shown).
Conclusions
In conclusion, the reasons for staff turnover in nursing homes are of interest to the industry and policymakers. These findings add one further important factor to the body of literature examining turnover—top management. One has to be careful in drawing conclusions from these cross-sectional analyses, but I believe this study provides preliminary evidence that the turnover of top managers of nursing homes has an important influence on staff turnover.
|
Footnotes |
|---|
Decision Editor: Linda S. Noelker, PhD
Received for publication January 13, 2004. Accepted for publication July 21, 2004.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Haley-Lock and J. Kruzich Serving Workers in the Human Services: The Roles of Organizational Ownership, Chain Affiliation, and Professional Leadership in Frontline Job Benefits Nonprofit and Voluntary Sector Quarterly, September 1, 2008; 37(3): 443 - 467. [Abstract] [PDF]
|
|
|
|
 |
|
 C. Howes Love, Money, or Flexibility: What Motivates People to Work in Consumer-Directed Home Care? Gerontologist, July 1, 2008; 48(suppl_1): 46 - 60. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Pillemer, R. Meador, C. Henderson Jr., J. Robison, C. Hegeman, E. Graham, and L. Schultz A Facility Specialist Model for Improving Retention of Nursing Home Staff: Results From a Randomized, Controlled Study Gerontologist, July 1, 2008; 48(suppl_1): 80 - 89. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. T. Barry, P. Kemper, and S. D. Brannon Measuring Worker Turnover in Long-Term Care: Lessons From the Better Jobs Better Care Demonstration: Kathleen Walsh Piercy, PhD, Editor Gerontologist, June 1, 2008; 48(3): 394 - 400. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Stearns and L. P. D'Arcy Staying the Course: Facility and Profession Retention Among Nursing Assistants in Nursing Homes J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2008; 63(3): S113 - S121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L M Wagner, E Capezuti, P C Clark, P A Parmelee, and J G Ouslander Use of a falls incident reporting system to improve care process documentation in nursing homes Qual. Saf. Health Care, April 1, 2008; 17(2): 104 - 108. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. A. Kash, N. G. Castle, G. S. Naufal, and C. Hawes Effect of staff turnover on staffing: a closer look at registered nurses, licensed vocational nurses, and certified nursing assistants. Gerontologist, October 1, 2006; 46(5): 609 - 619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Cohen-Mansfield and A. Bester Flexibility as a management principle in dementia care: the adards example. Gerontologist, August 1, 2006; 46(4): 540 - 544. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. G. Castle Measuring Staff Turnover in Nursing Homes Gerontologist, April 1, 2006; 46(2): 210 - 219. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. G. Castle and J. Engberg Organizational Characteristics Associated With Staff Turnover in Nursing Homes Gerontologist, February 1, 2006; 46(1): 62 - 73. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
