| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Dale E. Yeatts, Department of Sociology, P.O. Box 311157, University of North Texas, Denton, TX 76203. E-mail: yeatts{at}unt.edu
 |
Abstract |
|---|
 TOP Abstract MethodsFindingsDiscussionImplicationsReferences |
|---|
Key Words: Long-term care • Resident care • Nursing home administration • Turnover • Absenteeism
The reaction among practitioners and gerontologists alike has been the development of new models for nursing home care that address these concerns (Kane, 2001; Kane et al., 1997; Kovach & Krejci, 1998). For example, the Pioneer Network was organized in the 1990s with the purpose of identifying "deep systematic change" that will allow for a "person-centered" focus of care (Fagan, 2003, p. 125). Members of this network include practitioners, researchers, educators, and other professionals who seek to find alternatives to the strictly medical model. In 1992 the Eden Alternative was developed by Thomas and Thomas (Thomas, 1994) to encourage a more homelike environment in the nursing home and to allow direct-care workers to be more involved in decision making. A decade later, Thomas and his colleagues introduced the promising concept of "Green Houses," which provide a structural alternative in the form of specially designed houses for elderly residents as well as the empowerment of direct-care workers (Green House Project, 2006; also see Keane, 2004). In 1994, another approach, the Wellspring Model, was established to encourage nursing homes to work together to teach line staff the best clinical practices and to move away from the typical nursing home culture of control (Kehoe & Heesch, 2003; Reinhard & Stone, 2001). In 2001, the LEAP initiative was underway to alter how the nursing home workforce was viewed and treated by nursing home management (Hollinger-Smith, 2003; Hollinger-Smith, Ortigara, & Lindeman, 2001). In a fairly recent development, Grant and Norton (2003) identified various stages that nursing homes are expected to go through in the process of culture change. Additional recent initiatives include those by Gilster and colleagues (Gilster, Accorinti, & Dalessandro, 2005), Pillemer and colleagues (Pillemer, Suitor, & Wethington, 2003), Rosen and colleagues (2005), and Shields (2004).
One characteristic typically found as a component of these and other person-centered initiatives has been the empowering of direct-care nursing home staff (also referred to as frontline staff). It is reasoned that certified nurse aides (CNAs) have the most knowledge about nursing home residents: They know better than any other employees the likes and dislikes of residents, including what they want when they wake up in the morning (e.g., a glass of water or the newspaper), how they would like their hair combed, and what they want to wear when they go to bed at night. Consequently, the CNAs are in the best position to make decisions that are directly related to day-to-day resident care (Beck, Ortigara, Mercer, & Shue, 1999).
In the workplace as a whole, the empowerment of workers has occurred primarily through the use of empowered work teams. These were first implemented on a large scale within the manufacturing sector in the 1980s and 1990s. Empowered work teams typically consist of a group of frontline employees who hold similar job titles and do similar work. The teams are unique in that the team members do more than simply perform their technical or service responsibilities while being supervised. In an empowered work team, the team members take on some supervisory responsibilities and make recommendations regarding others. Evaluation studies of work teams in manufacturing settings have found that empowered work teams result in improved performance, higher job satisfaction, and reduced turnover (Lawler, 1986; Pasmore, Francis, Haldeman, & Shani, 1982; Wellins, Byham, & Dixon, 1994; Yeatts & Hyten, 1998). Further, studies have shown that the decisions made by the teams are at times more innovative and creative than those made by management, because the team members are more familiar with the whole work process (Hitchcock & Willard, 1995). However, these studies also typically report that the success of an empowered work team is highly linked to the success in its implementation. Where teams were poorly implemented, performance sometimes dropped and turnover increased (Lawler; Yeatts & Hyten).
The purpose of the research reported here was to examine the effects of empowered work teams within the nursing home industry. We expected that the empowerment of CNAs through empowered work teams would result in improved performance, job attitudes, and turnover. Although most recent nursing home initiatives encourage the empowerment of CNAs, this has been a secondary focus for them, with little direction regarding how to enact the empowerment of CNAs and no evaluation of its effects.
Review of the Literature
As we stated, empowered work teams are teams whose members typically hold similar job titles, do similar work, and are empowered to make decisions about some aspects of their work and recommendations about others. Empowered work teams are different from interdisciplinary teams in which members have varying job levels (e.g., nurse, social worker, nurse aide), with those at the lower levels providing input but making few if any decisions (Wellins, Byham & Wilson, 1991; Yeatts & Hyten, 1998). Empowered work teams are sometimes referred to as self-directed work teams, self-managed work teams, or autonomous work groups. The specific designation has sometimes mirrored the level of empowerment, with those in self-managed teams having more decision-making authority than those in self-directed teams. However, no clear distinctions have been established in the literature.
Attempts to explain the effects of employee empowerment and of empowered work teams have been presented at both the individual and group levels. At the individual level, there are two theoretical traditions—participative management and employee involvement. Theories of participative management focus on the sharing of decision making between employees and managers and how such sharing enhances organizational performance and employee satisfaction (e.g., Cotton, Vollrath, Groggatt, Lengnick-Hall, & Jennings, 1988; Locke & Schweiger, 1979; Miller & Monge, 1986; Spreitzer, Kizilos & Nason, 1997; Wagner, 1994). Theories of employee involvement emphasize pushing power down to nonmanagement employees in the form of information, rewards, and training, with the same effects of high employee performance and satisfaction (Lawler, 1986). From these perspectives, Kirkman and Rosen (1999), Thomas and Velthouse (1990), and others have expanded the concept of employee empowerment to include more than simply the experience of autonomy in decision making. They explain that to be empowered also means that the employee perceives her or his work to have important effects and to be meaningful, and that she or he is highly competent to do the work (Bandura, 1997; Ford & Fottler, 1995; Spreitzer, 1995; Spreitzer et al, 1997). It is a combination of these factors that instills a feeling of empowerment in the worker.
At the group level, the theoretical basis for the use of empowered work teams also stems from several different schools of thought. The human relations school argues that group participation in decision making allows employees to satisfy higher order needs such as self-actualization and respect, which in turn leads to higher job satisfaction and reduced turnover (Cummings, 1978; Hackman & Oldham, 1980; Lawler, 1986). The cognitive school suggests that group participation in decision making enhances the flow and use of important information (Kren, 1992; Pasmore & Purser, 1993). The contingency model argues that no single theoretical explanation for the effectiveness of empowered work teams holds across all groups and situations. Instead, the empowered work team will affect performance and satisfaction differently for different people and different situations (Galbraith, 1973; Vroom & Yetton, 1973).
Research on empowerment and performance in manufacturing and some service industries has occurred primarily through the study of empowered work teams. This research has, for the most part, supported these theoretical perspectives. In particular, studies have found that the use of these teams has resulted in higher performance and reduced turnover (Bowers, Faan & Jacobson, 2003; Eaton, 2001; Mobley, Griffeth, Hand, & Meglino, 1979; Spector, 1986). Further, associated with empowered work teams have been higher levels of cooperation and coordination among the workers, as well as more feedback and support from management (Belasco & Stayer, 1994; Black & Gregersen, 1997; Bowen & Lawler, 1992; Saxon, 2000).
Research on empowerment and employee attitudes has found that working in an empowered work team is associated with higher job satisfaction (Friedman, Daub, Cresci, & Keyser, 1999; Spector, 1997; Spreitzer et al., 1997; Teresi et al., 1993; Thomas & Tymon, 1994). The general consensus is that employees who feel empowered like their job more, although some research suggests that such positive effects depend on the personal characteristics of the workers themselves (Yeatts & Hyten, 1998). Empowered work teams have also been found to positively affect employee commitment (Manz & Sims, 1987; Somers, 1995; Wellins et al., 1991). In this case, those individuals working in empowered work teams have been found to identify more with the work organization and its goals. Similarly, it is reasonable to expect that an employee's self-esteem is positively affected by employee empowerment (Rosenberg, Schooler, & Schoenbach, 1989).
In summary, empowered work teams provide a setting for a variety of effects, including (a) increased empowerment of the team members, (b) improved performance, (c) better employee attitudes, and (d) reduced absenteeism and turnover. The large majority of this research has focused on the manufacturing industry. Much less has focused on the service industry, and we found little that focused on the long-term-care industry (Yeatts & Seward, 2000). To fill this research and knowledge gap, we tested the following propositions.
Proposition 1: Empowered work teams in nursing homes positively affect feelings of empowerment, including autonomy, impact, meaningfulness, and competence among the team members.
Proposition 2: Empowered work teams in nursing homes positively affect the performance of the team members.
Proposition 3: Empowered work teams in nursing homes have a positive effect on job attitudes, including higher levels of job satisfaction, commitment, and self-esteem, and lower levels of burnout.
Proposition 4: Empowered work teams in nursing homes reduce absenteeism and turnover among the team members.
|
Methods |
|---|
|
TOPAbstractMethodsFindingsDiscussionImplicationsReferences |
|---|
We chose the five nursing homes we selected as the experimental group from 18 volunteer nursing homes located in the north Texas region. The criteria used to select the five experimental nursing homes included the willingness of the nurse management and nursing home administrator to implement empowered work teams and the stability of the nursing home management in terms of job tenure. We based additional selection criteria on a desire to obtain variation in nursing home characteristics with regard to size, location (rural vs urban), and ownership (profit vs nonprofit). Once we selected the five experimental nursing homes, we made an attempt to identify comparison nursing homes of the same size, location, and ownership; we then invited these homes to participate in the study.
A nonequivalent control group design is a quasi-experimental design—that is, the participants in the experimental and control groups are not randomly placed in the two groups. Instead, the participants are already in the experimental and control groups. We made an effort to select a comparison nursing home that was as similar as possible to the matching experimental nursing home. To check for similarity, we conducted baseline surveys of CNAs, nurses, and the residents' family members or significant others at each of the experimental and comparison nursing homes. We also undertook an in-person survey of residents, but the number of residents who were able and willing to be interviewed at both the pretest and post-test time periods was too small for meaningful quantitative analyses. Contributing to the small number was one experimental nursing home that housed only residents with Alzheimer's disease and its comparison nursing home that housed a substantial number of such residents. None of these residents could be interviewed.
A comparison of demographic characteristics of the CNAs, nurses, and family members at baseline for the treatment and comparison groups shows that there were only a few differences between them (Table 1). The experimental group had a slightly lower percentage of female CNAs (83% vs 92%, respectively). The nurses from the experimental nursing homes tended to have worked at the nursing home longer (72 vs 43 months) and were more likely to be non-Hispanic white (79% vs 59%, respectively). When considering the family members or significant others surveyed (henceforth referred to as family members), we found that those from the experimental nursing homes were slightly more likely to be reporting on a female resident than were the family members from the comparison nursing homes (85% vs 78%, respectively).
|
We pretested the procedures that we used to implement the CNA-empowered work teams at an earlier date in a nursing home that is not included in the current study. We describe these procedures in detail elsewhere (Yeatts, Cready, Ray, DeWitt, & Queen 2004). In short, implementation included first orienting and training CNAs, nurses, and nurse management. Once familiar with their respective roles and the expected activities of the empowered work teams, nurse management members began making efforts to involve the teams in management decisions related to CNA work. For example, the empowered work team might be asked to provide revised work procedures for distributing breakfast trays so that residents would get hotter meals than the current procedures allowed, or the team might be asked to revise procedures for answering call lights so that less time elapsed before a CNA responded to a call light.
At the same time, CNAs began holding weekly 30-minute meetings (initially with the assistance of the Principal Investigator or Coprincipal Investigator as team facilitator) and short stand-up meetings during the week as needed. The weekly 30-minute meetings followed a set agenda that not only included addressing issues provided by nurse management but also included other areas of focus such as a review of resident health conditions, review of new residents and their specific needs, and any issues of concern to the CNAs. The team provided weekly written summaries (i.e., minutes) of each team meeting to nurse management, and any CNA recommendations or suggestions regarding their work or regarding the residents were typically provided in these weekly summaries. The nurse management members reviewed the team summaries and provided weekly written feedback to each empowered work team. Feedback typically included responses to any CNA questions or concerns.
If nurse management members were responding to a change to the work process that was proposed by an empowered work team (e.g., the process for handing out meal trays), they typically responded in one of several ways. In some cases, they approved the proposed change as submitted and they asked the empowered work team to begin using the revised procedure. In other cases, the nurse management sent the proposed change back to the team, pointed out some shortcomings of the proposed procedure, and asked the team members to resubmit their proposed change with this new information in mind. Once management and the empowered work team came to an agreement on what would work, the proposed change was implemented.
Short stand-up meetings were used by the empowered work team to address more immediate concerns handed to them by nurse management. For example, nurse management might call the CNAs together and ask them to determine the best way for the empowered work team to get all the residents to a planned activity, given that one of their team members was unexpectedly absent. The nurse manager would then leave and the CNAs would discuss the alternatives, make a decision, and implement it.
Although we initially selected five pairs of nursing homes for the study, we had to drop one pair. In this pair, empowered work teams were initially implemented within the experimental nursing home successfully. However, about 6 months after team implementation, the Assistant Nursing Home Administrator chose to assume the role of team facilitator. Subsequently, the recommendations made by the team were not only provided to nurse management but were immediately acted on by the Assistant Administrator. As a result, the nurse management members perceived this as usurping their authority; in other words, the CNAs were "going around" the nurse management and taking their concerns and suggestions directly to the nursing home administration. This empowered the CNA teams greatly until the Administrator left the institution for another position and the Assistant Nursing Home Administrator did likewise about 6 weeks later. The nurse management then chose to discontinue the empowered work teams; the new Nursing Home Administrator, being unfamiliar with the teams, followed the recommendations of the nurse management members in discontinuing the teams.
In addition, empowered work teams were initially started among night-shift CNAs, but after some problems were experienced, most were eventually discontinued. The primary problems encountered were the small numbers of CNAs who worked the night shift and the irregular staffing of these CNAs. In most of the experimental nursing homes, there were three or fewer night-shift CNAs working in a service area (e.g., one floor of the nursing home). Further, of these CNAs only a few were permanent night-shift staff—the others worked primarily on the day or evening shifts and occasionally worked on the night shift.
Data Collection
The timing for the collection of the post-test data varied somewhat from nursing home to nursing home, depending on the schedules and availability of the staffs to participate in the follow-up surveys. On average, there was a 16-month period between Time 1 and Time 2 for the experimental nursing homes and a 17-month period for the comparison nursing homes. Previous research has shown that the effects of interventions, such as empowered work teams, can look promising in the short term of 6 to 12 months (the "honeymoon" effect), but the effects become less pronounced over time (Lawler, 1986; Lawler & Mohrman, 1987). Therefore, we selected the 16- to 17-month follow-up in part to avoid making such premature conclusions.
The baseline and post-test data collection included self-administered questionnaires for the CNAs and nurses and a mail survey of the residents' family members. In most cases, the self-administered questionnaires were distributed by the research team at an all-staff meeting of CNAs and nurses. Each questionnaire was given a unique number and placed within an envelope addressed to a specific CNA or nurse, who was subsequently handed the envelope by the researcher. This resulted in each CNA having a unique questionnaire number. Once completed, the questionnaires were immediately collected by the research staff. Thus, the CNA or nurse could fill out the questionnaire without her or his name on it. Simultaneously, the researchers had a unique number for each CNA and nurse that was given to them again at the follow-up survey so that their responses from the second survey could be matched and compared with the first. CNAs and nurses who were not in attendance were later approached by a member of the research team and invited to participate. In a few cases, the nursing home management preferred that the questionnaires not be distributed at all-staff meetings. In these cases, the researchers approached the CNAs and nurses at their work locations during their work hours and invited them to participate. Not using the all-staff meeting to collect the data did not appear to affect the quality of the data collected or the response rates. For the survey of family members, the nursing home management typically provided the researchers with a mailing list that consisted of one family member (or significant other) for each nursing home resident. These individuals were mailed a questionnaire with an addressed, stamped envelope that was mailed back to the researchers.
The pretest response rates for CNAs at the experimental and comparison nursing homes were 84% and 92%, respectively, whereas the response rates of nurses were 84% and 80%, respectively, and the pretest mail response rates for family members were 69% and 52%, respectively. Response rates from the post-test surveys were similar though generally slightly lower.
Questionnaire Items, Concepts, and Indices
The survey instruments for the CNAs and nurses asked them to respond to a series of statements by using a 5-point Likert-type scale ranging from 1 for "strongly disagree" to 5 for "strongly agree." We drew many of the statements from a number of existing instruments that measure employee empowerment and the other work-related concepts. These instruments included those developed by Cook and colleagues (Cook, Hepworth, Wall, & Warr, 1979), Hackman and Oldham (1980), Maslach and colleagues (Maslach, Jackson, & Leiter, 1996), McGee and Ford (1987), Quinn and Staines (1979), Spreitzer (1995), and Yeatts and Hyten (1998). When necessary, we modified statements to reflect the uniqueness of the nursing home environment. For example, we replaced the word recipients by the word residents in burnout statements such as "I feel I treat some residents as impersonal objects" (Maslach et al, 1996). When we could not find statements in previous studies, we developed statements and pretested them at a nursing home not included in the study.
The family questionnaire used a 5-point Likert-type scale ranging from 1 for "yes, always" to 5 for "no, never." We reversed these scores to maintain consistency across the tables; that is, the larger the number (5), the more positive the response ("yes, always"). In constructing the family questionnaire, we took several questions from a 17-item satisfaction scale presented by Kruzich and colleagues (Kruzich, Clinton, & Kelber, 1992). Additional questions came from a nursing home satisfaction survey instrument developed jointly by the Scripps Gerontology Center and the Margaret Blenkner Research Center (Straker, 2001) and from an instrument presented by Uman (Cohen-Mansfield, Ejaz & Werner, 2000). We drew still other questions from instruments developed by Bliesmer and Earle (1993), Davis and colleagues (Davis, Sebastian & Tschetter, 1997), and Kleinsorge and Koenig (1991).
We typically used the questionnaire items to create indices to represent various concepts. The majority of indices consisted of three or more items. For example, the indexed variable for CNAs' perceptions of global empowerment consisted of 19 statements, including those items measuring its dimensions of autonomy (e.g., "The nurse aides decide the procedures for getting residents to the dining room"), competence (e.g., "When a new resident is admitted, I am given all the information I need about the new resident"), and impact or meaningfulness (e.g., "The management staff listens to the suggestions of CNAs"). In contrast, the indexed variable for general job satisfaction consisted of only three statements (e.g., "Generally speaking, I am very satisfied with my work"). Further, two of the concepts from the CNA and nurse questionnaires, CNA self-reported absenteeism and CNA satisfaction with scheduling, were measured by only a single statement. Four concepts from the family questionnaire also relied on a single item: spends time on needs, checks on comfort, satisfaction with the care provided, and satisfaction with staff friendliness.
We determined the specific items included in an index after examining factor analyses and Cronbach's alpha based on standardized items. We calculated standardized alphas twice for each concept—once at Time 1 (pretest) and again for Time 2 (post-test). They ranged in size from a low of 
= 0.50 (autonomy at Time 1) to a high of = 0.90 (nurse perception of CNA empowerment, Time 1, and nurse perception of time available for paperwork, Time 1). The majority of CNA indices ranged in their standardized alpha values from = 0.60 to = 0.85. All the nurse indices ranged between 0.71 and 0.90, and the majority of family indices ranged in their standardized alpha values from = 0.70 to = 0.85.
To calculate each index, we added the items for a specific concept together, and then we divided the resulting sum by the number of items added together. This calculation allowed the index score to remain in the original range of the individual items.
Concepts that were measured from CNA questionnaire statements included the following: global empowerment and its dimensions of autonomy, impact or meaningfulness, and competence, rating of CNA performance, self-esteem, burnout, job satisfaction, satisfaction with scheduling, commitment, intent to quit, and absenteeism. Originally, we designed the questionnaire to measure impact and meaningfulness as two separate concepts. However, a factor analysis and standardized alphas showed that the statements used to measure these concepts actually measured a single concept. Therefore, we created a single index; we refer to this as impact or meaningfulness. We determined CNA turnover by determining the percentage of CNAs working at Time 1 who were still working at the same nursing home at Time 2. We made a separate calculation for the experimental nursing homes and for the comparison homes.
Concepts that we measured from nurse questionnaire statements included the following: global CNA empowerment, CNA procedures, CNA coordination, CNA cooperation with nurses, and nurse time to complete paperwork. Concepts concerning the direct care to residents that we measured from family member questionnaire items included these: spends time on resident needs, checks on resident comfort, responds to resident complaints, and listens, talks, or cares for the resident. Other concepts from the family questionnaire included satisfaction with care provided, satisfaction with staff friendliness, residents have choice of bed time, residents have choice of meal time, and residents have choice of shower time.
Qualitative Approach
The qualitative approach consisted of (a) observations of over 270 CNA team meetings, (b) examination of weekly team-meeting summaries provided by the CNA-empowered work teams to nurse management, and (c) examination of written weekly responses and requests from nurse management to the empowered work teams. For each team, the first 8 to 12 weekly meetings were observed by the Principal Investigator or Coprincipal Investigator, who also served as the facilitators of the first two to four weekly meetings. We determined that the 30-minute team meetings provided the most fruitful information about the team's effects, because the CNAs had a block of time available to them to make decisions that affected their work processes and performance, to express their attitudes about their work, and to discuss turnover and absenteeism issues. The times of the weekly meetings were typically established by the teams themselves and were usually scheduled either during the slowest part of the team's shift or during the shift change so that CNAs from the next shift could join those from the former shift. We randomly observed additional meetings roughly once every 3 months for the following 12 months. In most cases, immediately following an observation, the observer jotted down a summary of what occurred at the meeting. The notes were organized chronologically. Additionally, copies of the weekly summaries submitted to nurse management by the empowered work teams were provided to the researchers for the first 12 weeks that the team met and then irregularly after that (roughly once every 3 months for 9 additional months). Further, copies of nurse management's written weekly responses to the empowered work teams each week along with any other information they wanted to share with the team were also provided to the researchers for the first 12 weeks and then roughly every 3 months after that for 9 additional months.
Data Analyses
To test the propositions, we analyzed both the qualitative and quantitative data. We conducted a review of the qualitative data (observations, empowered work team weekly summaries, and nurse management weekly written responses) separately for each proposition. During each review, we examined the data for trends that refuted as well as supported the proposition. We present the qualitative findings for each proposition.
For the quantitative analysis, we included only those responses from CNAs, nurses, and family members who participated in both the pretest and post-test surveys. We reasoned that changes in the responses of these CNAs, nurses, and family members, if any, would most accurately reflect any effects of the empowered work teams.
We performed several reliability checks of the quantitative data. For three pairs of statements included in the questionnaire for CNAs, each statement in a pair was either identical or the statements were the same except that one was worded in the opposite direction (positive statement vs negative statement). If the respondent was not consistent in his or her responses on at least two of the three pairs of questions, we examined the responses for possible coding errors or excessive influence on overall scores. Subsequently, we found no such respondents. In addition, for the multivariate analyses, we used Mahalanobis distance scores to uncover any respondents who were outliers on the variables used. Here again, we identified no respondent as needing to be removed from further analysis.
To examine change between Times 1 and 2, we used t tests for the experimental group and for the comparison group. To compare the amount of change in the experimental group with the amount of change in the comparison group, we again used t tests by first creating a new variable that reflects the difference between Times 1 and 2 for each respondent and then regressing this difference on a dummy variable coded 1 if she or he was in the experimental group and 0 if she or he was in the comparison group. The p value associated with the dummy variable coefficient estimate allowed us to determine whether the average change (or difference) across time was significantly different between the experimental and comparison groups. Further, because CNAs, nurses, and family members were clustered within nursing home, we used p values based on standard errors adjusted for this clustering (see StataCorp, 2003, p. 328). Additionally, because the large number of t tests increases the chance of committing a Type 1 error (rejecting the null hypothesis of no difference when it is in fact true), we performed a multivariate analysis of variance (MANOVA). In performing the MANOVAs, we met all assumptions and conducted appropriate diagnostics.
|
Findings |
|---|
|
TOPAbstractMethodsFindingsDiscussionImplicationsReferences |
|---|
Although observation notes clearly show that the CNAs' empowerment was increased, the amount of increase is less clear. Observations revealed that CNAs were at times given opportunities to make decisions but these were not routine. This appeared to be due to the following: (a) a perception by management that some decisions needed to be made quickly and so there was no time to consult with the empowered work teams; (b) nurse management would sometimes simply forget to include the empowered work teams in decision making; and (c) empowering workers is a different way of thinking that requires time and effort and so it was almost always easier for nurse management to make decisions without consulting with the CNA teams.
An examination of the CNA pretest (Time 1) and post-test (Time 2) questionnaire data confirms that there was an increase in empowerment, although the increase was modest (Table 2). The CNAs in the experimental nursing homes reported higher levels of global empowerment, autonomy, impact or meaningfulness, and competence at Time 2 than at Time 1, whereas the CNAs in the comparison group showed no significant change between time periods. In addition, the differences over time on these measures between the experimental and comparison groups were for the most part significant (Table 2, last column).
|
Another means of examining this proposition, while avoiding any contamination issues caused by CNA awareness of empowerment at Time 2, is an examination of the nurses' perceptions (Table 2, bottom panel). Consistent with Proposition 1, these findings suggest that the empowered work teams positively affected CNA empowerment. The nurses in the experimental group perceived more CNA empowerment at Time 2 than at Time 1, whereas the nurses in the comparison group did not perceive any change. Further, when comparing the difference over time between the experimental and comparison groups, we again find a significant effect (Table 2, last column).
Proposition 2 proposes that empowered work teams positively affect performance. This proposition is also supported by both the qualitative and quantitative data. The qualitative or observation data indicate that the empowered work teams assisted the CNAs in becoming more aware of resident health conditions. This reduced the possibility that a resident health problem would go unattended for an extended period of time and increased the possibility that proper health care would be provided. For example, in one team meeting, one of the CNAs informed the others that "Ms. Smith" preferred to be taken to the toilet rather than using a bed pan. The response from another CNA was that Ms. Smith had recently had a hip replacement and should not be gotten out of bed. This important bit of information may have saved Ms. Smith a serious reinjury of her hip, because all the CNAs quickly agreed that Ms. Smith should not be gotten out of bed in the future until the nurses and doctor found the hip to be in a condition to do so.
Second, the empowered work teams provided CNAs with information regarding the special care needs, uniqueness, and preferences of residents. For example, a CNA might report in a team meeting that a particular resident prefers using a particular bathroom. In another case a CNA reported having difficulty turning "Mr. Jones" in bed without his becoming combative. Another CNA responded by explaining that Mr. Jones is willing to be turned as long as he is slid up against the wall prior to being turned.
Third, team meetings provided an opportunity for CNAs to have frank discussions regarding any questionable behaviors of other CNAs. Such discussions provided the opportunity to clarify one's actions. Clarification often resulted in reduced frustration and animosity among CNAs and an increased willingness to cooperate and coordinate the care provided to residents. In some cases, poorly performing CNAs were confronted by the empowered work team and provided instruction by the team on how to improve their performance. This instruction was also found to be valuable for new CNAs. When a CNA was repeatedly performing at a low level, it would sometimes be brought up in a team meeting. Team members would describe why the performance was poor and clarify what should be done to improve the performance. This sometimes resulted in the poor performer learning from the team and subsequently improving. In extreme cases, if the poor performer was unreceptive to constructive criticism, the empowered work team contributed to the firing of the CNA by reporting the poor performance to nurse management on multiple occasions.
Fourth, new CNA work procedures appeared to be carried out more willingly by CNAs and to remain a part of their work routine longer when the CNAs participated in creating the new procedures. For example, in one nursing home there were three pairs of CNAs serving three halls. One of the pairs of CNAs was taking longer to get the work done, so nurse management suggested that the CNAs from the other two halls take turns assisting those on the third hall. The CNAs were unhappy with this solution because they believed that the slower pair of CNAs was simply not putting in enough effort. After some discussion with the CNAs, the Director of Nursing (DON) allowed the CNAs to implement their own solution, which was for the pairs of CNAs to take turns working on the third hall. The result of this solution was that the CNAs came to realize that the third hall was indeed more difficult to serve. Consequently, they no longer resented having to "help out" the CNAs on this third hall.
The qualitative or observation data also provided evidence that the empowered work teams may have simultaneously had some negative effects on resident care. This was caused by pulling CNAs away from providing direct resident care for 30 minutes while they attended a weekly team meeting. In all experimental nursing homes, the CNAs had difficulty finding time for their team meetings. The job of a CNA is a difficult one; the CNA does not have time to do all that needs to be done and consequently must choose which tasks will get done and which will not. The team meetings reduced the time available to provide resident care by at least 30 minutes each week.
In addition, the positive effects of the empowered work teams on performance appeared to be experienced in a repeating cycle from high effect on performance to less effect and back again. When the empowered work team had an issue to address that was perceived to be important to the nursing home or to the CNAs themselves, there was a high level of energy and enthusiasm applied to the issue. When there were no burning issues of interest or DON-requested problems to be solved during the team meetings, the CNAs showed less enthusiasm and subsequently it appeared that less was accomplished.
The quantitative or survey data also support the performance proposition. For example, when experimental group CNAs were queried about the empowered work teams in the post-test questionnaire, the majority of them agreed or strongly agreed that the teams allow them to learn from each other (58%) and learn what the residents like and dislike (59%), with 22% and 19% disagreeing or strongly disagreeing, respectively (Table 3). Similarly, the majority of nurses reported that the empowered work teams provide new ideas that are helpful (57%) and create helpful ways of doing their work (59%), with 11% and 9% disagreeing or strongly disagreeing, respectively. Sixty percent of the CNAs disagreed with the statement "We should stop using teams," and only 15% agreed or strongly agreed. Similarly, 63% of the nurses responded negatively to the suggestion of stopping teams, with 11% responding positively. Neutral responses to these statements included new CNAs and nurses who were not yet familiar with the effects or lack of effects of the empowered work teams.
|
|
Family member pretest and post-test questionnaire responses indicated that some measures of CNA performance were affected by the empowered work teams whereas others were not. We found three of four measures of direct care to be slightly higher at Time 2 than at Time 1 for the experimental group, whereas we found no differences for the comparison group (Table 4, third panel). The three measures that did improve included the following: spends time on resident needs (4.2 vs 4.4), checks on resident comfort (4.2 vs 4.4), and staff listens, talks, and cares (4.4 vs 4.5). However, when comparing the difference across time on these measures between the experimental and comparison groups, we found that the only significant difference was for staff listens, talks, and cares (Table 4, third panel, last column). Further, the treatment group effect in a MANOVA (not shown), including all four direct care variables, was not significant. Moreover, an examination of two additional performance measures, that is, satisfaction with care and satisfaction with staff friendliness, showed no differences between the experimental and comparison groups (Table 4, fourth panel).
Family members were also asked if residents were given an opportunity to make choices regarding the services provided (Table 4, fifth panel). Family members from the experimental nursing homes reported a significantly higher level of choice at Time 2 than at Time 1 regarding when to eat (2.5 vs 2.8) and when to shower (2.6 vs 2.9). Family members from the comparison nursing homes reported lower levels of choice at Time 2 than at Time 1 on these and the variable of when to sleep. The trend of more choice in the experimental group was significantly different from the trend of less choice in the comparison group (Table 4, fifth panel, last column). The result of a MANOVA that included the three choice variables confirms the findings (p =.005). It is reasonable to suspect that the increased choices available to residents in the experimental nursing homes were the result of the better knowledge that the CNAs had of the residents' preferences as noted from the qualitative or observation data.
Proposition 3 proposes that empowered work teams have a positive effect on job attitudes, including job satisfaction, commitment, and self-esteem. It is partially supported by qualitative data and unsupported by the quantitative data. In support of the proposition were observation data that showed that CNAs were able to distribute the work so that CNAs were more likely to do the kinds of tasks they most preferred and serve the particular residents they most enjoyed. For example, if one CNA preferred answering call lights (lights that residents turn on when they have a need) and another preferred assisting residents with eating, the CNAs were sometimes able to distribute their work so that each was doing what she or he preferred most. Similarly, CNAs were sometimes able to distribute the specific residents according to who preferred serving whom or to share equally in serving a resident whom no one wanted to serve.
In contrast, observations of meetings of empowered work teams suggested that the teams also have the potential to foster some negative attitudes. Some CNAs complained that the 30-minute weekly meetings kept them from completing their work on time. In other cases, when team meetings were held before or after a CNA's shift, some CNAs complained that they could not get to work early for the meetings or stay late for the meetings because they had other obligations such as taking a spouse to work or picking up children from school. Still other CNAs sometimes complained when one or more team members would disrupt the team meetings by repeatedly bringing up an issue that had already been discussed or bringing up a personal problem that had already been addressed. Finally, CNAs generally became dissatisfied in those cases when a DON forgot or neglected to read and respond to the CNAs weekly team notes.
Similarly, in general, the quantitative or survey data did not support the proposition that empowered work teams have a positive effect on job attitudes. The CNAs in the experimental nursing homes as well as the comparison nursing homes reported the same levels of general job satisfaction, burnout, and self-esteem at Time 1 and at Time 2 (Table 4, sixth panel). In one exception the experimental group reported higher satisfaction with scheduling at Time 2 than at Time 1, whereas we found no difference among the comparison CNAs. It is reasonable to suspect that this was the result of the empowered work teams' ability to influence scheduling decisions. In another exception, the comparison group reported lower commitment between Time 1 and Time 2 whereas the experimental group showed no change. However, neither the results of t tests comparing differences over time between the experimental and the comparison groups (Table 4, sixth panel, last column) nor the results of a MANOVA including all the job attitude variables (not shown) were statistically significant.
Proposition 4 proposes that empowered work teams reduce absenteeism and turnover. The proposition is partially supported. The qualitative or observation data showed that, during team meetings, CNAs sometimes discussed the problems that CNA absenteeism causes. This appeared to help those CNAs who were routinely absent to better understand the effects their absences were having on the other CNAs, and it appeared to result in their making more of an effort to be at work. In some cases, CNAs would report to the team that they were going to be absent and then another CNA would subsequently try to help them avoid the absence. For example, in one case a CNA reported that she would be absent the next day because her babysitter was not going to be available. A second CNA noted that her babysitter would not mind having additional children, and that the CNA could leave her kids with this babysitter. The subsequent result was that the CNA avoiding being absent.
When considering turnover, we again see that the observation data suggest some positive effects of the empowered work teams. On several occasions during empowered-work-team meetings, a CNA would comment that she would prefer to stay at the nursing home because the empowered work teams seemed to be helping and other nursing homes in the local area lacked such teams.
The observation data also uncovered some negative effects on turnover. In those cases in which a CNA was absent routinely, the CNA sometimes spoke of the need to quit because of her inability to avoid absences and of its subsequent negative effects on the other CNAs. In other cases, the empowered work team made an effort to have a CNA fired. This was found to occur when the CNA was clearly not getting the work done and this was having negative effects on resident care (e.g., when a CNA routinely allowed residents to sit in soiled, wet clothing for long periods of time rather than changing the clothing promptly).
When examining the quantitative or survey data, we found no differences in self-reported absenteeism between Time 1 and Time 2 for those in the experimental nursing homes (Table 4, bottom panel). Interestingly, self-reported absenteeism had a statistically significant increase in the comparison group (1.3 vs 1.6). However, although this increase among CNAs in the comparison nursing homes may seem to suggest that empowered work teams may help prevent absenteeism, it was not large enough to be significantly different from the stable trend observed over time for the experimental group (Table 4, bottom panel, last column).
In contrast, CNAs in experimental nursing homes were significantly less likely to quit or be terminated (Table 4, bottom panel, last column). Only 37% of the CNAs who were working in the experimental nursing homes at Time 1 were not working in these same nursing homes at Time 2. In contrast, at the comparison nursing homes, 51% were no longer working at the same nursing homes. However, we must view this difference with caution, because the average length of time between Time 1 and Time 2 was 1 month longer for the comparison nursing homes and consequently there was an additional month for CNAs to quit. Further, although the experimental nursing homes had a lower turnover rate, the data available do not allow us to determine whether there was a drop in turnover with the implementation of empowered work teams, because there is a lack of employment data prior to the study's beginning.
|
Discussion |
|---|
|
TOPAbstractMethodsFindingsDiscussionImplicationsReferences |
|---|
The generalizability of the study's findings is also affected by two other substantive limitations. First, because the study is based on a quasi-experimental design conducted in a single region of the United States, its findings may not apply to all nursing homes. Participating nursing homes volunteered to join the study and therefore may be somewhat different from the typical nursing home. CNAs, nurses, and family members were not randomly assigned to one nursing home or another, allowing for the possibility of differences between the groups beyond those already identified (Table 1). Finally, because the nursing homes are all located in North Texas, findings may be influenced to some extent by the uniqueness of the location itself. Second, the analytical approach results in the exclusion of missing data—only those who were present at both Times 1 and 2 were included in the analyses. Unfortunately, we were unable to survey those CNAs, nurses, and family members who were missing at Time 2 to obtain their attitudes. It is possible that the empowered work teams had a different effect on those who left the nursing home prior to the second survey. Further, it is reasonable to expect that there were baseline differences between those who chose to leave the nursing homes before Time 2 and those who stayed. We are currently undertaking a thorough analysis of these data in the hope of shedding light on the factors that contributed to CNA and nurse turnover; we will report the results of these analyses in a separate publication.
Given these limitations, we find that the qualitative or observation data suggest, in general, that the empowered work teams had positive effects on CNA empowerment and performance, possibly positive effects on absenteeism and turnover, and mixed effects on job attitudes. CNAs were more empowered after working in the teams because they were given new decision-making responsibilities, grew in competence at decision making, and experienced more positive impacts from their efforts. However, we also observed that although nurse management did involve the empowered work teams in decision making, this involvement could have been more routine.
The empowered work teams had positive impacts on CNA performance by allowing the CNAs to become more aware of resident health conditions, by providing them with more information on the special care needs of residents, by giving them the opportunity to question the poor performance of negligent team members, by giving them the time needed to clear up misinformation and communication, and because the team members were more willing to carry out decisions that they participated in making. Nevertheless, the observation data also suggest some negative effects of the teams on performance, related to the time spent in the meetings, which pulled CNAs away from their direct care duties. Moreover, the data also suggest that if nurse management members had been more consistent in their weekly communications and feedback to the teams, the positive effects that the empowered work teams had on CNA performance would likely have been larger.
It appears that team meetings may have contributed to reduced absenteeism and turnover. In some cases, team members were able to help member another avoid an absence, once her circumstances had been brought up during team meeting discussions. In addition, from some team members, who particularly liked the empowered work teams, there were comments that the teams were an important feature of the nursing home not offered by other facilities.
We found mixed results from the qualitative or observation data with regard to work attitudes. In some cases, positive effects were apparent when CNAs were given more opportunity to control who did what during the day and who served which residents. This allowed CNAs to do the kinds of CNA work they most preferred. However, on the negative side, CNAs sometimes expressed frustration at having to "hurry up" their direct-care duties in order to attend the team meetings. Furthermore, when meetings were scheduled before or after the CNAs' shifts, some CNAs experienced tension because of other, typically family, responsibilities that they had outside of work that conflicted with the meeting times.
When examining the quantitative or survey data, we see that the findings also show statistically significant positive effects of the empowered work teams on CNA empowerment and performance, although the substantive effects are small. We found no effects on job attitudes or self-reported absenteeism. In addition, although the quantitative or survey data showed no effect of the empowered work teams on CNAs' intent to quit, there is some evidence that turnover itself may have been reduced.
|
Implications |
|---|
|
TOPAbstractMethodsFindingsDiscussionImplicationsReferences |
|---|
For those nursing home administrators and nurse managers who want to implement empowered work teams, the first step should be reading about the uses of and strategies for implementing them. In this regard, we are preparing a "why and how to" book that shares the many lessons learned in the process of implementing empowered work teams and includes orientation materials for nursing home administrators, nurse management, CNAs, and team facilitators (Yeatts, Cready, & Noelker, in production). The next step should include the observation of existing empowered work teams or discussions with those nurse managers who have experience working with the teams. In this regard, success will depend greatly on nurse managers' understanding of what is involved and what they will be responsible for doing, and likewise their desire to create and support empowered work teams. A subsequent step is the orientation of all nursing home administrative and nursing staff, including the CNAs and team facilitators, to the purposes and advantages of empowered work teams. Further, it is important that the CNAs be given the opportunity to learn how to work together in a team meeting, to sometimes make mistakes, and to routinely contribute to the management decisions made about their work.
|
Footnotes |
|---|
1 Department of Sociology, University of North Texas, Denton, TX.
Decision Editor: Jeanne Teresi, EdD, PhD
Received for publication May 12, 2006. Accepted for publication January 8, 2007.
|
References |
|---|
|
TOPAbstractMethodsFindingsDiscussionImplicationsReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. E. Bishop, D. B. Weinberg, W. Leutz, A. Dossa, S. G. Pfefferle, and R. M. Zincavage Nursing Assistants' Job Commitment: Effect of Nursing Home Organizational Factors and Impact on Resident Well-Being Gerontologist, July 1, 2008; 48(suppl_1): 36 - 45. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
