Differences in Assisted Living Staff Perceptions, Experiences, and Attitudes

Abstract

Research within residential care/assisted living (RC/AL) settings has shown that the attitudes of personal care (PC) staff towards their organization, and its residents and families, can affect the quality of resident care. This paper describes the perceptions, experiences, and attitudes of PC staff and their supervisors and considers these data in the context of non-hierarchical staffing patterns – a philosophically expected, yet unproven tenet of RC/AL. Using data collected from 18 RC/AL communities, these analyses compared the characteristics, perceptions, experiences, and attitudes of PC staff (N=250) and supervisors (N=30). Compared to supervisors, PC staff reported greater burden, frustration, depersonalization, hassles, and feeling significantly more controlling of, and less in partnership with, families (p<0.05). Because the PC staff experience is crucial for their and resident outcomes, more work is needed to create a work environment where PC staff are less burdened and have better attitudes towards work and families.

Introduction

Residential care/assisted living (RC/AL) is a long-term care option that provides housing and services for close to a million older adults in the United States (Park-Lee et al., 2011), and next to nursing homes, is the largest provider of residential long-term care (Polzer, 2010). RC/AL is an attractive alternative to nursing homes because it is designed to be more home-like and to emphasize the resident as an individual who deserves autonomy, privacy, independence, and consistent engagement with staff empowered to provide for resident needs (Center for Excellence in Assisted Living, 2010; Fazio, 2008; Talerico, O’Brien, & Swafford, 2003).

In RC/AL settings, personal care (PC) staff typically take care of residents’ daily needs including personal hygiene, housekeeping, meals, and assisting with medication administration (Chou & Robert, 2008). The retention of PC staff is one of the biggest challenges to quality of care in long-term care, in that turnover disrupts continuity of resident care, creates burden for other staff, and incurs costs in hiring and training new staff (Sikorska-Simmons, 2005). Given these consequences, a number of studies - conducted primarily in nursing homes (NHs) – have examined factors that contribute to staff retention and staff turnover (Angelelli, Gifford, Shah, & Mor, 2001; Castle, 2001, 2005; Castle & Engberg, 2005; Castle & Lin, 2010; Fitzpartick, 2002). For one, staff relationships with families affect not only the quality of resident care provided, but also job satisfaction and turnover (Lerner, Resnick, Galik, & Flynn, 2011). PC staff attitudes, such as burden, work stress, and hassles, can also cause staff to be less satisfied with their jobs and thus more likely to leave. Protective factors, such as knowing a resident well, being better trained, maintaining a resident’s independence, and having a family member follow recommendations, help staff deal with day-to-day challenges and are associated with higher rates of staff satisfaction and staff reporting that they intend to stay in their position (Deveraux, Hastings, Noone, Firth, & Totsika, 2009; Lerner et al., 2011; Zimmerman et al., 2005). Although more limited in scope and quantity, research conducted within RC/AL settings has yielded similar findings as the work done in NHs; PC staff attitudes towards their organization, the residents they care for, and residents’ families, can affect the quality of the care they provide to residents (Aud & Rantz, 2004; Maas & Buckwalter, 2006).

The attitudes of supervisors, and especially those towards their care and management duties, affects the care PC staff provide and residents outcomes (Anderson, Issel, & McDaniel Jr., 2003; Barry, Brannon, & Mor, 2005). Previous work in nursing homes has shown that supervisors who are flexible, responsive, and collaborative engender work environments characterized by high teamwork and shared decision making among all staff types (Tellis-Nayak, 2007). Not surprisingly, other work has found that nursing homes with low teamwork have less interaction among staff and more animosity between supervisors and PC staff (Scott-Cawiezell et al., 2004; Tyler & Parker, 2011). Thus, it seems that both the structure and function of the nursing home team are crucial to a high functioning work environment.

In addition to the quality of the team, the quality of the relationship between supervisors and PC staff can also affect staff and resident outcomes. For example, in settings where supervisors are empathetic, reliable and focus on connecting with staff PC staff have lower job stress and higher job satisfaction (Chou & Robert, 2008; McGilton, McGillis Hall, Wodchis, & Petroz, 2007). Similarly, in settings where staff share similar attitudes, including a common understanding of the work they do, mutual respect, and shared goals, staff report better job satisfaction and residents report better quality of life (Gittell, Weinberg, Pfefferle, & Bishop, 2008) and greater satisfaction with care (Sikorska-Simmons, 2006). Similarly, NHs with a flattened staffing hierarchy that includes open communication, shared decision making, and relationship oriented leadership have lower rates of restraint use, and residents exhibit fewer aggressive and disruptive behaviors, and complications from immobility (Anderson et al., 2003). Evidence suggests that a more equitable working environment provides the opportunity for better relationships between staff and residents and family. In a study of the long-term care work environment, the staff working in organizations with a more person centered management approach had more positive attitudes towards management and these attitudes correlated with families having higher ratings of satisfaction and care quality (Tellis-Nayak, 2007). These findings suggest that if this structure has been effectively translated into practice, one would expect PC staff and supervisors have similar perceptions about work, experiences of burden and stress, and attitudes towards families and their co-workers (Barry et al., 2005; Stone et al., 2002).

This paper explores the experiences, perceptions, and attitudes of staff in RC/AL communities, comparing those held by PC staff to their supervisors. The unique contribution of this paper is that it examines staff perceptions, experiences, and attitudes (both towards their work and towards families) in RC/AL communities, a setting that has been understudied. This paper concludes with a discussion of the implications of this work on future research, RC/AL organizational and care practices, and gerontological nursing.

Design and Methods

The data for these analyses were derived from RC/AL staff members who participated in Families Matter, a group randomized controlled trial conducted in North Carolina. This trial involved data collection at both baseline and six-month follow-up; the data for these analyses are based only on those collected at baseline. Data were obtained from 18 RC/AL communities, which each received monetary reimbursement for participation so as to defray the cost of staff time incurred by study procedures. The Institutional Review Boards of the XXX and XXXX reviewed and approved all study materials and procedures.

Sample

The Families Matter study consisted of a sample of approximately 20 residents from each long-term care setting and the staff members who were most familiar with the selected residents. In addition to being familiar with a participating resident, eligible staff members were at least 18 years old, worked at least 20 hours/week, and were employed by the setting for at least one month prior to study. All eligible staff members were approached in person, provided details about the study and its requirements, and asked to provide written informed consent prior to participation. Participation comprised a 20-minute in-person interview. As part of this interview, staff participants reported their job title and position, which were categorized as supervisory or PC during analyses. Job titles and positions categorized as supervisors included administrators, business managers, activity directors, life enrichment coordinators, health and wellness coordinators, licensed practical nurses, and other supervisors. Job titles and positions categorized as PC staff included CNAs, medication technicians, and unlicensed care assistants. Of note, in [name of state], RC/AL settings are not required to have a registered nurse on-site, and thus there were few to recruit for this study. Further, when registered nurses are on-site, they typically act in supervisory rather than direct care roles. Because this study was primarily targeted at direct caregivers who interact with families and residents, in most cases, the inclusion of RNs was inappropriate. Indeed, two RNs were enrolled in the study, however because they numbered too few, were excluded from analyses.

Measures

Administrators from the participating RC/AL communities provided information about the community characteristics including profit status, years in operation, number of beds, occupancy rate, number of administrators in the past 3 years, staffing, monthly charges, affiliation and percent of residents with dementia, of minority race, and receiving Medicaid.

Staff characteristics

Staff participants provided information about their own demographic characteristics, health status, perceptions about and attitudes towards the caregiving role, families, and residents; and experiences. To measure overall health, staff were asked “In general, would you say your health is: excellent, very good, good, fair, or poor?” To assess depression, the 10 item Center for Epidemiologic Studies Depression Scale was used to identify the absence or presence of depressive symptoms in the past month (Radloff, 1977). Work history included the number of years working in that setting, years of long-term care experience, hours worked in a typical week, and whether the work role included personal care or other functions (e.g., supervisor, coordinator, director, or administrator).

Staff perceptions, experiences, and attitudes

Staff perceptions were evaluated using the Staff Perception of the Caregiving Role Instrument -- a 78 item self-report measure with four subscale measures of task burden (α = .61–.84), frustration (α = .70–.82), dominion (control in relation to family members; α = .64–.71), and exclusion of families (α = .70) (Maas & Buckwalter, 1990; Maas et al., 2004; Specht et al., 2005). Staff perceptions were also examined using the Family Behaviors and Family Empathy Scales (α = .55) which ask staff their perceptions of how families behave towards them, how well family members understand their job, and are sensitive to their feelings (Pillemer et al., 2003).

Staff experiences were measured using the 22 item self-report Maslach Burnout Inventory that includes three sub-scales measuring emotional exhaustion (α = .90), depersonalization (α = .79), and lack of personal accomplishment (α = .71) (Maslach, Jackson, & Leiter, 1996); hassles and uplifts measured by the Hassles and Uplifts Scale (Elder, Wollin, Hartel, Spencer, & Sanderson, 2003); work stressors measured by the Work Stress Inventory subscale related to caring for residents (α = .82) (Schaefer & Moos, 1993, 1996); and interpersonal conflict measured using the Interpersonal Conflict Scale (α = .79), which asks how frequently staff have conflict with family members regarding resident care tasks (Pillemer et al., 2003). Staff attitudes towards their jobs and residents’ families were assessed with the 16 item self-report Attitudes Towards Family Checklist (α = .70–.91) which includes three subscales: families cause disruption (α = .56–.64); partnership with family (α =.58–.63); and family relevance (Maas & Buckwalter, 1990; Maas et al., 2004). The Staff Perceptions of Caregiving Role and the Attitudes toward Families Checklist, were developed for use with staff from special care units for persons with dementia, rather than RC/AL. Still, because the majority of RC/AL residents have some cognitive impairment, it is likely that the staff share similar experiences and the measures are similarly valid (Magsi & Malloy, 2005).

Analyses

Descriptive statistics related to RC/AL communities (means, standard deviations, frequency counts, and percentages) were generated using SPSS version 16.0. Because of the clustering of staff within communities, linear and nonlinear mixed models were used in analyses of differences between staff types, as appropriate to the measure. The mixed models specified a random effect for setting and a fixed effect for staff type. Models were also run adjusting for staff race and educational level. All mixed models analyses were completed using SAS software, version 9.2 of the SAS System for Windows.

Results

Table 1 presents the characteristics of the 18 RC/AL communities participating in this project. All were for-profit; they had been in operation an average of 8.3 years (SD 4.7), had an average bed size of 81.7 (SD 26.8) beds, an occupancy rate of 83.3% (SD 14.7), and an average monthly charge of $3,095 (SD $722). Eight (44%) of communities reported having three or more administrators in the past three years. Three communities (17%) were affiliated with a continuing care retirement community, twelve (67%) were affiliated with another RC/AL community, and two (11%) were affiliated with a nursing home. Close to half of the residents had a diagnosis of dementia (48.4%, SD 31.8) and 22.6% were racial minorities (SD 27.9).

Table 1.

Characteristics of RC/AL communities (n=18)

	N (%) or Mean (SD)
For-profit (n, %)	18 (100)
Years in operation (mean, SD)	8.3 (4.7)
Number of beds (mean, SD)	81.7 (26.8)
Occupancy rate (mean, SD)	83.3 (14.7)
Average monthly charge (mean, SD)	$3,095 ($722)
Three or more administrators in past 3 years (n, %)	8 (44)
Affiliated with a continuing care retirement community (n, %)	3 (17)
Affiliated with another RC/AL living community (n, %)	12 (67)
Affiliated with a nursing home (n, %)	2 (11)
Percent of residents with dementia diagnosis (mean %, SD)	48.4 (31.8)
Percent of residents with minority race (mean %, SD)	22.6 (27.9)

A total of 280 data (250 categorized as PC staff and 30 categorized as supervisors) provided data for these analyses. Table 2 describes and compares the PC staff and supervisors. Regardless the classification, the sample was overwhelmingly female (96% of PC staff and 93% supervisors), but differed in age (PC staff 37.6 years, supervisors 43.7 years; p<0.05), race (74% of PC staff and 27% of supervisors were minorities; p<0.001), and education level (8% of PC staff and 30% of supervisors held a Bachelor’s degree or higher; p<0.001). PC staff worked in the setting an average of 2.7 years versus 4 years for supervisors (p<0.05), but the two groups did not differ in overall years of long-term care experience (PC staff 6.9 years versus supervisors 7.4 years). PC staff worked fewer hours each week (36.8) than did supervisors (40.2; p<0.01).

Table 2.

Characteristics, perceptions, experiences, and attitudes of personal care staff and supervisors

CHARACTERISTICS	N (%) or Mean (SD)
	Personal Care Staff (n=250)	Supervisors (n=30)	p
Age (mean, SD)	37.6 (13.2)	43.7 (12.1)	.013
Gender (n, % female)	240 (96)	28 (93)	.65
Marital status (n, % married)	100 (40)	14 (47)	.50
Minority race (n, % minority)	184 (74)	8 (27)	<.001
Hispanic	6 (2)	0 (0)	.73
Education level (n,%)
High school or less	118 (47)	6 (20)	<.001
Some college or Associate’s degree	113 (45)	15(50)
Bachelor’s degree or higher	19 (8)	9 (30)
Health is excellent	47 (19)	10 (33)	.07
Depression (0–10; lower is better)	2.6 (2.1)	2.5 (2.0)	.80
Years working in that setting (mean, SD)	2.7 (3.3)	4.0 (3.3)	.032
Years of long-term care experience (mean, SD)	6.9 (6.7)	7.4 (7.0)	.69
Hours worked in typical week (mean, SD)	36.8 (6.0)	40.2 (7.3)	.005
PERCEPTIONS			p Unadjusted	p Adjusted1
Perceptions of the caregiving role
Burden (1–5; lower is better)	2.2 (0.5)	1.9 (0.5)	<.001	.002
Frustration (1–5; lower is better)	3.7 (0.5)	3.4 (0.5)	.05	.07
Dominion (1–5; lower is better)	3.2 (0.5)	2.7 (0.5)	<.001	.004
Exclusion (1–5; lower is better)	2.1 (0.4)	2.1 (0.4)	.63	.88
Perception of family empathy towards staff (3–15; higher is better)	8.5 (2.4)	9.2 (1.8)	.15	.14
EXPERIENCES
Maslach Burnout
Emotional exhaustion (0–54; lower is better)	13.3 (11.4)	10.9 (8.9)	.31	.26
Depersonalization (0–30; lower is better)	2.1 (3.1)	0.8 (1.4)	.035	.038
Lack of personal accomplishment (0–48; higher is better)	39.9 (7.6)	42.5 (4.9)	.09	.048
Work Stress Inventory, caring for residents (4–20; lower is better)	11.0 (3.6)	11.1 (3.3)	.73	.85
Hassles (17–85; higher is better)	45.4 (9.7)	42.9 (7.2)	.09	.06
Uplifts (20–100; higher is better)	87.1 (10.2)	86.7 (9.5)	.64	.71
Interpersonal conflict - disagreements (7–35; lower is better)	11.9 (5.6)	10.7 (4.3)	.23	.08
ATTITUDES
Attitude towards job (1–4; higher is better)	3.3 (0.8)	3.6 (0.6)	.11	.18
Attitudes towards families - partnership with family (1–5; higher is better)	3.9 (0.5)	4.1 (0.3)	.013	.028
Attitudes towards families - families cause disruption (1–5; higher is better)	3.4 (0.5)	3.8 (0.4)	.002	.014
Attitudes towards families - family relevance (1–5; higher is better)	3.0 (0.6)	3.3 (0.5)	.024	.025

Note: Linear and nonlinear mixed models used to adjust for clustering within communities when testing for statistically significant differences between staff types.

¹Adjusted for educational level and minority race.

Table 2 also presents the unadjusted and adjusted differences between PC staff and supervisors on perceptions, experiences, and attitudes. The completeness of these measures was high with no single item of any of the measures having more than one missing response. In adjusted analyses, PC staff reported being significantly more burdened (p<.01) and controlling in relation to families (dominion; p<.01) than supervisors. They also reported more burnout related to depersonalization (p<0.05) and accomplishment (p<0.05). Finally, PC staff reported significantly lower scores on the attitude scales than supervisors, including partnership with families, families cause disruption, and family relevance (p<0.05). Additional adjustments for the setting characteristics of percent of residents with dementia, staff-to-resident ratio, size (total beds), and percent of residents receiving Medicaid/public assistance had no substantive effect on the results of the analyses.

Discussion

The purpose of this study was to compare RC/AL PC staff and supervisors in terms of their perceptions, experiences, and attitudes. Not unexpectedly, PC staff and supervisors differed by demographic characteristics including age, race, and education level. PC staff were more likely to be younger, racial minorities, and less educated. These differences reflect the nature of the long-term care setting and workforce – supervisors tend to be higher educated and non-minorities, while PC staff are primarily middle aged, minority women with at least a high school education (Bureau of Labor Statistics, 2010).

Even after adjusting for differences in race and educational attainment, there were numerous differences between PC staff and supervisors. Overall, PC staff had poorer perceptions, experiences, and attitudes towards their jobs. PC staff perceived their work as being more burdensome and were less willing to grant families control over resident care. At the same time, when compared to supervisors, PC staff reported that they experienced more depersonalization and felt less accomplished. These findings suggest that the PC staff may benefit from a more supportive environment that emphasizes team work and allows for participation in decision making. Research examining staff-supportive cultures suggests that RC/AL communities that value teamwork, PC staff empowerment, and shared decision-making have more organizational commitment, less turnover, and better quality of care (Sikorska-Simmons, 2008). Future work focused on identifying and relieving the specific burdens PC staff experience may improve staff attitudes toward their work and also their relationships with families.

Even though the relationship between residents’ families and PC staff are a key interface essential for day-to-day RC/AL resident care (Gaugler & Ewen, 2005), PC staff had more negative attitudes towards families than supervisors. PC staff had poorer scores on the three Attitudes Towards Families subscales (partnership with family, families cause disruption, and families relevant subscales). A potential explanation for this difference may be the nature of the PC staff work, meaning that while providing hands on care to residents, they, more so than supervisors, come into contact with families, or are more often aware when families are not present. These situations may engender conflict, either in their own right, or because PC staff are not comfortable interacting with families, or perhaps because families are not as supportive as desired. Our study and the existing literature suggest that the attitudes of PC staff towards families warrants more detailed investigation to examine not only staff attitudes, but factors that may influence these attitudes, such as organizational culture, supervisory support, and the relationship between RC/AL residents and families (Gaugler & Ewen, 2005; Maas & Buckwalter, 2006; McGilton et al., 2007; Sikorska-Simmons, 2005). Understanding the factors that influence PC staff attitudes could help focus efforts aimed at improving PC staff satisfaction with their work.

RC/AL is often viewed as preferable to nursing homes because it provides a home-like environment, presumably emphasizing choice, independence, and connection to a larger community (Center for Excellence in Assisted Living, 2010; Fazio, 2008; Talerico et al., 2003). The RC/AL setting is further presumed to empower PC staff to focus on individual resident needs and thereby achieve more of an equal partnership between supervisors and PC staff than evidenced in traditional nursing homes (Center for Excellence in Assisted Living, 2010). In this study, PC staff reported significantly poorer attitudes on the burden, dominion, burnout, and attitudes towards family scales. The differences observed between PC staff and supervisors are inconsistent with what would be expected from a less hierarchical structure where perceptions, experiences, and attitudes should be more similar. On the other hand, it must also be considered that these findings may in fact be consistent with what would be expected in a setting wherein PC staff do have control, and are empowered to make decisions, but there are poor relations with families nonetheless or even in consequence. Thus, this finding suggests a need for further research examining the leadership structure in place and the relationships among supervisors, PC staff, and families. Future research observing contextual factors such as enactment of the philosophy and mission of care, may provide insight into the factors that may influence these relationship including the organizational structure (whether they have a person-centered or hierarchical focus), staff attitudes, and resident/family/staff outcomes.

This analysis was limited because the cross-sectional nature of the data did not examine the consequences of staff perceptions, experiences and attitudes, such as staff turnover and residents care outcomes. The analyses for this paper were derived from baseline data from an intervention study and did not focus on collecting data about the flattened or hierarchical nature of the organizational structure itself. Regardless these limitations, these findings have implication for future research efforts and gerontological nursing practice. Overall, PC staff had more negative perceptions of their work were more burdened, and had more negative attitudes towards residents’ families than supervisors. This highlights a need for a more concentrated effort to create an environment that decreases the stresses and burdens experienced by PC staff. Prior work aimed at improving the work environment in nursing homes has found that efforts such as consistent assignment, improving employee benefits, and interventions to improve communication between staff and family can affect staff work stress and staff retention (Advancing Excellence in America’s Nursing Homes, 2009; Pillemer et al., 2003). Adapting these intervention efforts for staff in RC/AL may provide positive outcomes. In particular, focusing policy and management efforts on providing routine assessment of PC staff perceptions, experiences, and attitudes, may inform programs to improve PC staff stress and burden, as well as PC staff and family relationships.

Clinical Implications

Currently, there is a wide array of roles a nurse can enact in RC/AL. Although fewer than 50% of states currently require a nurse to be involved in RC/AL care, between 47–70% of these settings employ an RN or LPN (Maas & Buckwalter, 2006; Mitty et al., 2010). RNs may oversee care at a single RC/AL site, or may oversee care for multiple settings, while LPNs may coordinate clinical care as health care supervisor or case manager (Mitty et al., 2010). For settings that do not employ nurses, nurses who provide care to RC/AL residents may be employed by outside agencies such as home health or hospice (Park-Lee et al., 2011; Stearns et al., 2007; Zimmerman et al., 2003). Given the various roles nurses can have in overseeing the care provided to RC/AL residents, it is important that they critically examine the care that is provided to these residents, and understand the factors that may influence resident outcomes, such as staff relationships with residents and families. Nurses employed in RC/AL may be in the position to implement interventions to improve staff attitudes and experiences. For nurses who provide care to RC/AL residents but may not necessarily be employed by the setting, it is important for to understand the staff experiences and attitudes that may be facilitators or barriers to providing high quality care.