Getting Along in Assisted Living: Quality of Relationships Between Family Members and Staff

Abstract

Background and Objectives

Assisted living facilities (ALFs) have quickly expanded as an alternative to nursing homes. Research on nursing homes has revealed problems in relationships between family members and staff. However, little is known about these relationships within ALFs. The purpose of the current study was to examine the prevalence of conflict and positive and negative interactions from the perspective of both family members and staff and to examine the effects of positive and negative aspects of the relationship on salient staff and family outcomes in ALFs.

Research Design and Methods

Data were collected from 252 family members and 472 staff members across 20 ALFs who participated in the Partners in Care in Assisted Living study. Participants completed measures including interpersonal conflict, depressive symptoms, perception of treatment, and stress related to caregiving.

Results

Conflict among family and staff members was found to be relatively low. For staff, interpersonal conflict and treatment by family members significantly predicted burnout and depressive symptoms. For families, only female gender significantly predicted burden. Subgroup analyses, however, indicated that the effect of interpersonal conflict was significantly associated with perceived caregiver burden among family members whose relative had dementia.

Discussion and Implications

Despite the relatively harmonious relationships among family–staff in ALFs, sources of conflict and negative interactions were identified, revealing the importance of collaborative relationships and the influence these relationships have on both family and staff outcomes. These findings can inform intervention efforts to improve family–staff interactions within ALFs.

As people are living longer with chronic illness and disability, the need for long-term residential care facilities continues to rise. Despite the critical role residential facilities play in providing a range of care to the growing older adult population, these institutions tend to be under-resourced (Barken & Lowndes, 2018) and struggle with staff dissatisfaction and turnover (Kemp, Ball, Perkins, Hollingsworth, & Lepore, 2009). In recent decades, the movement toward relationship-centered care in long-term facilities has emphasized the need to establish close and collaborative relationships not only between staff and residents, but also between family members and staff. Improving the family–staff relationship is a potential mechanism to foster improved coordination of care and, in turn, optimal quality of life and quality of care for residents (Beach, Inui, & Relationship‐Centered Care Research Network, 2006; Koren, 2010; Puurveen, Baumbusch, & Ghandi, 2018; Robison et al., 2007).

Most of the literature examining family–staff interactions has been limited to nursing homes (Kemp et al., 2009; Pillemer, Schultz, Cope, Meador, & Henderson, 2018). However, due to the high demand that exists for residential care, assisted living facilities (ALFs) are growing as an alternative to nursing homes (Koren, 2010; Silver, Grabowski, Gozalo, Dosa, & Thomas, 2018). As of 2015, approximately 1.4 million individuals were residing in nursing homes, and close to 1 million lived in ALFs (Harris-Kojetin et al., 2019). In recent decades, ALFs have evolved to provide a comprehensive continuum of care to their residents. For example, approximately 42% of ALF residents are living with dementia, and ALFs have adapted to now provide specialized dementia care services for this population (Caffrey & Sengupta, 2018; Zimmerman, Sloane, & Reed, 2014). Additionally, research examining family involvement in ALFs and nursing home settings has found that caregivers with a relative in ALFs report greater involvement in care (e.g., monitoring resident health and well-being) and higher levels of perceived burden compared with caregivers of residents in nursing homes (Port et al., 2005). Although ALFs continue to expand, they have remained under-researched, and the comparability of the nature of family–staff dynamics to nursing home settings remains unknown. Given the high level of involvement of family caregivers in ALFs and the likely frequent interactions they have with staff, additional research is needed to examine the interpersonal dynamics among family members and staff within these settings.

Conceptual Framework

Gittell’s (2005) theory of relational coordination suggests that effective coordination between families and staff will depend on frequent and high-quality interactions characterized by respect, shared goals, and shared knowledge, placing emphasis on families as members of the primary care team. These characteristics are particularly important when residents have conditions, such as dementia, which result in increasing impairment, dependency, and communication difficulties. However, both theoretical and empirical work examining the relationships between families and staff in nursing homes indicate that structural barriers can impede the formation of collaborative relationships between the two groups.

Litwak’s (1985) influential theoretical approach highlights the fundamental differences between primary groups and secondary groups. Primary groups, such as families, are categorized as emotional, socially intimate, and involve long periods of interaction. Alternatively, secondary groups, such as large-scale formal organizations, are characterized by bureaucratic structure, formal behavioral rules, technical knowledge, and impersonal ties. In the context of transitioning individuals to long-term care, secondary groups tend to assume responsibilities that were previously held by members of primary groups. However, formal organizations, such as residential care facilities, are limited in their capacity to manage care tasks across the many individuals they are responsible for, and care delivery by secondary groups is accomplished through task simplification (e.g., routinizing feeding and bathing processes). Thus, a process of “dual specialization” is suggested by the theory, in which family members provide psychosocial support whereas staff provide routine hands-on care (Gaugler, 2005).

This theoretical framework suggests that problems can emerge in ALFs when there is a mismatch between the structure of the formal organization and the types of tasks it seeks to take over from families. In ALFs, there is increased potential for such interpersonal conflict because these institutions represent a classic example of a formal organization seeking to take over primary care tasks, while fitting such tasks into a bureaucratic, routinized, organizational framework (Litwak, 1985; Litwak, Jessop, & Moulton, 1994). These factors can, in turn, impede the establishment of positive interpersonal dynamics among family members and staff, which can have negative impacts on the outcomes of families, staff, and residents. Therefore, the current study seeks to apply this theoretical approach to identify and examine interpersonal relationships between family members and staff in ALFs.

Family–Staff Relationships and Outcomes in Long-Term Care

In long-term care, family members are important contributors to the quality of care a resident receives, and they often remain actively involved in many aspects of care following placement (Bauer, 2006; Gaugler & Kane, 2007; Roberts & Ishler, 2018). However, family members can also experience feelings of guilt related to their decision to relocate the care recipient (Majerovitz, Mollot, & Rudder, 2009; Martz & Morse, 2017). Furthermore, although residential placement may relieve family members of direct caregiving responsibilities, feelings of burden have been shown to persist following placement (Majoerovitz et al., 2009), and families may find it difficult to surrender control over the provision of care.

Family members serve as key members of the resident’s long-term care team and can provide valuable insights into their loved one’s needs. Additionally, when asked to rate the quality of the resident’s care, family members have reported quality of family–staff relationships (in addition to security and medical care) to be a key determinant of care quality (Wodchis, Kwong, & Murray, 2015). However, institutional-level restrictions (e.g., heavy workload, scarce resources, inadequate staffing) can be a significant barrier to the formation of positive and collaborative family–staff relationships (Barken & Lowndes, 2018; Baumbusch & Phinney, 2014; Haesler, Bauer, & Nay, 2010). Moreover, as Litwak’s framework suggests, a lack of clarity in the expectations of relationships between family members and staff, and ambiguity around the level of involvement that family members should take in their loved one’s care, can create feelings of conflict and tension among family members and staff (Bauer, 2006; Barken & Lowndes, 2018).

In addition to imposing stress on families, poor relational dynamics between families and staff have been found to contribute to the persisting issues of job-related burnout and high rates of staff turnover in nursing home settings (Abrahamson, Suitor, & Pillemer, 2009; Majerovitz et al., 2009). The importance of these relationship issues on work-related outcomes is not surprising, given the frequency of contact and interaction between family members and direct care workers.

Eldercare professionals are tasked with heavy workloads and are required to provide comprehensive care to a range of individuals with varying illnesses and levels of functional dependency (Gallego-Alberto et al., 2018), and burnout is quite common among staff within these facilities. Burnout is defined as an adverse response to chronic interpersonal job stressors, where job-related demands exceed an individuals’ perceptions of their capacity to cope (Edvardsson, Sandman, Nay, & Karlsson, 2009; Woodhead, Northrop, & Edelstein, 2016). Burnout has been shown to negatively affect staff members’ mental and physical health, compromise the quality of care provided to residents, impose additional burden and workload on other staff, and increase costs associated with staff turnover, such as hiring and training new employees (Beeber et al., 2014; Woodhead et al., 2016).

Studies focused on ALFs have reported that, consistent with rates in nursing homes, approximately 40% of direct care workers reported symptoms of depression and anxiety (McKenzie et al., 2011), and a study in nursing homes found that negative family–staff relationships were associated with increased levels of anxiety in staff (Gallego-Alberto et al., 2018). Furthermore, positive family–staff communication has been shown to have an impact on the health and well-being of family members and residents and reduces levels of burnout and staff turnover (Majerovitz et al., 2009). Robison and colleagues (2007) found that after exposing family and staff in dementia special care units to an intervention designed to enhance communication and problem-solving, families experienced significant improvements in communication with staff, and staff reported reduced family conflict and lower levels of depression. These findings highlight the importance of positive relationships on the outcomes of both groups.

Research that examines the causes and consequences of family–staff relationships in ALFs, however, remains scarce. Given the degree to which ALFs are caring for more physically and mentally impaired residents, similar to the care provided to residents in nursing homes, there is reason to hypothesize that similar relational dynamics, such as sources of conflict and perception of others’ behaviors, may also exist in this care setting. Therefore, the current study tests several hypotheses reflected in the analyses conducted. Analysis 1 examines the prevalence of conflict and positive and negative interactions from the perspective of both family members and staff. Although we expect that both groups will report conflict and relationship difficulties, we predict that staff will experience higher conflict and negative interactions, and fewer positive interactions, compared with families. The reason for this pattern is that staff have the opportunity to interact with more family members than the reverse and thus have more opportunities for negative communication events.

Analysis 2 uses cross-sectional data to examine the degree to which reported positive and negative aspects of the relationship contribute to family and staff outcomes. In both groups, depressive symptoms are used as an indicator of psychological well-being. Two different but comparable scales measure perceived stressful outcomes of the caregiving experience. For staff, burnout was measured; for family caregivers, the measure was a caregiver burden scale.

In Analysis 2, we test several hypotheses. Hypothesis 1 proposes that across families and staff, interpersonal conflict and perceptions of family treatment will predict burnout among staff, caregiver burden among families, and depressive symptoms in both groups. Hypothesis 2 proposes that interpersonal conflict and perceptions of family treatment will be stronger predictors of burnout and depressive symptoms in direct care staff compared with those in other job positions (e.g., housekeeping staff or dining assistants). We base this hypothesis on the greater frequency of contact direct care staff (e.g., resident assistants) have with family members around the care of residents.

Hypothesis 3 proposes that in the family member sample, conflict and treatment by staff will be stronger predictors of burden and depressive symptoms in individuals whose relative has dementia compared with those with residents without dementia. In such cases, family members must rely more heavily on staff for information about the residents, who are often unable to report for themselves. Furthermore, the behavioral symptoms frequently manifested by persons with dementia may lead to stressful interactions with staff about care (Robison et al., 2007).

Methods

Participants and Procedure

We employed data collected from the Partners in Care in Assisted Living (PICAL) study (Pillemer et al., 2018), a randomized, controlled intervention to improve communication between family members and staff in ALFs. The purpose of the intervention was to provide staff and family members with educational, interactive training sessions on effective techniques (e.g., active listening skills and guidelines for dealing with conflict) to foster cooperative communication among family members and staff.

The PICAL intervention study included 20 ALFs across eight states. The states were chosen to be broadly representative of regions of the country and included Colorado, Florida, Michigan, Ohio, Illinois, New York, Connecticut, and Pennsylvania. Participating communities were purposively selected from two national senior living providers based on their interest in the program and perceived capacity to engage in an evaluation research project. All communities had at least 50 beds and were able to assign a staff person to work with the research team.

In the present study, we use only Time 1 data because of the likelihood that the intervention affected variables of interest (Pillemer et al., 2018). Data were collected via an online survey sent to respondents; for those unable or unwilling to complete the survey online, a paper copy was made available. Across the 20 facilities, approximately 40% of eligible staff responded to the survey. In terms of families, the individual listed as the responsible relative or caregiver was invited to participate in the survey, and approximately 15% did so.

The current study includes 252 family members between the ages of 45 and 87 years and 432 ALF staff members between the ages of 18 and 79. Both family and staff respondents were primarily female. Almost all family members were white, whereas staff members were more diverse (76% white). Additionally, many ALFs are relatively small, and many staff interact with family members in these facilities. We therefore did not restrict the survey to direct care staff only, but included all staff members who have contact with family members. The largest staff category was direct care staff, such as resident assistants, nursing assistants, and medication administration aides (58.5%). A fifth of the respondents (20.5%) were in administrative positions, such as receptionists, administrators, assistant administrators, and dining managers. The remainder of the staff were in dining, housekeeping, and maintenance. Characteristics of staff and family members included in the current study are presented in Table 1. Cronbach’s alphas were calculated to determine reliability for each of the measures in the current sample.

Table 1.

Demographic Characteristics for Staff (n = 432) and Family Members (n = 252)

	Staff			Family
Sociodemographic variable	%	M	SD	%	M	SD
Age		40.47	14.8		65.17	7.05
Sex (female)	87.5			67.8
Race/ethnicity
White	76			97.4
Black or African American	14.8			0.4
Hispanic or Latino	6.8			0.4
Asian/Pacific Islander	1.7			0.4
Native American	0.2			0.4
Other	0.5			0.9
Highest level of education
8th grade or less	0.5			0
Some high school	4			0
High school diploma	40.8			6.9
Some college	22.8			16.7
College degree	22.2			41.2
Postgraduate	0.9			32.6
Job title				n/a
Direct care staff	58.5
Business administration	17.5
Dining	14.3
Housekeeping	6.9
Maintenance	2.7
Length of employment in assisted living				n/a
Less than one year	22.2
One year or more	77.8
Depressive symptoms		5.18	1.4		2.6	1.14
Stress related to caregiving		18.43	2.65		14.05	4.47
Relationship to resident	n/a
Spouse				5.3
Parent				83
Other				11.7
Resident dementia diagnosis (yes)	n/a			48

n/a = not applicable.

Measures

Family Measures

Family members completed scales assessing their perceptions of interactions (both positive and negative) with staff. The Interpersonal Conflict Scale (Pillemer & Moore, 1989) asks participants to indicate how frequently they experience conflicts with staff members over issues, such as personal care, meals/foods, laundry/clothing, and attentiveness to the resident’s needs (α = .85). A shortened version of the Nursing Home Hassles Scale (Stephens, Ogrocki, & Kinney, 1991) was administered to assess the frequency in which family members experience negative staff behaviors, such as staff members exhibiting rude behavior (α = .80). Caregiver burden was assessed using a 6-item version (α = .80) of the Zarit Burden Interview (Zarit, Todd, & Zarit, 1986), and depressive symptoms were measured using a seven-item version (α = .73) of the Center for Epidemiologic Studies-Depression (CES-D) scale (Radloff, 1977).

Staff Measures

Two measures administered to staff were identical to the measures administered to family members: The Interpersonal Conflict Scale (α = .90) and the CES-D (α = .90). Staff members were also administered the Family Behaviors Scale (Pillemer, Hegeman, Albright, & Henderson, 1998), which assesses staff perceptions of the frequency in which family members treat them with respect, act rude toward them, or ignore them (α = .68). To measure staff burnout, a modified version (Pillemer & Moore, 1989) of the Depersonalization subscale (α = .71) of the Maslach Burnout Inventory (Maslach, 1982) was administered. All measures were selected based on their established psychometric properties and prior use with family and staff in long-term care settings.

Analysis Plan

For Analysis 1, descriptive statistics were computed to generate the frequency of participant responses for each scale of interest. In Analysis 2, multilevel modeling was first utilized to account for community nesting to examine differences in variance among the 20 ALFs included in the analysis. For all outcomes of interest, very little variance in family and staff outcomes [with intraclass correlations (ICC) of less than 3%] were attributed to ALFs. For this reason, linear regression analyses were instead used to examine predictors of caregiving-related stress and depressive symptoms in family members and staff, controlling for relevant covariates, including age, education, gender, and length of employment in the ALF for staff models.

Two-block hierarchical regression analyses were conducted for families and staff, with covariates included in Block 1 and relevant predictors included in Block 2. Follow-up analyses examining subgroups of families (e.g., having a resident with dementia) and staff (e.g., direct care workers) were examined using hierarchical regression models. Job category was dichotomized to reflect direct care workers versus other. Therefore, to examine subgroup differences in the effects of interpersonal conflict and perception of family/staff behaviors on our outcomes of depressive symptoms and stress-related to caregiving, we conducted separate analyses by job category in staff and dementia status of residents for the conflict and treatment variables and compared the similarity of coefficients across models using the following equation (Paternoster, Brame, Mazerolle, & Piquero, 1998):

$${\displaystyle \begin{array}{l}{\displaystyle Z=\frac{b_1-b_2}{\sqrt{SE{b}_1^2-SE{b}_2^2}}}\end{array}}$$

Where, b = beta coefficient, SE = Standard error, SEb2 = Squared Standard Error.

Results

Analysis 1: Family and Staff Perceptions of Conflict and Support

To shed light on the nature of family–staff relationships in ALFs, we examine the responses to the scale items used in this study. We begin by examining family and staff perceptions of conflict with one another. We then examine staff perceptions of treatment by families, followed by family perceptions of similar issues.

Conflict Between Family Members and Staff

Table 2 presents and compares rates of perceived conflict between families and staff members. The data show that the majority of both families and staff experience low levels of conflict in all areas measured, with a majority of participants in both groups reporting that conflicts in a particular domain occur less than once a month. In general, families reported less conflict than staff across all domains.

Table 2.

Perceived Interpersonal Conflict Between Staff and Families

	Staff	Family
Variables	n (%)	n (%)
Conflicts over personal care
Less than once a month	297 (75.2%)	174 (82.5%)
Once a month	43 (10.9%)	16 (7.6%)
A few times a month	38 (9.6%)	14 (6.6%)
A few times a week	14 (3.3%)	6 (2.8%)
Every day	4 (.9%)	1 (.5%)
Conflicts over meals or food
Less than once a month	258 (64.3%)	168 (77.8%)
Once a month	54 (13.5%)	13 (6.1%)
A few times a month	58 (14.5%)	21 (9.8%)
A few times a week	21 (5.2%)	9 (4.2%)
Every day	10 (2.5%)	3 (1.2%)
Conflicts over administrative rules
Less than once a month	283 (73.1%)	192 (89.8%)
Once a month	48 (12.4%)	9 (4.2%)
A few times a month	36 (9.3%)	10 (4.7%)
A few times a week	14 (3.6%)	3 (1.4%)
Every day	6 (1.4%)	0
Conflicts over laundry or clothing
Less than once a month	238 (61.3%)	165 (78.2%)
Once a month	53 (13.7%)	19 (9%)
A few times a month	71 (18.3%)	23 (10.9%)
A few times a week	18 (4.6%)	4 (1.6%)
Every day	8 (2.1%)	0
Conflicts over resident appearance
Less than once a month	308 (79.4%)	175 (83%)
Once a month	44 (11.3%)	14 (6.6%)
A few times a month	23 (5.9%)	17 (8.1%)
A few times a week	6 (1.5%)	5 (2.4%)
Every day	7 (1.8%)	0
Conflicts over toileting
Less than once a month	299 (77.1%)	169 (80.9%)
Once a month	41 (10.6)	14 (6.7%)
A few times a month	30 (7.7%)	19 (9.1%)
A few times a week	8 (2.1%)	7 (3.3%)
Every day	10 (2.6%)	0
Conflicts over attentiveness to resident needs
Less than once a month	287 (72.8%)	145 (67.2%)
Once a month	38 (9.6%)	34 (15.7%)
A few times a month	44 (11.2%)	24 (11.1%)
A few times a week	17 (4.3%)	12 (5.6%)
Every day	8 (1.9%)	1 (.5%)

The most commonly reported conflicts experienced by staff relate to the domains of laundry or clothing, in which 13.7% reported experiencing conflict at least once a month; and conflicts related to meals or food, in which 13.5% of staff participants reported experiencing conflict at least once a month. Family respondents reported the most frequent conflicts to be in the areas of attentiveness to resident needs, laundry or clothing, and meals or food—in which 15.7%, 9%, and 6.1% reported conflicts at least once a month, respectively.

Staff Perceptions of Family Behaviors

Table 3 presents items administered to staff respondents related to the treatment of staff by family members. Overall, a majority of staff members indicated that family members are either sometimes (31.1%) or always (68.3%) likely to treat staff with respect, with a similar pattern emerging for the item inquiring about the frequency in which family members are likely to smile and greet the staff member. Approximately 44% of staff respondents indicated that family members are sometimes likely to be rude with requests, and 20% indicated that family members are sometimes likely to ignore or brush the staff member aside.

Table 3.

Staff Perceptions of Family Behaviors

Item	n (%)
How often are family members likely to treat you with respect?
Never	0
Rarely	0
Sometimes	133 (31.1%)
Always	295 (68.3%)
How often are family members likely to be rude with requests?
Never	90 (21.3%)
Rarely	137 (32.4%)
Sometimes	188 (44.4%)
Always	8 (1.9%)
How often are family members likely to smile and greet you?
Never	1 (0.2%)
Rarely	4 (0.9%)
Sometimes	161 (38%)
Always	258 (60.8%)
How often are family members likely to ignore you or brush you aside?
Never	166 (39.6%)
Rarely	159 (37.9%)
Sometimes	84 (20%)
Always	10 (2.4%)

Family Perceptions of Staff Behaviors

Table 4 presents items administered to family respondents who were asked about a range of staff behaviors toward them and toward their family member. Families had relatively positive perceptions of staff behaviors. When asked about the frequency in which staff members are rude, ignore their family member’s requests, and are intolerable toward their family member, a majority of participants reported no occurrences of these behaviors. In contrast, 50% of family respondents reported having to tell staff how to care for the resident at least once a while.

Table 4.

Family Perceptions of Staff Behaviors

Item	n (%)
Staff members being rude to you
Never	197 (83.8%)
Once in a while	34 (14.5%)
Often	4 (1.7%)
Very often	0
Your family member complaining about staff
Never	89 (37.7%)
Once in a while	106 (44.9%)
Often	26 (11%)
Very often	15 (6.4%)
Needing to tell the staff how to care for your family member
Never	87 (37.5%)
Once in a while	116 (50%)
Often	21 (9.1%)
Very often	8 (3.4%)
Staff leaving your family member ungroomed or untidy
Never	137 (59.8%)
Once in a while	68 (29.7%)
Often	16 (7%)
Very often	8 (3.5%)
Staff complaining to you about your family member
Never	187 (79.9%)
Once in a while	40 (17.1%)
Often	6 (2.6%)
Very often	1 (0.4%)
Staff ignoring your family member’s requests
Never	123 (53%)
Once in a while	87 (37.5%)
Often	17 (7.3%)
Very often	5 (2.2%)
Staff being intolerant toward your family member
Never	188 (81%)
Once in a while	29 (12.5%)
Often	13 (5.6%)
Very often	2 (0.9%)

Consistent with our first hypothesis, these analyses reveal that family members report less conflict in comparison to staff. Staff members are more likely to report conflicts with families over issues, such as laundry/clothing and meals/food. Families most frequently reported conflict in the domains of attentiveness to resident needs, laundry/clothing, and meals/food. In terms of staff perceptions of family, most respondents reported positive perceptions such as families treating staff with respect and smiling and greeting them. Similarly, a majority of respondents indicated that they have never encountered staff members who complained about the resident (79.9%), left the resident ungroomed or untidy (59.8%), or were intolerant toward the resident (80%).

Analysis 2: Impact of Family–Staff Relations

Our previous analyses suggest that family and staff held generally positive perceptions that were reciprocal. However, we also showed that levels of conflict and negative interactions do, in fact, occur in assisted living. We now turn to the question: To what degree do family–staff interactions predict psychological well-being (measured by a depression scale) and stress related to caregiving? In the case of families, stress is measured by caregiver burden; for staff, it is measured by job burnout. The analysis will include control variables and the two perception scales used for each group.

Staff Burnout

A two-step hierarchical multiple regression was conducted with staff burnout as the dependent variable. Age, education, gender, and length of time employed at the ALF were entered as control variables in Block 1, and interpersonal conflict and perceived treatment by family members were entered in Block 2. Results show that in Step 1, age (β = .32) and length of time employed at the ALF (β = −.17) were the only statistically significant predictors of staff burnout [F(4, 334) = 9.83, p < .001] and accounted for 10.5% of the variation in staff burnout. Introducing the interpersonal conflict and perception of treatment by family variables explained an additional 16% of the variance in staff burnout, and this change in R² was significant [F(6, 332) = 10.48, p < .001]. Age (β = .29), shorter length of employment in the ALF (β = −.14), greater frequency of interpersonal conflict (β = .10), and worse perception of family treatment (β = .19) significantly predicted higher levels of staff burnout.

The hypothesis regarding job category was supported. Tests of the difference between the coefficients by job status (direct care worker vs other) indicated that the interpersonal conflict coefficient was significantly larger for direct care staff versus staff in other positions (p < .05, one-tailed). This finding suggests that the more intense interactions with residents and families on the part of direct care staff provide greater opportunities for interpersonal conflict. Job status did not have an effect on perceptions of family treatment.

Staff Depressive Symptoms

A two-step hierarchical multiple regression was conducted with staff depressive symptoms as the dependent variable controlling for age, education, gender, and length of employment in the ALF. In Block 1, which included the same covariates, only age (β = .19) emerged as a statistically significant predictor of staff depressive symptoms [F(4, 350) = 4.70, p < .001] and accounted for 5% of the variation in staff burnout. The addition of the perception of family treatment and interpersonal conflict variables in Block 2 explained an additional 11.7% of the variation in staff depressive symptoms, with a significant change in R² [F(6, 348) = 7.66, p < .001]. Age (β = .20), greater frequency of interpersonal conflict (β = .14), and worse perception of family treatment (β = .20) significantly predicted higher levels of staff depressive symptoms. Subgroup analyses examining the difference between the coefficients by job status (direct care worker vs other) revealed no significant differences.

Family Burden

As with the staff models, a two-step hierarchical multiple regression analysis was conducted with family caregiver burden as the outcome variable, controlling for age, gender, and education. In Block 1, which included age, education, and gender as covariates, gender (β = .19) was the only statistically significant predictor of burden [F(3, 175) = 3.18, p < .05] and the model explained 5% of the variance. After including the variables representing interpersonal conflict and perception of staff behaviors in Block 2, the model was significant [F(5, 173) = 2.84, p < .05] and explained an additional 7.6% of the variance in family burnout. However, gender was the only significant predictor to emerge (β = .19), such that women experienced higher levels of burden. Subgroup analyses further revealed that in family members whose resident had a dementia diagnosis, higher caregiver burden was associated with worse perceptions of treatment by staff (p < .05, one-tailed). No differences were identified for interpersonal conflict based on dementia status.

Family Depressive Symptoms

For the outcome of depressive symptoms, neither the covariates [F (3, 173) = 2.03, p = .11] nor the inclusion of the interpersonal conflict and perception of staff behavior variables in Block 2 [F(5, 171) = 1.32, p = .26] significantly predicted depressive symptoms levels in family members. Subgroup analyses further revealed that the regression coefficients did not significantly differ based on dementia status.

Discussion

ALFs continue to expand nationally and provide a comprehensive continuum of care to medically complex patients and individuals with dementia, which has evolved to become similar to the services provided in nursing homes. Within nursing homes, the quality of the relationships that exist between family members and staff affects quality of care. Both family members and long-term care staff are confronted with unique stressors in relation to their roles as caregivers, and the perception of these issues may influence the interpersonal dynamics that exist between them. However, little is known about the nature of family–staff dynamics in ALFs. The current study was designed to provide the first evidence regarding the actual prevalence of conflict and positive and negative interactions from the perspectives of both family members and staff. Overall, a majority of family and staff members reported low levels of conflict and generally positive perceptions of each other, with family members reporting lower levels of conflict compared with staff.

Additional analyses indicated that older age, shorter length of employment in ALFs, worse perception of family treatment, and greater frequency of interpersonal conflict significantly predicted staff burnout. As hypothesized, the effect of interpersonal conflict was more pronounced in direct care staff, which is likely due to the greater frequency of contact and interactions that direct care staff have with family members of residents. Additionally, age, interpersonal conflict, and worse perception of family treatment emerged as significant predictors of depressive symptomatology, although no differences were identified by subgroup.

When examining burden in family members, results indicated that female gender was the only significant predictor. However, subgroup analyses revealed that worse perceptions of staff treatment were associated with higher levels of caregiver burden in family members whose resident had a dementia diagnosis. Alternatively, no significant predictors emerged for family depression. Family caregivers tend to experience feelings of guilt surrounding the decision to place their loved one in long-term care, making it plausible that interpersonal conflict with staff members within these facilities can exacerbate feelings of guilt-related distress, rather than depressive symptoms (Bauer, Fetherstonhaugh, Tarzia, & Chenco, 2014; Majerovitz et al., 2009). In fact, a prior study has shown that poor family–staff relationships contributed to higher levels of guilt and distress in family caregivers (Majerovitz et al., 2009).

The findings of this study indicate that family–staff relationships in ALFs appear to be relatively harmonious when compared with the interpersonal dynamics that exist within nursing homes. Research has shown that care provided to individuals in nursing homes appears to be more task-centered than person-centered, which is likely due to institutional-level barriers, such as high workload, leaving minimal time to attend to family members (Hertzberg et al., 2003; Liu, Guarino, & Lopez, 2012). It is also plausible that person-centered care approaches in ALFs are more easily implemented due to the characteristics of the residents themselves. Interestingly, however, the rates of dementia in ALFs (42%; Caffrey & Sengupta, 2018; Zimmerman et al., 2014) and nursing homes (47.8%; Harris-Kojetin et al., 2019) are relatively similar. Thus, it is possible that the movement toward person-centered care and a departure from the traditional hospital-like models in ALFs, along with the emphasis ALFs place on maintaining the autonomy and privacy of residents, may partially account for the positive relationships between staff and family members in the current study (Bauer & Nay, 2010; Barken & Lowndes, 2018; Koren, 2010). Future research would benefit from examining differences in resident characteristics and their impact on person-centered care approaches across long-term care settings.

However, we also found that, similar to nursing homes, interpersonal conflict and worse perception of treatment are significant predictors of burden and depressive symptoms in staff members (Abrahamson et al., 2009; Gallego-Alberto et al., 2018; Majerovitz et al., 2009; McKenzie et al., 2011). Alternatively, in family members, worse perception of staff treatment was a stronger predictor of caregiver burden in families whose resident had a dementia diagnosis. When caring for an individual with dementia, family members rely more heavily on staff for information about their loved one’s care and condition (Robison et al., 2007), and staff members—without malicious intent—may place more emphasis on addressing the needs of the patient while overlooking the family caregiver (Morris & Thomas, 2002; Stajduhar, 2003; Weinberg, Lusenhop, Gittell, & Kautz, 2007). This situation can contribute to worse appraisals of staff, which can further interfere with the ability to establish collaborative relationships among staff and families.

Two limitations of the study should be noted. First, the cross-sectional nature of the analysis does not allow for causal inferences about the relationship between family–staff relationship quality and the outcomes in our analyses. Although this study includes the standard limitations of cross-sectional data, it is highly useful to establish baseline data on the quality of family–staff relationships in ALFs and to examine cross-sectional relationships among variables of interest, which can guide future longitudinal research to verify the findings of the current study. The present study lays the groundwork for longitudinal research by both establishing the prevalence of negative and positive assessments of these relationships and showing that they are related to outcomes of interest.

Second, the response rate from family members was relatively low (15% responded), which may limit the generalizability of our findings. Some participating communities found it challenging to recruit family members for the project. This pattern was in contrast to the experience of conducting similar programs in nursing homes in which family members were substantially more eager to participate (Pillemer et al., 2003; Robison et al., 2007). One possible reason for this difficulty is that family members in ALFs may experience fewer issues with care than did individuals with relatives in nursing homes; therefore, they may have been less motivated to respond. We do not have information on staff or family members who chose not to participate in the data collection, and we therefore cannot compare responders to nonresponders. Future studies should utilize additional recruitment efforts to achieve a broader, more representative sample of family members to confirm the results presented in this article.

Despite these limitations, taken together, the findings provide insight into the dynamics among family members and staff in ALFs and have several important implications. Staff members need to be cognizant of the importance of family–staff collaborations and work to engage and communicate with families more frequently. Consistent communication between family members and staff becomes especially important when residents have chronic conditions, such as dementia, that result in increasing dependency and impairment (Gittell, 2005).

Personalized resident care would be more efficient if care plans consider the importance of the family member as a part of the primary care team to provide the most optimal quality of care to residents. Family members can provide invaluable information about behaviors and care preferences and can identify changes in the status of the resident based on the deep knowledge of the resident. At the institutional level, efforts should be made to allow this information to be easily accessible by staff members who interact with the resident to promote more efficient person-centered care. Not only do improved family–staff relationships contribute to better patient care, it may also reduce feelings of burnout and depression in staff, and feelings of burden in family members, essentially benefitting all members of the care team (Majerovitz et al., 2009).

Moreover, the current study sheds light on how staff outcomes can be influenced by their interactions with family members, particularly among direct care staff. Professional caregiving burden has been related to worse physical and mental health and increased rates of turnover (Beeber et al., 2014; Woodhead et al., 2016). Given the difficulty with staff retention and satisfaction in long-term care facilities, greater cohesion in staff members’ relationships with families constitute important targets for intervention efforts to decrease stress in long-term care staff. However, Hengelaar and colleagues (2018) reported that care providers may feel unsure or unable to effectively collaborate with family members, suggesting that strategies for supporting staff to further support family members should be prioritized in the educational curricula of health care professionals. In conclusion, the results of this study provide a baseline in establishing a profile of family–staff relationships in ALFs that can be used in the design of interventions targeting improvement of family–staff communication and relationships in ALFs and other residential care settings.

Funding

Support for this project came from the University Research Grant Program of the American Seniors Housing Association (K. Pillemer, PI). Support was also received from an Edward R. Roybal Center Grant from the National Institute on Aging (NIA; P30AG022845). Falzarano acknowledges support from a NIA-funded T32 training program grant (AG049666).

Conflict of Interest

None reported.