Relationship Between Functional Capacity and Community Responsibility in Patients with Schizophrenia: Differences Between Independent and Assisted Living Settings

Abstract

We examined factors potentially associated with level of community responsibility among middle-aged and older patients with schizophrenia. Participants in residential care facilities engaged in significantly fewer community responsibilities than those residing in the community. However, demographic and clinical characteristics did not explain these differences. Further, greater functional capacity was associated with greater community responsibility among participants residing in the community, but not those in residential care facilities. These results suggest that, despite capacity, patients residing in residential care facilities are not engaging in community responsibilities. Further, among participants residing in the community, functional capacity may predict level of responsibility.

Introduction

The functional disabilities associated with schizophrenia include diminished capability to live independently, maintain employment, attend school, manage finances, or communicate with others (Bowie et al. 2006; Green 1996; Green et al. 2000). These disabilities are often associated with implantation of supportive services, such as case management and, for some, residential care. Recovery-oriented models for psychosocial rehabilitation (Mueser et al. 2006) emphasize increasing daily living skills and enhancing individuals' engagement in meaningful involvement in the community (e.g., working, volunteering, going to school). However, it is unclear how functional abilities and environment interact in schizophrenia to facilitate or deter community involvement.

We previously demonstrated performance on a brief measure of functional capacity, known as the UCSD Performance-based skills assessment (UPSA-Brief), was effective in predicting whether people with schizophrenia were residing independently in the community (i.e., in a house or apartment by oneself) (Mausbach et al. in 2007). In that study, patients who scored 60 or above on the UPSA-Brief were optimally classified as capable of residing independently in the community. However, we did not assess the UPSA-Brief's ability to predict level of community involvement, such as working for pay, volunteer work, or attending school.

Accurate prediction of one's capacity to perform community responsibilities is a difficult task because prediction may be confounded by factors unrelated to capacity, such as patient motivation, expectations placed upon the patient, and/or opportunity provided by the residential environment. That is, while it may be that patients have the capacity to engage in community responsibilities, barriers such as lack of motivation or low expectations result in a low rate of community involvement. This may be particularly salient among patients residing in residential facilities (e.g., board & care homes, group homes, assisted living), where opportunity and/or expectations may be lacking compared to those of patients residing alone or with family. For example, studies have shown that people with schizophrenia residing in residential facilities appear to do so “for life” (Trieman et al. 1998), with very few patients ever achieving greater residential independence (de Girolamo and Bassi 2004; de Girolamo et al. 2002; Trieman et al. 1998). Documentation of differential engagement in community responsibilities among patients residing in the community (e.g., alone or with family) vs. those in residential care facilities, despite equivalent capacity, might help raise awareness for helping patients engage in meaningful activities that may subsequently improve their quality of life.

The purpose of this study was threefold. First, we were interested in determining whether patients residing in residential facilities (e.g., board & care facilities, group homes, assisted living) engaged in fewer community responsibilities relative to those residing in the community (independently or with family). Our second aim was to examine potential reasons why participants residing in residential care facilities differed from those residing in the community on their level of community responsibility (e.g., older age, greater functional impairment, and greater symptoms of psychosis). Finally, we also examined whether level of community responsibility was significantly associated with scores on our brief measure of functional capacity (i.e., UPSA-Brief).

Methods

Participants

A total of 240 middle-aged and older patients with a DSM-IV based chart diagnosis of schizophrenia or schizoaffective disorder were the participants in this study. All were enrolled in a large, randomized control trial known as Functional Adaptation Skills Training (FAST), which was designed to improve functional capacity among this population (Patterson et al. 2006). Data for this study were from the baseline assessment only; prior to participants receiving any skills training. To be eligible, patients had to be 40 or older and have a DSM-IV-based chart diagnosis of schizophrenia or schizoaffective disorder. Patients were excluded if they had dementia, were a serious risk for suicide, could not complete the assessment battery, or were participating in a separate psychosocial intervention or drug research at the time of enrollment.

Measures

We used a subset of questions from the Quality of Life Interview (QOLI; Lehman 1983; Lehman et al. 1993) to assess participants' community responsibilities. Four questions on this scale ask participants whether or not they engaged in the following activities during the previous month: (a) work at a job for pay; (b) go to school; (c) volunteer work; and (d) keep house or take care of children. The sum of ‘yes’ responses was used to indicate total number of responsibilities (range = 0−4). Functional capacity was assessed using the brief version of the UCSD Performance-based Skills Assessment (UPSA-Brief; Mausbach et al. 2007). Symptoms of psychosis were assessed using the Positive and Negative Syndrome Scale (PANSS; Kay et al. 1987). The 17-item Hamilton Depression Rating Scale (HAM-D-17) was used to assess depressive symptoms (Hamilton 1960). Overall cognitive functioning was assessed using Mattis' Dementia Rating Scale (DRS; Mattis 1973). Clinical interviews were conducted by trained raters employed by UCSD.

Statistical Analyses

Our first aim was to examine whether community-dwelling participants engaged in a greater number of community responsibilities compared to residential care participants. This was analyzed using a Mann–Whitney U test, with responsibilities as our dependent variable and living situation as our independent variable. Our second aim was to examine whether participants residing in residential care facilities differed from those residing in the community on a number of clinical and demographic characteristics, including age, education, gender, minority status, diagnosis, UPSA-Brief scores, DRS scores, PANSS positive and negative symptoms, and depressive symptoms. For this aim, we used chi-square and independent samples t-tests to compare categorical and linear variables, respectively. Our third aim was to determine whether UPSA-Brief scores were significantly related to level of community responsibility. To analyze this relationship, we conducted a one-way ANOVA, with UPSA-Brief score as the dependent variable and number of community responsibilities as the independent (grouping) variable. In the case of a significant omnibus test, follow-up (post-hoc) LSD tests were conducted to compare individual groups of participants. Partial eta-squared values (${\eta }_{\mathrm{p}}^2$) were calculated as an estimate of effect size, with values of .01, .09, and .25 representing small, medium, and large effects, respectively (Cohen 1988). Because previous research has shown that UPSA-Brief scores are significantly related to age, educational attainment, and negative symptoms, we re-ran our original ANOVA analysis while including these factors as covariates (i.e., ANCOVA) to determine whether our original results remained significant.

Study Approval

The protocol was approved by the UCSD Institutional Review Board (IRB), and all patients provided written, informed consent. In addition, all authors had no conflicts of interest in conducting this study or reporting the results of this manuscript, and all authors certify responsibility for the manuscript.

Results

Of our entire sample, residential status was missing from 1 participant. Of the remaining 239 participants, three failed to provide answers to the community responsibility questions, and were therefore excluded from the analysis. Therefore, our final sample consisted of 236 participants. Characteristics of these 236 participants are presented in Table 1. Because so few participants engaged in more than two community responsibilities (n = 9), these participants were collapsed with participants who engaged in two responsibilities. Therefore, our final analyses consisted of three groups: (a) those who engaged in ‘0’ responsibilities, (b) those who engaged in ‘1’ responsibility, and (c) those who engaged in ‘2 or more’ responsibilities.

Table 1.

Characteristics of the sample

Characteristic	M	SD	N	%
Age	48.7	7.3
High school education			77	32.6
Male			154	65.3
Ethnicity
Caucasian			125	53.0
Hispanic/Latino			59	25.0
African American			31	13.1
Asian/Pacific Islander			11	4.7
Other			10	4.2
Diagnosisa
Schizophrenia			189	80.1
Schizoaffective			46	19.6
Community responsibilities
0			108	45.8
1			90	38.1
2			29	12.3
3			7	3.0
4			2	0.8
UPSA-Brief	54.1	23.5
DRS	125.4	13.7
PANSS positive	14.9	5.5
PANSS negative	14.8	4.8
HAM-D	10.4	7.2

^an = 235. UPSA = UCSD Performance-based Skills Assessment; DRS = Mattis' Dementia Rating Scale; PANSS = Positive and Negative Syndrome Scale; HAM-D = Hamilton Depression Rating Scale

Comparison of Participants Residing in Residential Care Facilities vs. Community Dwellers

Our first analysis examined whether or not participants in residential care facilities engaged in fewer community responsibilities than those residing in the community. Results of our Mann–Whitney U test were significant (Mann–Whitney U = 4,286.5, z = 3.73; p < .001), whereby participants in the community (residing alone or with a family member) performed significantly more responsibilities than those in residential care facilities. Closer examination of these data indicated that whereas nearly 73% of those residing in the community engaged in at least 1 responsibility, only 46% of those in residential facilities did so (see Table 2).

Table 2.

Participant characteristics by living situation

	Living situation
	Assisted living (n = 163)	Community living (n = 73)	χ2, t	p-value
Age, M (SD)	49.3 (7.2)	47.3 (7.5)	1.92	.056
High school education, n (%)	116 (71.2)	43 (58.9)	3.45	.063
Male, n (%)	111 (68.1)	43 (58.9)	1.88	.170
Minority, n (%)	64 (39.3)	47 (64.4)	12.77	<.001
Schizoaffective, n (%)	19 (12)	27 (37)	20.39	<.001
Community responsibilities, n (%)
0	88 (54.0)	20 (27.4)	14.64	.001
1	54 (33.1)	36 (49.3)
2	21 (12.9)	17 (23.3)
UPSA-Brief, M (SD)	53.9 (22.4)	54.7 (25.9)	−0.24	.812
DRS score, M (SD)	125.6 (13.8)	125.0 (13.6)	0.32	.753
PANSS Positive, M (SD)	14.3 (5.3)	16.2 (5.7)	−2.44	.015
PANSS Negative, M (SD)	14.6 (5.1)	15.2 (4.2)	−0.77	.440
HAM-D, M (SD)	9.9 (6.6)	11.4 (8.2)	−1.44	.152

Note. UPSA = UCSD Performance-based Skills Assessment; DRS = Mattis' Dementia Rating Scale; PANSS = Positive and Negative Syndrome Scale; HAM-D = Hamilton Depression Rating Scale

To examine whether participants in residential facilities were more functionally impaired or had greater symptoms compared to their community-dwelling counterparts, we compared community and residential care participants in terms of demographic, functional, and psychiatric characteristics (see Table 2). There were no significant differences between the two groups in terms of functional capacity (i.e., UPSA-Brief scores; p = .812), global cognitive functioning (p = .753), negative symptoms (p = .440), or depressive symptoms (p = .153). The sole clinical variable to emerge as statistically different between groups was that participants in residential care facilities had less severe positive symptoms (p = .015). In addition to symptoms and functioning, participants did not differ significantly in terms of age, educational attainment, or gender. Interestingly, participants who were ethnic minorities and those with schizoaffective disorder were significantly more likely to reside in the community than in a residential care facility. However, analyses of our data indicated that number of community responsibilities was not different by ethnicity (${\chi }_{\left(2\right)}^2=5.34$, p = .069) or diagnostic category (${\chi }_{\left(2\right)}^2=3.13$, p = .210). These data suggest that clinical and demographic characteristics could not adequately explain why participants in residential care facilities had a restricted range of community responsibilities.

Relations Between Functional Capacity and Number of Community Responsibilities

Given the discrepancies in responsibility between those in residential care facilities and those residing in the community, we analyzed the relationship between community responsibilities and UPSA-Brief scores separately for these groups. Mean (95% CI) UPSA-Brief scores by number of responsibilities are presented in Fig. 1. Among patients residing in residential care facilities, our ANOVA was not significant (F_2,160 = 1.20; p = .303; ${\eta }_{\mathrm{p}}^2=.015$) (Fig. 1, Panel a). In contrast, among those living in the community, results of our ANOVA were significant (F_2,70 = 7.26; p = .001; ${\eta }_{\mathrm{p}}^2=.172$) (Fig. 1, Panel b). Among the community-dwelling group, post-hoc LSD analyses indicated that participants who performed ‘0’ responsibilities had significantly lower UPSA-Brief scores than those who performed ‘1’ responsibility (p = .043). Similarly, participants performing ‘2’ responsibilities scored higher than those performing ‘0’ (p < .001) and those performing ‘1’ responsibility (p = .023).

Fig. 1.

Mean (95% CI) UPSA-Brief scores by number of community responsibilities for participants in residential facilities (Panel a above) and participants residing in the community (Panel b below)

Among our community-dwelling sample, our second (adjusted) model, which included participant age, education, and negative symptoms as covariates, accounted for 36% of the total variance in UPSA-Brief scores. Both education (F_1,67 = 14.87; p < .001; ${\eta }_{\mathrm{p}}^2=.182$) and negative symptoms (F_1,67 = 6.73; p = .012; ${\eta }_{\mathrm{p}}^2=.091$) were significant predictors of UPSA-Brief scores. Community responsibilities were no longer significant (F_1,67 = 2.95, p = .059; ${\eta }_{\mathrm{p}}^2=.081$). Covariate-adjusted UPSA-Brief means (95% CI) by level of community responsibility were as follows: (a) 0 responsibilities = 46.0 (36.5−55.6); (b) 1 responsibility = 55.2 (48.3−62.1); (c) 2 or more responsibilities = 63.7 (53.3−74.2).

Discussion

This study assessed factors that were related to level of community responsibility in a sample of patients with schizophrenia. As expected, participants residing in the community engaged in a greater number of community responsibilities (i.e., working at a job for pay; going to school; volunteer work; and keeping house or taking care of children) than those residing in residential care facilities (e.g., board & care agencies). Contrary to expectations, participants residing in residential care facilities and those residing in the community, on average, were very similar in terms of demographic, clinical, and functional characteristics. Furthermore, our data suggest that community involvement was less dependent on the degree of functional capacity among residentially placed individuals than among people residing in the community. This implies that psychosocial rehabilitation aimed at increasing involvement in the community should consider not only the functional abilities and limitations exhibited by the individual, but the capacity of the environment to support their engagement in psychosocial goals. In addition, the manner in which other aspects of the environment moderate the relationship between functional capacity and community responsibility should be quantified (e.g., stigma, income from disability entitlements), in order to inform psychosocial rehabilitation. Finally, the focus may need to be on changing the contingencies in residential care that influence community involvement.

A number of measures have been developed for assessing functional living skills, including self- and clinician-rated instruments, direct observation, and performance-based measures (see review by Moore et al. 2007). Despite good reliability, few of the instrument development studies provided data on the instrument's ability to predict actual real-world functional independence. Based on the review, the authors recommended the Independent Living Scales (ILS; Loeb 1996) and the UPSA for assessing functional ability in patients with schizophrenia. With regard to the UPSA-Brief, we previously demonstrated that scores on the UPSA-Brief predicted one's ability to live alone in an apartment or house vs. in a residential care facility (Mausbach et al. 2007). Our current results extend the validity of the UPSA-Brief for predicting “real-world” outcomes to include level of community responsibility (e.g., work for pay, attend school), at least among those participants who reside in the community. That is, higher UPSA-Brief scores were related to greater level of community responsibility. These results may help treatment providers develop appropriate outcome goals (e.g., employment) for community-dwelling patients based on current level of functional capacity, and the UPSA-Brief may further be useful for monitoring progress toward these outcomes.

Our results are revealing in that they point to a need to examine why functional capacity (UPSA-Brief scores) did not predicts level of community responsibility in participants residing in residential care facilities. For example, future research should examine other factors that may contribute to placement and maintenance in assisted living. One factor that has been extensively studied has been stigma (Corrigan and Watson 2007). For example, Corrigan and Watson (2004) state that society often views schizophrenia as a disease from which people do not recover. That is, patients are perceived as forever needing care and oversight in their daily lives. Indeed, this perception may perpetuate patients' continual placement in assisted living facilities and/or hinder care facilities from employing psychosocial treatments to help patients become more independent. However, our data suggest that patients with greater functional capacity, as measured by the UPSA-Brief, are indeed capable of engaging in what society perceives as responsible behavior. Therefore, one means of reducing stigma in this patient population may be to increase use of psychosocial interventions in the community that target functional skills, which in turn may improve functional independence and responsibility.

We did not measure social support, but it may be that middle-aged and older persons with psychotic disorders who have few social resources are more likely to enter residential care, regardless of whether they can participate in community activities. It may also be that providers are less likely to initiate interventions to enhance community involvement (e.g., vocational rehabilitation) once a person enters a residential facility. Qualitative studies that examine barriers to and motivators for community involvement would be of particular interest as these studies may uncover factors that standardized assessment measures might overlook. Nonetheless, understanding the environmental factors that support or deter community involvement will be a vital step in implementing individually tailored rehabilitation programs.

Our results contrast with those of Bartels et al. (1997), who found that patients with schizophrenia who resided in nursing homes were more functionally impaired, exhibited greater overall psychiatric symptom severity (particularly positive symptoms), and were more cognitively impaired relative to community-dwelling patients. However, a number of differences between these studies may explain the contrasting results. First, our sample was considerably younger than that of Bartels and colleagues. Specifically, the average age of participants in that study was 72 years, whereas our sample averaged 49 years of age. Also, only 1 participant in our study resided in a nursing home. Rather, participants in our study resided in facilities requiring intermediate levels of care relative to nursing homes, such as board & care homes and assisted living facilities. Other research has investigated predictors of community functioning without examining differences between community-dwelling patients and those residing in residential care facilities. This research suggests that individual factors, such as neurocognitive functioning, are associated with community functioning. For example, both Green (1996) and Velligan et al. (2000) found that verbal memory impairment was negatively associated with multiple measures of community functioning, including social and vocational functioning. Further, Green found that negative symptoms were associated with social problem solving, whereas Velligan reported that scores on tests of executive function were inversely related to work and productivity. Future studies may want to relate scores on the UPSA-Brief to these multiple measures of community functioning.

One limitation of this study was that we assessed four specific community responsibilities (e.g., work for pay, going to school). These specific activities may not be the most salient for middle-aged and older adults with schizophrenia, and future studies may wish to include assessment of more age-relevant activities (e.g., engagement in community organizations; caring for grandchildren). Further, it is possible that place of residence may have been determined, at least in part, by the need to engage in community responsibilities. For example, participants residing in the community may have been more likely to remain in the community due to the need to care for children. However, because the item asked participants whether or not they “keep house or care for children”, it is unclear which of these responsibilities patients were endorsing. Ultimately, future research should include more detailed assessment of which specific responsibilities participants engage.

This was not a random sample. As such, it may be that a greater proportion of individuals residing in residential care facilities did not participate in the study because they were too impaired. However, it is worth noting that there are people residing in residential care facilities who are functionally indistinguishable from their counterparts in the community. Our measure of community engagement had a high “floor”, as only nine participants were participating in more than two activities. Future research should refine the indicators of community involvement so that incremental steps toward the types of community involvement measured in this study could be detected.

In our study, we relied on patients' self-reported engagement in community responsibility and did not require validation on whether or not participants indeed engaged in these responsibilities (e.g., verify reports with a confidant; obtain copies of pay stubs, etc.). While existing research indicates that patients with schizophrenia may be able to provide valid reports on their use of health services (Goldberg et al. 2002), we recommend that future studies validate patients' engagement in community responsibilities.

In sum, we found that participants in residential care facilities engaged in significantly fewer community responsibilities than those residing in the community (alone or with someone). However, few demographic, clinical, or functional differences were observed between these two groups, suggesting that reduced engagement in community responsibilities among participants in residential care facilities was not based on reduced capacity or greater impairment. These results suggest that psychosocial rehabilitation aimed at increasing involvement in the community should consider the capacity of the environment to support their engagement in psychosocial goals. We also found that, among participants residing in the community, higher scores on the UPSA-Brief were associated with increased level of community responsibility. These results suggest that in this subpopulation of persons with schizophrenia, the UPSA-Brief may be a useful measure for treatment providers for develop and monitoring progress toward “real-world” treatment outcomes.