Introduction

It is widely known that the quality of life (QOL) of people with severe mental illness (SMI) is significantly lower than that of the general population (Kao et al., 2011). People with SMI often experience reduced opportunities for social participation owing to social stigmas and community stereotypes (Abdullah & Brown, 2011). Many spend their time sleeping and resting and have difficulty engaging in meaningful occupations (Leufstadius & Eklund, 2014). Thus, community adaptation for them should target improving QOL instead of just avoiding hospitalization (Chan et al., 2005).

In Japan, most patients with SMI have been hospitalized for extended periods, with an average duration of hospitalization of 13.5 years (Oshima et al., 2007). Recently, the Ministry of Health, Labour and Welfare announced that it would be able to promote the discharge of 72,000 inpatients. The outreach service was enhanced to support the community life of discharged patients. The multidisciplinary outreach team consists of nurses, occupational therapists, and psychiatric social workers. Occupational therapists implement home-visit occupational therapy (OT) in this multidisciplinary outreach team. We implemented the home-visit OT that enable people with SMI to do what they “want to do” and “need to do” by using management tool for daily life performance (MTDLP) (Mashimo et al., 2020). MTDLP is paper-based tool that is developed to achieve the meaningful occupation of clients by Japanese Association of Occupational Therapists (JAOT). According to the OT theory, their QOL ought to be enhanced through their engagement in meaningful occupations (Hammell, 2004). However, this hypothesis has not revealed among the users of home-visit OT using MTDLP.

QOL is defined as individuals’ perceptions of their position in life in the context of the culture and value systems within which they live and in relation to their goals, expectations, standards, and concerns (WHO QOL Group, 1998). Existing QOL literature shows discrepancies among the concepts used by researchers. While researchers have conceptualized QOL as a subjective affair that only patients can report, others have argued that there are ‘‘objective’’ indicators of QOL, such as housing and health status, that can be used to evaluate patients’ QOL (Eack & Newhill, 2007). Recently, the recovery paradigm has enhanced the subjective aspects (e.g., the life story, strength, preferences) (Lloyd et al., 2008) of people with SMI. It has been also demonstrated that self-reported QOL by psychiatric patients is correlated significantly with the clinician’s estimates, and been able to evaluated accurately and consistently (Voruganti et al., 1998). Thus, this study focuses on improving subjective QOL rather than objective QOL of people with SMI.

Investigating the predictors that improve subjective QOL among people with SMI is useful to implement an effective home-visit OT. Kao et al. (2011) reported that the negative predictors of subjective QOL in 104 inpatients with schizophrenia were depressive symptoms, extrapyramidal side effects of psychotropics, hopelessness, and early age of onset. Furthermore, having reduced hospitalization time, fewer side effects due to psychotropic medication, and good social functioning have shown a positive-significant correlation with subjective QOL (Hasan, 2019; Chino et al., 2009). Therefore, the onset of age, depressive symptoms, psychotropic dosage, hospitalization periods, and social functioning seem to influence subjective QOL.

In OT research, a cross-sectional study of 45 community-dwelling people with SMI reported that perceived pleasure in the domains of work and rest was positively correlated with subjective QOL (Aubin et al., 1999). Another cross-sectional study on 103 community-dwelling people with SMI by Eklund and Leufstadius (2007) reported that satisfaction with and value of daily activities were positive predictors of subjective QOL. The pleasures, satisfaction, and value of daily activities may therefore be related to subjective QOL in community dwelling people with SMI.

Thus, this study aimed to identify the predictors of subjective QOL of community-dwelling people with SMI who use home-visit OT using MTDLP in Japan. The stepwise multiple regression analysis of factors already identified as affecting subjective QOL—such as demographic factors, social functioning and psychological factors related to activities (i.e., pleasure and satisfaction with activities)—will reveal the factor that most improves subjective QOL of home-visit OT users. We reported the effects of home-visit OT using MTDLP on social functioning in a previous randomized controlled trial (Mashimo et al., 2020). This study is cross-sectional study for the people with SMI receiving home-visit OT using MTDLP, and may contribute to find the therapeutic elements for improving their subjective QOL.

Methods

Recruitment

We invited applications from research participants between January and October 2018. The application process comprised an open call at the academic conference and a web advertisement. Twenty multidisciplinary outreach teams agreed to participate in this study. Team locations covered both rural and urban areas. People with SMI were given explanations about the study by the multidisciplinary outreach team staff. Those willing to participate were asked to provide informed written consent prior to participation (Mashimo et al. 2020).

Participants

Participants were from the subdivision groups after randomization process in the previous study (Mashimo et al., 2020). The inclusion criteria were adults aged 18–65 years, who were diagnosed with ICD-10 F2 (i.e., schizophrenia, schizotypal, and delusional disorders) or F3 (i.e., mood [affective] disorders) by an attending physician, and receiving home-visit OT using MTDLP. Exclusion criteria were a diagnosis of severe and moderate mental retardation, dementia, or substance-use disorder; refusing to participate; an evaluation of being unable to give consent; and having psychiatric symptoms that were predicted to worsen because of participation by the attending physician or the director of the team. All participants were already using psychotropics when the study began (Chlorpromazine equivalent doses: mean = 582.2 mg/day, SD = 502.2 mg/day; Imipramine equivalent doses: mean = 29.0 mg/day, SD = 6.3 mg/day).

Home-visit OT Using MTDLP Protocols

MTDLP includes intake, assessment, goal setting, planning, and intervention of tasks (Japanese Association of Occupational Therapists, 2017). They received home-visit OT using MTDLP once a week for 30 min to one hour for four months. Home-visit OT using MTDLP protocols were as follows: (1) The occupational therapist clarified the participants’ desired daily activities by asking them what activities they needed or wanted to perform in their daily lives using MTDLP sheets.; (2) The occupational therapist evaluated the factors involved in the promotion and inhibition of these activities based on the International Classification of Functioning, Disability and Health (ICF) model.; (3) The occupational therapists and participants created “collaborative goals” to perform the meaningful daily activities; (4) The occupational therapists made created the following three-stage program: basic program (physical function and structural domain approaches), application programs (activity and participating domain approaches), and social adaptation programs (environmental factor domain adaptation approaches).; (5) The participants implemented these programs and plans. The occupational therapist created an environment that facilitated their daily activities, encouraged them to remain motivated to achieve their goals, and provided positive feedback.; (6) Participants assessed the scores of performances of and satisfaction with daily activity. These contents were recorded in the MTDLP sheets by occupational therapists.

Prior to administering the home-visit OT, all occupational therapists attended a one-day training program on using MTDLP.

Data Collection

Both clinical and demographic data were collected. In this study, 20 occupational therapists administrated the home-visit OT using MTDLP. They had an average clinical experience of 12.76 (SD = 5.63) years. We provided them with training on recording data and administering the questionnaire before the study.

They collected demographic and clinical data from medical records. Demographic data included age, gender, education, and marital, living, and employment status. Clinical data included the diagnosis, illness duration, duration and number of hospitalizations, psychotropic dosage, and experience with outreach service.

The WHOQOL-BREF consists of 24 items in four domains of QOL (physical, psychological, social, and environmental) as well as two items that measure the overall QOL in general (WHO QOL Group, 1998),with a higher score represent a better subjective QOL. In this study, we used the Japanese version of the WHOQOL-BREF, which has been tested for content equivalence with the original English version (Tazaki & Nakane, 1997). If the participants had difficulty administering this assessment, the occupational therapist in charge assisted them by reading out the question items and selections response options.

The MTDLP self-rating scales consist of performance of and satisfaction with daily activity of the participants’ desired daily activity. They were evaluated in the range of 1–10 points on an MTDLP sheet by participants (Japanese Association of Occupational Therapists, 2017). The score of performance of daily activity is the subjective evaluation of one’s frequency of performing a desired daily activity. A high score is given for frequent performance. The score of satisfaction with daily activity is the subjective satisfaction with the desired daily activity. A higher score indicates higher satisfaction.

The GAF (Jones et al., 1995) was evaluated by medical professionals who were not involved in this study. Symptoms and psychological, social, and occupational functions are comprehensively evaluated. Responses are rated on a scale of 0 to 100, with a higher score indicating good social functioning. We used Japanese version of GAF (Takahashi et al., 2004).

Th e SFS (Birchwood et al., 1990) was evaluated by family members and group home caretakers. Participants who did not have a corresponding evaluator were assessed via a semi-structured interview with an occupational therapist. The SFS consists of 79 items divided into seven categories (social engagement/withdrawal, interpersonal behavior, prosocial activities, recreation, independence-competence, independence-performance, and employment/occupation.)

We used the Japanese version of the SFS (Nemoto et al., 2008; Sumiyoshi & Sumiyoshi, 2015).

Data Analysis

The data from participants who had experienced four months of home-visit OT using MTDLP were analyzed. To create a predictive model of subjective QOL, multiple regression analysis was performed based on the WHOQOL-BREF total score as a response variable. Independent variables were selected by referring to those reported as predictors of the WHOQOL-BREF score in previous studies (Kao et al., 2011; Hasan, 2019; Chino et al., 2009; Aubin et al., 1999; Eklund and Leufstadius, 2007 ). Pearson’s correlation analysis was conducted to evaluate the relationship between variables. To prevent multicollinearity, a variable whose correlation coefficient was statistically significant with r more than 0.9 was excluded from the independent variables. In the multiple regression analysis, the stepwise selection method was used, and the adjusted β was used as a measure of how strongly each independent variable affected the response variable. R2 was used to evaluate the variance of the independent variables of the predictive model. Variance inflation factor (VIF) was calculated to confirm co-linearity. VIF greater than 10.0 indicates multicollinearity (Chatterjee & Hadi, 2012). The variables indicating multicollinearity were excluded from the model. Since the significance of the regression model corresponds to the F distribution, if the normality of the residuals can be confirmed, the regression model is considered effective, without error. The normality of the residuals of the predictive model was confirmed using the Shapiro–Wilk test. Statistical significance was less than 5%. Data were analyzed using IBM SPSS Statistics version 25.

Results

Demographics

Table 1 shows the demographic and clinical variables. The gender ratio of the participants was almost the same, and 80% were in the schizophrenia classification. They were subjected to frequent and long hospitalization, with the average illness duration being 21.1 years, the average number of past hospitalizations being 3.6 times, and the average hospitalization period being 703.1 days. There were no significant differences in psychotropic dosage between the beginning of the study and when the data were collected (P > 0.05).

Table 1 Demographic and clinical characteristics of the participants (n = 25)
Full size table

Assessment Scale Results

Participants were asked by an attendant occupational therapist about their desired daily activities based on MTDLP protocols. Then, a collaborative goal was set to achieve them. Examples of collaborative goals were “working at the supported employment workshop twice a week” and “cooking independently.” Twenty-five participants mentioned their desired daily activities, which included work (n = 10, 40%), health management (n = 6, 24%), housework (n = 4, 16%), outing (n = 2, 8%), and other activities (n = 3,12%). A participant and an attendant occupational therapist planned together to implement the desired activity.

Table 2 shows the WHOQOL-BREF, MTDLP self-rating scale, GAF, and SFS scores after home-visit OT for four months. The total score of the WHOQOL-BREF was 84.2 (SD = 12.93), the score of performance of daily activity of MTDLP was 6.44 (SD = 3.01), and that of satisfaction with daily activity of MTDLP was 7.28 (SD = 2.30), the GAF score was 58.64 (SD = 13.62), and the total SFS score was 104.8 (SD = 29.35).

Table 2 Assessment scales' scores (n = 25)
Full size table

Predictors of Subjective QOL

Prior to performing a multiple regression analysis, a correlation analysis of the demographic and clinical variables and assessment scores were performed (Table 3). As the independence-performance and independence-competence scores of SFS subscales showed a strong correlation with r = 0.939, the independence-competence variable was excluded from the independent variables of the multiple regression analysis.

Table 3 Inter-correlations and significance levels of the variables (n = 25)
Full size table

The results of a stepwise multiple regression analysis with the WHOQOL-BREF total score as the response variable are shown in Table 4. Predictors of subjective QOL were satisfaction in daily activity of MTDLP (adjusted β = 0.584) and social engagement/withdrawal of the SFS subscale (adjusted β = 0.405). The VIF was 1.004, indicating no multicollinearity. The residuals of the created model were normally distributed (p = 0.462, Shapiro–Wilk test). The R2 was 0.534, explaining 53.4% of the variance in subjective QOL.

Table 4 Multiregression analysis predicting subjective quality of life (n = 25)
Full size table

Discussion

This study aimed to identify the predictors of subjective QOL of community-dwelling people with SMI who use home-visit OT using MTDLP. According to the stepwise multiple regression analysis, the predictors of subjective QOL were satisfaction with daily activity and social engagement. Approximately 53.4% of the variance in subjective QOL can be explained by these two variables. The study found that satisfaction with daily activity and social engagement, previously reported in separate studies, explained the subjective QOL of people with SMI. In the present sample, predictors reported in previous studies—such as length of illness duration and psychotropic dosage—did not affect subjective QOL. This suggests the strength of the influence of satisfaction with daily activity and social engagement on subjective QOL.

Note that the performance score of participants’ desired daily activity was not a predictor of subjective QOL. This suggests that the subjective QOL of people with SMI is dependent on their perception of satisfaction with the desired activity rather than feeling like they can perform it frequently. Similarly, Eklund and Leufstadius (2007) reported that the activity level and domain of the occupation are not correlated with subjective QOL; however, the value and satisfaction derived from daily activities were correlated with subjective QOL. Becker et al. (2005) also found that unmet daily life needs were associated with poor QOL. In other words, to improve subjective QOL, interventions focusing on subjective elements—such as satisfaction with daily activity of people with SMI—may be effective.

Social engagement was also determined to be a predictor of subjective QOL. This makes sense in view of the inclusion of social relationships among the constituents of subjective QOL (WHO QOL Group, 1998). Similar our result, Chino et al. (2009) reported that the social engagement/withdrawal score in SFS was significantly correlated with subjective QOL. Thus, promoting social engagement through meaningful daily activity for people with SMI with difficulties making social contacts may contribute to subjective QOL.

This study was cross-sectional study and we could not explain causality. Although, the results of this study have implications for occupational therapy oriented toward improving subjective QOL. Occupational therapists help people with SMI integrate into the community based on assessments and interventions that consider the complex interactions of people, the environment, and occupation by strategically and therapeutically using their desired daily activities. These therapeutic interventions may positive affect on subjective QOL.

Taken together, the results in this study suggest that increasing satisfaction with desired daily activities in people with SMI and restoring their contact with society may improve their subjective QOL. The satisfaction with desired daily activity and the social engagement may be important therapeutic elements in improving subjective QOL of people with SMI.

This study has several limitations. First, the sample size was small. In Japan, there are still only a few occupational therapists who conduct the newly started home-visit OT. Therefore, there was a limitation in obtaining a large sample. As the predictive model was established with a small sample size, a retest with a larger sample is necessary. Second, our sample was community dwelling patients. The subjective QOL of acute and inpatients may show different results. Thus, the study of patients in varying stages of illness will be necessary. Third, since 80% of the participants were diagnosed with ICD-10 F2, the characteristics of the illness may have influenced this study’s results. Therefore, further research is needed to identify illness-specific predictors of subjective QOL. Finally, predictors of subjective QOL identified in previous studies (e.g., depressive symptoms; Kao et al. 2011) were not examined in this study. These potential variables may influence the subjective QOL of SMI patients. Additional validation of these variables is needed in the future.

Conclusion

This study aimed to identify the predictors of subjective QOL of people with SMI using home-visit OT designed to enable them to engage in their desired daily activity. The results indicated that satisfaction with desired daily activity and social engagement were positive predictors of subjective QOL of people with SMI. Interventions focusing on enhancing satisfaction with desired daily activities in people with SMI and reconnecting them with society may contribute to an overall improvement in subjective QOL.