Productivity loss of caregivers of schizophrenia patients: a cross-sectional survey in Japan

Abstract

Background: When a family member is diagnosed with schizophrenia, it causes stress to the caregiver that can eventually result in missed work days and lower work performance.

Aim: This study aims at revealing productivity costs for caregivers of schizophrenia patients in Japan.

Method: A cross-sectional survey of caregivers was conducted and resulted in 171 respondents. The assessment of work productivity included calculating the costs of absenteeism, presenteeism and total productivity costs. This was accomplished using the “Work Productivity and Activity Impairment Questionnaire” (WPAI).

Results: A relative majority of caregivers in the sample provided care for their spouse (47%), 18% cared for their brother or sister and 16% provided care for their child. Per capita productivity costs totaled JPY 2.42 million, with JPY 2.36 million (97%) of that amount being due to presenteeism.

Conclusions: The burden on caregivers is substantial enough to warrant structured support programs aimed at maintaining careers’ physical and mental health, helping them provide more effective care to schizophrenia patients and eventually increase productivity at work.

Keywords:

• Schizophrenia
• caregiver burden
• productivity cost
• Japan

Background

Schizophrenia is a chronic, severe and disabling brain disorder that significantly impacts a person’s thought processes and emotional responsiveness. The disorder typically consists of one or more symptoms, such as hallucinations, delusions, disorganized communication, poor planning, reduced motivation and blunted affect. Schizophrenia is associated with significant social and/or occupational dysfunction, reducing a patient’s quality of life and ability to engage in everyday activities (Saha et al., 2005). In Japan, schizophrenia occurs at a rate of approximately seven per 1000 persons (Sado et al., 2013).

The cost of illness for schizophrenia is substantial (Jin & Mosweu, 2017; Nakamura & Mahlich, 2017). In Japan, the total cost of schizophrenia has been estimated at approximately JPY 2.77 trillion, which is the equivalent of USD 23.8 billion (Sado et al., 2013). Out of this total, JPY 0.77 trillion (USD 6.59 billion) was the direct cost of healthcare and social services, while JPY 1.85 trillion (USD 20.2 billion) was the cost of morbidity and mortality. The literature indicates that both direct and indirect costs are important cost components in schizophrenia, which aligns with findings from other countries (Phanthunane et al., 2012; Wu et al., 2005).

The research conducted by Sado et al. (2013), however; did not account for the indirect costs incurred by caregivers of schizophrenia patients when care is provided in the family setting. Previous studies across a variety of disease conditions have noted an association between the level of caregiver burden and constraints in social activities, negative impact on family life and feelings of loss (Magliano et al., 2005a).

For schizophrenia, a number of studies have reported on caregiver burden in countries outside Japan, including European countries (Gupta et al., 2015a; Magliano et al., 2005a; Papastavrou et al., 2012; Zendjidjian et al., 2012), middle income countries, such as Chile (Gutierrez Maldonado et al., 2005) or developing countries like India (Jagannathan et al., 2014) and Nigeria (Lasebikan & Ayinde, 2013).

Most studies of caregiver burden use clinical or health-related outcome measures. A study conducted in Italy found that strong emotions (e.g. expressed hostility or critical comments) were positively correlated with more subjective burden among relatives of people with schizophrenia (Magliano et al., 2005a). One review reported how caregivers frequently experience deteriorated health, including stress-related conditions, anxiety and depression (Caqueo-Urizar et al., 2009). Furthermore, there appears to be a linear relationship between the amount of time spent providing care and reported levels of burden, with investigators noting that caregivers who provided cared to high-to-moderately dependent patients reported greater impairment in health, professional and family/social domains (Aranda-Reneo et al., 2013). Literature about the impact of such burdens on healthcare resource use or productivity among caregivers is sparse, with only a few studies attempting to measure productivity losses associated with caregiving. One cross-country study by Gupta et al. (2015b) reported that employed schizophrenia caregivers in Europe experienced serious work impairment resulting in average productivity costs of €6667, which was significantly higher than that of caregivers for other conditions. However, to our knowledge, limited information has been determined in Japan. For this reason, this study aims at assessing the productivity loss of caregivers of Schizophrenia patients in Japan. We acknowledge that the productivity loss is only a fraction of the overall economic burden of caregivers as it does not capture direct costs, such as health resource utilization. Nevertheless, we believe that this study addresses a gap in the literature as it sheds some new light on the situation of a specific population of caregivers in Japan.

Methods

The cross-sectional study was accomplished by a nationwide online survey of caregivers who are taking care of schizophrenia patients administered in July–August and November–December 2016 using an Internet survey service. We included caregivers who had a family member diagnosed with schizophrenia for at least 1 year and currently receiving at least one schizophrenia treatment. Applying those filters resulted in an eligible population of 371 caregivers in the database of the survey service company. Among the 371 caregivers who were invited to the survey, a total of 171 caregivers responded. The study was conducted under the guidelines of Johnson & Johnson and was approved by a Janssen approval committee. All participants provided informed consent and responses were automatically de-identified in the database.

The survey asked about the nature of the relationship to the patient as well as a number of demographic characteristics including age, gender, marital status, occupation, geographic region and annual income. Clinical characteristics including time since diagnosis (of the patient), and medication class [oral vs. long-acting injectable (LAIs)] were also recorded.

Work productivity was assessed through the “Work Productivity and Activity Impairment Questionnaire” (WPAI), a six-item validated instrument (Reilly et al., 1993). The six questions ask about missed and unproductive work hours during the last seven days and produce the following four metrics:

1. Absenteeism (percentage of work time missed due to health problems).

2. Presenteeism (percentage of impairment while working due to health problems).

3. Overall work productivity loss (percentage of overall work impairment due to health problems).

4. Activity impairment (percentage of activity impairment due to health problems).

The percentages of absenteeism and presenteeism were extrapolated for a whole year and multiplied by participants’ annual salaries to calculate productivity losses in monetary values. This approach assumes that workers are paid according to their productivity or in other words, that labor markets work frictionless. We are aware that this is a strong assumption for Japan, where performance-based pay is not necessarily the norm and seniority-based pay is still common (Lee et al., 2011).

Those who were no longer in the labor force were asked if they had quit their job in order to be a caregiver. If they answered yes, they were then asked to state their last compensation level before quitting their job. This number was used to calculate productivity costs for this subgroup of caregivers. According to the definition by both the international labor organization and the Japanese government, the status unemployment indicates that the unemployed are actively looking for a job. Those who quit their job due to caregiving on the other hand are not in the labor force anymore.

Because previous research has revealed a higher treatment satisfaction level of caregivers when patients were treated with LAIs (Han et al., 2005), a subgroup analysis according to antipsychotic medication type (oral vs. LAIs) was also conducted. The additional subgroup analysis was conducted to examine whether such higher satisfaction translates into economic benefits.

Results

Table 1 reports caregiver characteristics. The relative majority of caregivers provided care for their husband or wife (47%), 18% provided care for a sibling and 16% provided care for their child. In addition, 19% of caregivers reported that they used to work but had to quit their job because of caregiving. Most of the patients in this subgroup had previous income of around JPY 1.3 million. More than half (55%) of responding caregivers in the sample had a university degree. While caregivers reported a wide range of annual household income, 20% fell in the range of JPY 3.7–5.7 million and another 19% fell in the range of JPY 5.7–7.7 million.

Table 1. Caregiver characteristics.

Characteristics	Number (%)
Respondent	171
Age (mean ± SD)	47.88 ± 7.18
Gender
Male	121 (71)
Marriage status
Single	58 (34)
Married	113 (66)
Highest education
High school or less	50 (29)
College	27 (16)
Bachelor’s degree or higher	94 (55)
Current job
Full-time employee	72 (42)
Part-time employee	19 (11)
Self-employed	11 (7)
Unemployed	37 (22)
Quit work for caregiving	32 (19)
Relationship with patient
Mother	23 (13)
Father	5 (3)
Brother/sister	30 (18)
Husband/wife	80 (47)
Son/daughter	15 (9)
Other family members	18 (10)
Household income (JYP)
<800 000 Yen	3 (2)
0.8–1.6 M Yen	6 (4)
1.6–3.1 M Yen	31 (18)
3.1–3.7 M Yen	16 (9)
3.7–5.7 M Yen	34 (20)
5.7–7.7 M Yen	33 (19)
7.7–9.7 M Yen	20 (12)
9.7–11.6 M Yen	9 (5)
11.6–15.1 M Yen	5 (3)
>15.1 M Yen	1 (0)
I do not know	13 (8)
Productivity loss index (%)
Absenteeism [mean (SD)]	4.70 (11.89)
Presenteeism [mean (SD)]	24.96 (24.42)
Annual Productivity loss (JYP)
Absenteeism [mean (SD)]	197 355 (447 275)
Presenteeism [mean (SD)]	2.36 M (3.56 M)
TOTAL productivity loss	2.42 M (3.58 M)
Productivity loss due to quitting job	1.33 M (1.40 M)

SD: standard deviation; JPY: Japanese Yen; M: million.

Turning to patient characteristics, the results of which are summarized in Table 2, it showed that 40% graduated from university, 77% were currently unemployed and 61% were considered disabled. Regarding the type of medication used for treatment, only 7% were treated with an LAI while the remaining 93% were treated with oral medications.

Table 2. Patient characteristics.

Characteristics	Number (%)
Patients	171
Age (mean ± SD)	48.61 ± 13.95
Gender
Male	45 (26)
Marriage status
Single	54 (32)
Married	117 (68)
Highest education
High school or less	75 (44)
College	28 (16)
Bachelor’s degree or higher	68 (40)
Current job
Full-time employee	14 (8)
Part-time employee	16 (9)
Self-employed	2 (1)
Unemployed	130 (77)
Others	9 (5)
Time since diagnosis  (mean ± SD) [median (IQR)]	11.65 ± 9.51 [9 (5–17)]
Special condition
Diagnosed as disabled patients	105 (61)
Antipsychotic medication class n (%)
Oral	159 (93)
Long-acting injectable	12 (7)

IQR: interquartile range; SD: standard deviation; JPY: Japanese Yen.

For productivity loss of these participants, only 5% reported missing work due to taking care of schizophrenia patients in the last 7 d while their productivity reduced approximately 25%. The calculated annual productivity cost was JPY 2.42 million per year. Among total annual productivity loss, JYP 2.36 million was due to reduced work performance (presenteeism) while at work and JYP 197 355 is due to sick leave (absenteeism). In addition, we also found that productivity losses for the LAI subgroup were smaller than non-LAI group. However; due to limited number of respondents, we did not find any statistically significant. See Table 3 for complete results.

Table 3. Subgroup analysis according to antispsychotic medication.

	Total	LAI	Non-LAI	p Value
Productivity loss index (%), mean (SD)
Absenteeism	4.70 (11.89)	2.45 (4.36)	5.11 (13.22)	0.529
Presenteeism	24.96 (24.42)	11.00 (9.94)	26.15 (24.94)	0.059
Annual productivity loss (JYP), mean (SD)
Absenteeism	197 355 (447 275)	110 862 (217 558)	204 431 (460 903)	0.548
Presenteeism	2.36 M (3.56 M)	1.51 M (2.25 M)	2.43 M (3.64 M)	0.458
Total productivity loss per year	2.42 M (3.58 M)	1.51 M (2.25 M)	2.50 (3.67 M)	0.430

JPY: Japanese Yen; SD: standard deviation; M: million.

Discussion

Study results indicate that productivity losses for caregivers of schizophrenia patients are substantial and impose a significant burden on Japanese society. For people who are still in the workforce, productivity costs totaled JPY 2.42 million per capita, largely due to presenteeism (JPY 2.36 million per capita). In addition, 19% of caregivers reported they used to work but had to quit their jobs because of caregiving. Most of the caregivers who quit their jobs had relatively low income in their previous employment (JPY 1.3 million), indicating they had been working on a part-time basis.

What are the implications of these findings? The cost burdens on caregivers are significant enough to warrant efforts to reduce them. One such effort could be providing better treatment since it is known that better-functioning patients reduce the burden on caregivers (Flyckt et al., 2015; Hanzawa et al., 2010). Also, the subgroup analysis by medication type suggests productivity losses are less for caregivers whose sick relatives are treated with an LAI, although the difference is not statistically significant due to the small number of respondents in this subgroup (only 12 LAI patients). A confirmatory study to test this relationship is warranted. In addition, an acknowledged limitation of this aspect of the study is the lack of control for potential confounders of LAI usage. LAI results could be biased if LAI prescriptions are not random. For example there could be a correlation between disease severity and the likelihood of an LAI prescription. However, a recent claims database study in Japan did not find significant differences in the use of LAIs vs. oral medications with regards to patient characteristics (Cheung et al., 2017).

Previous studies suggest LAIs are associated with better adherence to treatment plans and relevant patient outcomes, such as relapses and re-hospitalizations (Cheung et al., 2017; Grimaldi-Bensouda et al., 2012; Tiihonen et al., 2011). In spite of higher acquisition costs, several cost-effectiveness studies demonstrated the economic superiority of LAIs over oral antipsychotics, although these studies do not take indirect costs, such as caregiver burden into consideration (Mahlich et al., 2015; Olivares et al., 2008; Zeidler et al., 2013).

Other avenues for reducing cost burdens are structured support programs for caregivers in order to maintain their health, provide more effective care and eventually increase productivity at work. Formal care programing, including family and social support interventions, may help to enhance social functioning in people with schizophrenia and decrease the frequency of hospitalizations, thereby resulting in reduced burdens on caregivers (Awad & Voruganti, 2008; Magliano et al., 2005b; Sharif et al., 2012).

Any such structured programs must be tailored to the Japanese cultural context. As was argued by Hanzawa (2012), Japanese families of patients with schizophrenia are likely to feel stigmatized and are also likely to feel schizophrenia should be hidden from others in the community, which can lead to the coping strategy of resignation. Furthermore, Japanese families tend to resist allowing others to care for family members with schizophrenia, which in turn can make the implementation of effective intervention programs more difficult.

Several limitations should be discussed. First, it was cross-sectional design. We cannot track both progression of the disease and its impact on productivity over time. Second, the surveyed sample might not be representative of the Japanese population of caregivers. The study participants completed the survey online, precluding participation from patients without internet access. Compared to a recent national survey (National Federation of Mental Health and Welfare Party in Japan, 2010) where most of the caregivers were parents (85%), the majority of the caregivers in our sample were couples taking care of the partner. Therefore, our sample population was younger with the oldest respondent being 60 years of age. For this reason, average household incomes were higher compared to the national survey which in turn would result in higher productivity losses. Lastly, due to retrospective self-reported data, participants’ responses were susceptible to recall bias. On the other hand, we tried to mitigate this bias by limiting the recall period to the last 7 days.

Ethics approval and consent to participate

This article is based on a web-based survey and does not involve any interventions conducted on human subjects by any of the authors. The authors were not involved in the collection of this data. Data were collected by the database provider and caregivers were informed that their data would be used for research (opt-out system). Data were de-identified before addition to the database.

Declaration of interest

JM and RS are employed at Janssen KK. The study design, data collection, data analysis, data interpretation, and writing of the report, was solely done by the authors. The findings and conclusions in this report do not necessarily reflect the views of Janssen Pharmaceutical KK.

We received funding from Janssen pharmaceutical KK to conduct the survey.

Acknowledgments

We would like to thank Medilead Inc, Tokyo to help us with data collection.