Health-related quality-of-life, work productivity, and economic burden among patients with Parkinson’s disease in Japan

Abstract

Aims: This study aimed to characterize the burden of Parkinson’s disease (PD) by examining health-related quality-of-life (HRQoL), impairments to work productivity and daily activities, healthcare resource use, and associated costs among Japanese patients with PD.

Materials and methods: This retrospective cross-sectional study used data from the 2009–2014 Japan National Health and Wellness Survey (NHWS) (n = 144,692). HRQoL (Short Form 36-Item Health Survey version 2), impairments to work productivity and daily activities (Work Productivity and Activity Impairment Questionnaire), healthcare resource utilization, and annual costs were compared between respondents with PD (n = 133) and controls without PD (n = 144,559). The effect of PD on outcomes was estimated using propensity score weighting and multivariable regression models.

Results: HRQoL was lower in patients with PD compared to the control group, with reduced physical (41.3 vs 51.3) and mental (35.7 vs 45.4) component summary scores and health state utility scores (0.62 vs 0.77; p < .001 for all). Patients with PD also reported higher levels of absenteeism (19.3% vs 3.3%), presenteeism (45.2% vs 18.5%), overall work impairment (52.8% vs 20.3%), and activity impairment (49.6% vs 20.8%) than controls without PD (p < .001 for all). In addition, patients with PD had higher healthcare resource utilization, direct (¥3,856,921/$37,994 vs ¥715,289/$7,046), and indirect (¥2,573,938/$25,356 vs ¥902,534/$8,891) costs compared with controls without PD (p < .001 for both).

Limitations: Data were cross-sectional and did not allow for causal inferences. Although the NHWS demographically represents the Japanese adult population, it is unclear whether it adequately represents the adult population with PD in Japan.

Conclusions: PD was associated with poorer HRQoL, greater work productivity loss, and higher direct and indirect costs. The findings suggest that an unmet need exists among patients with PD in Japan. Improving PD treatment and management could benefit both patients and society.

Keywords:

• Activity impairment
• direct costs
• healthcare resource utilization
• health-related quality-of-life
• indirect costs
• work productivity impairment

JEL codes:

• I11
• I13
• I19

View correction statement:

Correction

Introduction

Parkinson’s disease (PD) is a debilitating neurodegenerative disorder, which develops via a complex interplay of genetic and environmental factors. PD is characterized by bradykinesia, rigidity, resting tremors, impaired postural reflexes, occasionally leading to death^1,2. Although disability in patients with PD is limited initially, disease progression results in increased difficulties in mobility, balance, and speech, apart from depression and cognitive decline³. PD is most common in the elderly, with mean age of symptom onset typically being 62–70 years⁴. In Japan, prior studies reported a higher prevalence of PD in women than in men across all age groups^5–7. The adjusted prevalence of PD in Japan increased over time from 145.8 in 1980 to 166.8 in 2004 (per 100,000 population)⁷.

PD markedly impacts patients' health-related quality-of-life (HRQoL)^8–13. For example, a Japanese cross-sectional survey among community-dwelling patients found that intractable neurological diseases, including PD, were associated with low scores on all HRQoL domains, such as physical function, role physical, and role emotional, when assessed using the Japanese standardized version of the Short Form 36-item Health Survey version 2 (SF-36v2)¹⁴.

A substantial economic burden is also attributed to PD. Specifically, patients with PD also experience increased medical expenses^15–17, higher healthcare resource use^18–21, and work productivity loss^17,18,22. The direct medical expenses of patients with PD at a Japanese medical university hospital were found to be $5,828 annually²³. Average health-related absenteeism costs were found to be $1,344 higher for patients with PD than controls without PD in a privately insured US population¹⁵. Furthermore, in a retrospective study²⁴, the total direct annual costs for patients with PD in the US were two-times higher, compared with controls without PD, while another study of elderly Medicare beneficiaries reported $7,710 higher annual healthcare expenses for patients with PD vs patients without PD²⁵.

Although various studies have assessed the health-related and economic burden of PD, there is a paucity of studies on HRQoL among patients with PD in Japan¹⁴. In addition, there are no studies that quantify work-related productivity loss and activity impairment among the Japanese population with PD. Therefore, the current study was undertaken to comprehensively characterize both the health-related and economic burden of PD in Japan by evaluating HRQoL, impairments to work productivity and daily activities, healthcare resource use, and associated costs among patients with PD.

Methods

Sample

The sample for this retrospective cross-sectional study included all unique respondents who completed the 2009–2014 Japan National Health and Wellness Survey (NHWS; n = 147,270). If an individual completed the NHWS in more than 1 year, only the most recent response was included in the study. Potential NHWS respondents were selected through the Lightspeed Research online panel. NHWS respondents were recruited using stratified sampling intended to mirror the age and gender distribution of the Japanese adult population. The study was found to be exempt from review by the Pearl Institutional Review Board (Indianapolis, IN). Invitations to participate in the NHWS were sent by e-mail, with reminder e-mails sent as needed. To be eligible for the current study, all participants were required to be: (1) able to read and understand Japanese, (2) able to operate a computer to access the online survey, (3) located in Japan, (4) aged 18 or older, and (5) willing to provide informed consent to participate. The NHWS was completed by respondents using a confidential self-administered online survey. Thus, all data for this study were patient-reported. In exchange for their participation, NHWS respondents received points that could be redeemed for small prizes or accumulated over time for larger items.

Measures

Demographic characteristics

Demographic measures included age (entered continuously), gender (male/female), education (university degree vs all else), household income [<¥3,000,000 ($29,553); ¥3,000,000 ($29,553) to <¥5,000,000 ($49,254); ¥5,000,000 ($49,254) to <¥8,000,000 ($78,807); ¥8,000,000 ($78,807) or more, or declined to answer], and health insurance (national health insurance, social insurance, late stage elderly insurance, other, or no insurance).

Health characteristics

Health characteristics were measured using data on smoking status (current, former, or never smoked), exercise behavior (did not exercise in past month vs exercised in past month), alcohol use (currently consume alcohol vs abstain), body mass index (BMI) category (as per the World Health Organization’s recommendation for Asian populations²⁶: underweight [<18.5 kg/m²], acceptable risk [18.5–<23 kg/m²], increased risk [23–<27.5 kg/m²], high risk [≥27.5 kg/m²], or decline to provide weight), and the Charlson comorbidity index (CCI; a measure of comorbidity burden)²⁷.

PD status

Respondents who self-reported being diagnosed with PD were included in the PD group, whereas those who did not report a diagnosis of PD served as the control group.

HRQoL

HRQoL was assessed using scores on the PCS, MCS, and health state utilities using the SF-36v2 and Short Form-6 Dimensions (SF-6D), respectively. The SF-36v2 is a generic HRQoL instrument, which was used to generate MCS, PCS, and SF-6D scores²⁸. Scores on the MCS and PCS can theoretically range from 0–100; based on US population norms, these two measures have a mean of 50 and standard deviation of 10. The SF-6D is a preference-based single index measure with interval scoring properties; scores on this measure can theoretically range from 0–1²⁹. Higher scores on the MCS, PCS, and SF-6D health utilities signify better HRQoL.

Work productivity and activity impairment

Work productivity and activity impairment were assessed using the Work Productivity and Activity Impairment Questionnaire-General Health (WPAI-GH)³⁰. Four sub-scales (absenteeism, presenteeism, overall work impairment, and activity impairment) were generated from the WPAI-GH in the form of percentages, with higher values indicating greater impairment due to the patient’s health in the past 7 days. Only respondents currently employed (full-time, part-time, or self-employed) provided data with respect to absenteeism, presenteeism, and overall work impairment. All respondents provided data on activity impairment.

Healthcare resource utilization

Healthcare resource utilization was defined by the number of healthcare provider visits, the number of ER visits, and the number of times hospitalized in the past 6 months.

Costs

Indirect and direct costs were estimated from the available NHWS data. Annual indirect costs were calculated by integrating information from the WPAI-GH and hourly wage rates from the Japan Basic Survey on Wage Structure, 2011 using the Lofland method³¹. For each employed respondent, his or her annual wage was estimated based on median weekly rates (provided by demographic strata) multiplied by the number of work weeks in a year (52 weeks). Direct costs were estimated by multiplying the number of physician visits, ER visits, and hospitalizations by two (to estimate annual number of visits) and then multiplying by the corresponding unit cost for each, which was obtained from the literature. In the case of hospitalizations, only a cost per day was obtained from the literature³², and the NHWS asked respondents for the number of hospitalizations. To align them, the cost per day was multiplied by the average number of days per hospitalization, as reported by the Organization for Economic Cooperation and Development³³. The costs were converted to USD as per the rate at the midpoint of the year 2014 (July 1) using an online currency conversion website³⁴.

Statistical analyses

All analyses described below were conducted using Statistical Analysis System (SAS) v9.3.

Treatment of outliers and extraneous controls

The sample was checked before analysis for potential outliers and extraneous controls (i.e. respondents without PD whose values on the covariates were outside the range of the same covariates for those diagnosed with PD). Outliers were identified by examining the distribution of variables. Based on the distribution of the CCI score, 15 respondents were removed from the PD group, as they had values of CCI ≥ 30.0. Extraneous controls were trimmed solely based on age. The maximum and the minimum age for the PD group was 81 years and 21 years, respectively. Controls over the age of 81 or under 21 were excluded from the sample (n = 2,563), accordingly. After treating the sample for outliers and extraneous controls, the sample used for analysis included a total of 144,692 respondents (Figure 1).

Figure 1. Flow diagram of the study respondents.

Independent group comparisons

Differences in demographic and health characteristics were examined by PD status (PD group vs controls without PD). These results served to characterize the differences between these populations and aided in the selection of covariates for multivariable analyses. Chi-square and one-way analysis of variance tests were used to determine significant differences for categorical variables and continuous variables, respectively.

Creation of sample weights

A propensity score weighting process was used to minimize the large baseline differences, in both sample size and demographic and health characteristics, between the PD group and control group. The weights were derived using the Toolkit for Weighting and Analysis of Nonequivalent Groups (TWANG)³⁵. Based on the independent sample comparisons, the statistically different (p < .05) variables were entered into a generalized boosted model to predict PD presence, and to ultimately balance the study groups. In the final analysis, each person was assigned a weight that reflected how closely that person was matched; these weights were then used to derive weighted sample sizes that were more closely aligned across groups than the original ns.

Multivariable analyses

Weighted generalized linear models (GLMs) were used to further adjust baseline differences that existed after weighting, and to assess the unique burden associated with PD on health and economic outcomes and test if a statistically significant difference existed between the two study groups. Only age and CCI score were included, as they were the covariates that significantly differed after using the weighting procedure. A normal distribution for the error terms and an identity link function were used to estimate the HRQoL (MCS, PCS, and health state utilities) data. Due to the skewing of the work productivity impairment, activity impairment, healthcare resource use, and direct/indirect costs variables, a negative binomial distribution with a log-link function was used to best fit the data. Based on the data, estimated means (M), standard errors (SE), confidence intervals (CIs), and p-values were calculated for each dependent variable.

Results

Demographics

A total of 144,692 respondents were included in the analyses. On average, they were 48.2 years old, 51.6% were male, 60.5% were employed, 52.1% had less than a university education, 17.7% earned less than ¥3 million ($29,553), and 44.5% were on national health insurance (Table 1). In addition, 7.1% had a BMI that classified them as high risk, 71.5% consumed alcohol, 21.4% currently smoked, and 44.3% had exercised in the past month. The sample had an average CCI score of 0.15.

Table 1. Comparisons of sociodemographic and health characteristics between patients with PD and controls without PD.

Parameters	Bivariate comparisons – unweighted			Bivariate comparisons – weighted
	PD (n = 133)	Control (n = 144,559)	p-value	PD (n = 133)	Control (n = 125)	p-value
Age [mean ± SD/SE]
Years	61.4 ± 14.3	48.2 ± 15.3	<.001	61.4 ± 1.2	62.4 ± 0.2	<.001
Sex [count (%)]
Male	72 (54.1)	74,619 (51.6)	.562	72 (54.1)	64 (51.5)	.551
Currently employed [count (%)]
No	85 (63.9)	57,036 (39.5)	<.001	85 (63.9)	78 (62.8)	.787
Yes	48 (36.1)	87,523 (60.5)		48 (36.1)	47 (37.2)
University education [count (%)]
Less than University education	86 (64.7)	75,370 (52.1)	.004	86 (64.7)	81 (65.1)	.923
University education or higher	47 (35.3)	69,189 (47.9)		47 (35.3)	44 (34.9)
Annual household income [count (%)]
<¥3 million	29 (21.8)	25,637 (17.7)	.061	29 (21.8)	26 (20.8)	.999
¥3 million to <¥5 million	32 (24.1)	37,416 (25.9)		32 (24.1)	31 (24.9)
¥5 million to <¥8 million	21 (15.8)	36,521 (25.3)		21 (15.8)	20 (15.8)
¥8 million or more	32 (24.1)	30,345 (21.0)		32 (24.1)	30 (24.0)
Decline to answer	19 (14.3)	14,640 (10.1)		19 (14.3)	18 (14.6)
Type of insurance [count (%)]
National Health Insurance	80 (60.2)	64,318 (44.5)	<.001	80 (60.2)	74 (58.8)	.961
Social Insurance	36 (27.1)	71,749 (49.6)		36 (27.1)	37 (29.7)
Late Stage Elderly Insurance	11 (8.3)	1,902 (1.3)		11 (8.3)	9 (7.1)
Other	2 (1.5)	2,503 (1.7)		2 (1.5)	2 (1.5)
None of the above	4 (3.0)	4,087 (2.8)		4 (3.0)	4 (2.9)
Body mass index (BMI) category [count (%)]
Underweight	14 (10.5)	15,128 (10.5)	.131	14 (10.5)	13 (10.3)	.994
Acceptable risk	58 (43.6)	72,987 (50.5)		58 (43.6)	56 (45.1)
Increased risk	45 (33.8)	40,505 (28.0)		45 (33.8)	41 (32.7)
High risk	14 (10.5)	10,232 (7.1)		14 (10.5)	13 (10.1)
Decline to provide weight	2 (1.5)	5,707 (3.9)		2 (1.5)	2 (1.8)
Alcohol use [count (%)]
Abstain	43 (32.3)	41,247 (28.5)	.332	43 (32.3)	43 (34.6)	.580
Drink alcohol	90 (67.7)	103,312 (71.5)		90 (67.7)	82 (65.4)
Smoking behavior [count (%)]
Never smoker	72 (54.1)	77,876 (53.9)	.946	72 (54.1)	69 (55.2)	.749
Former smoker	34 (25.6)	35,769 (24.7)		34 (25.6)	34 (27.0)
Current smoker	27 (20.3)	30,914 (21.4)		27 (20.3)	22 (17.8)
Exercise behavior [count (%)]
Do not exercise in past month	64 (48.1)	80,592 (55.8)	.077	64 (48.1)	60 (48.1)	.994
Exercised in past month	69 (51.9)	63,967 (44.2)		69 (51.9)	65 (51.9)
CCI [mean ± SD/SE]	1.04 ± 3.64	0.15 ± 0.48	<.001	1.04 ± 0.32	0.44 ± 0.04	<.001

Cost conversion: ¥3 million = $29,553, ¥5 million = $49,254, ¥8 million = $78,807 as per conversion rates on July 1, 2014. SDs and SEs are provided for unweighted and weighted comparisons, respectively.

Abbreviations: BMI, Body Mass Index; CCI, Charlson Comorbidity Index; PD, Parkinson’s Disease; SD, Standard Deviation. SE, Standard Error.

Health-related outcomes before weighting

Based on self-reported doctor’s diagnosis of PD, the unweighted sample included 133 respondents in the PD group and 144,559 respondents categorized as controls without PD. In the unweighted sample, patients in the PD group were significantly more likely to be older, unemployed, more frequently having national health or late stage elderly insurance, and with significantly higher CCI scores, compared with the control group (p < .001 for all; Table 1).

In the unadjusted sample, patients in the PD group reported significantly lower MCS, PCS, and SF-6D health state utility scores, as well as significantly higher absenteeism, presenteeism, overall work impairment, and activity impairment than the control group (p < .001, for all). In addition, patients in the PD group reported significantly higher healthcare resource utilization and direct and indirect costs than controls without PD (p < .001, for all; Table 2).

Table 2. Effect of PD on HRQoL, healthcare resource utilization, and costs – Bivariate analysis.

	Outcomes stratified – unadjusted			Outcomes Stratified – adjusted
	PD (n = 133)	Control (n = 144,559)	p-value	PD (n = 133)	Control (n = 125)	p-value
MCS [mean ± SD/SE]	40.1 ± 12.8	47.6 ± 10.0	<.001	40.1 ± 1.1	50.1 ± 0.1	<.001
PCS [mean ± SD]	40.3 ± 8.8	52.2 ± 6.7	<.001	40.3 ± 0.8	50.5 ± 0.1	<.001
Health state utility score [mean ± SD]	0.62 ± 0.14	0.76 ± 0.12	<.001	0.62 ± 0.01	0.77 ± 0.001	<.001
Absenteeism [mean ± SD (count)]	20.6 ± 32.3	3.1 ± 12.7	<.001	20.6 ± 5.1	11.7 ± 2.5	<.001
Presenteeism [mean ± SD (count)]	44.5 ± 33.7	18.2 ± 23.2	<.001	44.5 ± 5.3	16.0 ± 0.3	<.001
Overall work impairment [mean ± SD (count)]	19.9 ± 25.3	19.9 ± 25.3	<.001	50.7 ± 5.9	18.1 ± 0.3	<.001
Activity impairment [mean ± SD]	49.7 ± 29.9	20.3 ± 24.4	<.001	49.7 ± 2.6	20.6 ± 0.4	<.001
Healthcare provider visits in past 6 months [mean ± SD]	15.9 ± 22.7	4.8 ± 8.2	<.001	15.9 ± 2.0	7.6 ± 0.2	<.001
Hospitalizations in the past 6 months [mean ± SD]	5.4 ± 18.4	0.5 ± 4.3	<.001	5.4 ± 1.6	0.9 ± 0.1	<.001
ER visits in the past 6 months [mean ± SD]	0.9 ± 4.8	0.1 ± 1.0	<.001	0.9 ± 0.4	0.2 ± 0.04	<.001
APP indirect costs ($) [mean ± SD (count)]	21,713 ± 17,158	8,891 ± 11,832	<.001	21,713 ± 2,713	7,467 ± 109	<.001
APP direct costs ($) [mean ± SD]	63,529 ± 204,261	6,513 ± 48,300	<.001	63,529 ± 17,712	11,459 ± 728	<.001
APP physician costs ($) [mean ± SD]	2,580 ± 3,685	786 ± 1,327	<.001	2,580 ± 320	1,229 ± 27.9	<.001
APP hospital costs ($) [mean ± SD]	3,265 ± 11,026	306 ± 2,594	<.001	3,265 ± 956	547 ± 38.2	<.001
APP ER costs ($) [mean ± SD]	549 ± 2,950	62.5 ± 627	<.001	549 ± 256	114 ± 22.0	<.001

SDs and SEs are provided for unweighted and weighted comparisons, respectively.Abbreviations: APP, Annual per-patient; ER, Emergency Room; PD, Parkinson’s Disease; MCS, Mental Component Summary; PCS, Physical Component Summary; SD, Standard Deviation. SE, Standard Error.

Health-related outcomes after propensity weighting

The weighted sample included a total of 258 respondents, of which 133 were in the PD group and 125 were in the control group. Among the demographic and health characteristics, only age and CCI scores were different among the groups, as patients with PD were significantly younger and had a higher CCI score than controls without PD (p < .001 for both; Table 1).

Patients in the PD group reported significantly lower MCS (40.1 vs 50.1), PCS (40.3 vs 50.5), and SF-6D health state utility scores (0.62 vs 0.77), and significantly higher absenteeism (20.6% vs 11.7%), presenteeism (44.5% vs 16.0%), overall work impairment (50.7% vs 18.1%), and activity impairment (49.7% vs 20.6%) than the control group (p < .001, for all). In addition, PD patients reported a significantly higher number of healthcare provider visits (15.9 vs 7.6), ER visits (0.89 vs 0.19), and hospitalizations (5.4 vs 0.9) in the past 6 months, compared with controls (p < .001, for all). Both direct and indirect costs were also significantly higher for patients in the PD group compared with controls without PD (p < .001; Table 2).

Multivariable analysis

After further controlling for age and CCI scores, patients in the PD group reported significantly lower MCS (35.7 vs 45.4), PCS (41.3 vs 51.3), and SF-6D health state utility scores (0.58 vs 0.73) than the control group (p < .001 for all). In addition, patients in the PD group had significantly higher absenteeism (19.3% vs 3.3%), presenteeism (45.2% vs 18.5%), overall work impairment (52.8% vs 20.3%), and activity impairment (49.6% vs 20.8%) than the control group (p < .001, for all). In terms of healthcare resource use, patients in the PD group had a significantly higher number of healthcare provider visits (8.5 vs 4.9), ER visits (0.39 vs 0.11), and hospitalizations (3.7 vs 0.6) in the past 6 months, compared with the control group (p < .001, for all). Moreover, those in the PD group incurred significantly higher direct and indirect costs than controls without PD (p < .001, for both; Table 3).

Table 3. Effect of PD on HRQoL, healthcare resource utilization, and costs after further adjustment for age and CCI between PD diagnosed respondents and controls without PD.

Dependent variable	Adjusted mean ± SE (95% CI)		p-value
	PD	Control
MCS	35.7 ± 0.1 (35.6–35.8)	45.4 ± 0.1 (45.3–45.5)	<.001
PCS	41.3 ± 0.04 (41.2–41.4)	51.3 ± 0.04 (51.3–51.4)	<.001
Health state utility score	0.58 ± 0.00 (0.58–0.58)	0.73 ± 0.00 (0.73–0.73)	<.001
Absenteeism	19.3 ± 0.6 (18.2–20.5)	3.3 ± 0.1 (3.2–3.4)	<.001
Presenteeism	45.2 ± 0.9 (43.5–47.0)	18.5 ± 0.1 (18.3–18.8)	<.001
Overall work impairment	52.8 ± 1.0 (50.8–54.8)	20.3 ± 0.1 (20.1–20.6)	<.001
Activity impairment	49.6 ± 0.5 (48.6–50.6)	20.8 ± 0.1 (20.7–21.0)	<.001
Healthcare provider visits in the past 6 months	8.5 ± 0.2 (8.0–8.9)	4.9 ± 0.03 (4.8–4.9)	<.001
Hospitalizations in the past 6 months	3.7 ± 0.2 (3.4–4.0)	0.6 ± 0.02 (0.56–0.63)	<.001
ER visits in the past 6 months	0.39 ± 0.07 (0.28–0.55)	0.11 ± 0.00 (0.10–0.12)	<.001
APP indirect costs ($)	25,355 ± 185 (24,996–25,720)	8,891 ± 62.8 (8,769–9,015)	<.001
APP direct costs ($)	37,994 ± 1,070 (35,953–40,151)	7,046 ± 180 (6,703–7,407)	<.001
APP hospital costs ($)	2,191 ± 113 (1,981–2,424)	357 ± 11.5 (335–381)	<.001
APP ER costs ($)	242 ± 30.2 (190–309)	68.7 ± 1.7 (65.5–72.1)	<.001
APP physician costs ($)	1,376 ± 34.5 (1,310–1,446)	790 ± 3.9 (783–798)	<.001

Abbreviations: APP, Annual per-patient; CI, Confidence Interval; ER, Emergency Room; PD, Parkinson’s Disease; MCS, Mental Component Summary; PCS, Physical Component Summary; SE, Standard Error.

Discussion

The present patient-reported health outcomes study is the first to provide evidence on multiple dimensions of HRQoL and economic burden in a large, nationwide cohort of patients with PD in Japan. The results showed that patients with PD had significantly lower HRQoL; higher absenteeism, presenteeism, overall work impairment, and activity impairment; increased healthcare resource use; and higher costs; compared with controls without PD. Differences in all patient-reported outcomes remained significant following propensity score weighting and further adjustment using weighted GLMs.

The present study provided evidence of the deleterious effect of PD on HRQoL, which is consistent with earlier studies that demonstrated lower HRQoL in patients with PD than for controls without PD^14,36. Notably, prior research suggests that PD has a more negative impact on well-being than diabetes mellitus, thereby underscoring the severity of PD on patient HRQoL³⁷. Treatment strategies effective in delaying disease progression could potentially improve HRQoL.

The present study showed almost 6.0-fold and 2.5-fold higher absenteeism and presenteeism, respectively, in patients with PD, compared with controls without PD, after adjusting for confounding variables. These results suggest that PD is independently associated with considerable work productivity loss. These findings were consistent with a study that combined information from nationally representative surveys, which reported that employed patients with PD missed eight additional work days per year¹⁷. As productivity loss accounts for a share of the economic burden imposed by PD, treatments and healthcare programs with the potential for returning patients to higher productivity are needed. Development of new outpatient services may reduce the need for hospitalization, thereby reducing the number of days patients are absent from work.

In the current study, the number of healthcare provider visits, hospitalizations, and ER visits in the past 6 months reported by patients with PD were 1.6-times, 1.7-times, and 2.7-times higher than the control group, respectively. These results are in accordance with a retrospective Canadian study that reported the number of physician visits by patients with PD were 1.6-times higher per year than those of controls without PD¹⁹. Moreover, previous research indicated patients with PD may also have increased utilization of ancillary and community services, such as home healthcare providers, occupational therapists, physiotherapists, visiting general practitioners and nurses, congregate meals, transportation services, and home chores services²¹.

PD imposes a substantial economic burden to patients, the healthcare system, and society. Both the annual per-patient direct and indirect costs in this study were found to be 5.4-fold and 2.9-fold higher, respectively, in patients with PD than in controls without PD. Although the study did not statistically compare the difference between direct and indirect costs, the direct costs associated with PD were numerically higher than the indirect costs. The main factor contributing to direct costs was hospitalization costs, followed by costs for healthcare provider and ER visits. The present study’s estimate of $37,994 (¥3,856,921) in total annual per-patient direct costs is higher than those reported by studies conducted in the US by Huse et al.²⁴ ($23,101), Kowal et al.¹⁷ ($12,800), and O’Brien et al.¹⁶ ($12,491). Also, our estimated annual per-patient indirect costs of $25,355 (¥2,573,939) are higher than those of O’Brien et al.¹⁶ ($9,135). There are several possible explanations for these discrepancies, such as varying study populations, different patterns of treatment practices, and difference in gross national income per capita among various countries.

Ultimately, improved treatment of PD may have a substantial impact on ameliorating HRQoL and reducing associated costs. Nevertheless, a prior study indicated significantly lower HRQoL in almost 70% of Japanese patients with PD due to sub-optimal treatment in which the duration of the effect or dosage was insufficient. Therefore, it is imperative for physicians to raise greater awareness among patients regarding insufficient drug efficacy, and to offer treatment with suitable medications, which can help facilitate better HRQoL and better disease management in patients with PD³⁸.

The current study has limitations, including the following. The diagnosis of Parkinson’s disease and other survey responses were self-reported, and data could not be independently verified. However, the survey was low-stakes, and the questions were relatively benign, as the NHWS does not contain items that would be deemed offensive or stressful. In addition, the survey was confidential, diminishing the incentive to misrepresent one’s reporting. As the data were collected in 2009–2014, it is possible somewhat different results would be obtained with more recent data. Although relevant health and demographic characteristics were controlled by weighting and inclusion in multivariable models, the possibility that the observed pattern of results may be associated with other variables not included in the analyses cannot be excluded. Additionally, causal inferences cannot be drawn from the study results, and some types of variables cannot be assessed over the longer-term due to the cross-sectional nature of the data. Although the NHWS demographically represents the population of adults in Japan, it is unclear whether it adequately represents the adult population with PD in Japan.

Conclusions

PD in Japan poses a significant burden on patients due to the decline in HRQoL, greater work productivity loss leading to higher indirect costs, and increased healthcare resource utilization resulting in higher direct costs. Our results suggest that there are unmet needs among the PD population in Japan. Improved management of PD could, therefore, be beneficial to both the patient and society.

Transparency

Declaration of funding

This study was funded by Takeda Pharmaceutical Company Limited.

Declaration of financial/other interests

KY is an employee of Takeda Pharmaceutical Company Limited. RL, at the time of this study, was an employee of Kantar Health, a paid consultant of Takeda Pharmaceutical Company Limited. NF, at the time of this study, was an employee of Kantar Health, a paid consultant of Takeda Pharmaceutical Company Limited. CP is an employee of Takeda Pharmaceuticals International Co. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

The authors acknowledge Leo J. Philip Tharappel and Ramu Periyasamy, PhD, Indegene Pvt Ltd. for assistance with literature review and writing, as well as Martine C. Maculaitis, PhD, for editing assistance on behalf of Kantar Health.