Parkinson's disease (PD) is a progressive neurodegenerative disorder affecting millions worldwide. In the United States alone, approximately one million individuals were diagnosed with PD in 2017. Beyond the debilitating motor symptoms (tremor, bradykinesia), PD patients experience significant comorbidities (infections, cardiac and gastrointestinal disorders, fall-related injuries), impacting their healthcare needs, work ability, and requiring substantial caregiving support. These direct and indirect consequences generate substantial economic burdens. While PD is a rapidly growing neurological disorder, its economic impact in high-income nations remains poorly understood. This study aimed to comprehensively quantify the current and projected economic burden of PD in the U.S. Utilizing a human capital approach, the study assigned monetary values to health loss, encompassing productivity losses due to illness, disability, and premature mortality. This included less well-understood cost components like future earnings losses from premature death, productivity losses in both work and social life, and caregiver burdens. The study relied on a multi-source data approach, combining large, nationally representative claims databases from public and private payers with a primary survey (Social and Financial Impact of Parkinson's Disease Survey, n=4,548) to capture indirect productivity losses for patients and caregivers, disability benefits, and non-medical costs (home renovations, vehicle modifications, etc.).

The introduction briefly mentions previous research highlighting the burden of PD on society, families, and individuals, citing Whetten-Goldstein et al.'s finding of an average of 22 hours per week spent by family caregivers. However, a more extensive review of existing literature on the economic burden of PD in the U.S. and other high-income countries is lacking in the introduction. The discussion section compares the current study's findings to several previous U.S. studies (Kowal et al., O'Brien et al., Huse et al., Noyes et al., Dieleman et al.), highlighting discrepancies in prevalence and cost estimates primarily due to methodological differences (data sources, cost calculation approaches, inclusion/exclusion of cost components). These previous studies are summarized in Table 3, showing variations in prevalence estimates (ranging from 500,000 to 647,000 in previous studies versus 1,040,000 in the current study), total costs, and per capita costs. The current study differentiates itself by using more recent data, a more comprehensive primary survey, and a more rigorous methodology for estimating direct medical costs (comparing PD patients to matched controls).

This study employed a prevalence-based approach to estimate the economic burden of PD in 2017. The estimated number of individuals with PD was combined with per-capita costs to derive national economic burden estimates stratified by population characteristics. The methodology is detailed in Figure 4 and includes several key steps: 1. **PD Prevalence Estimation:** Age-, gender-, and insurance-specific prevalence rates were calculated using 2015 Medicare Current Beneficiary Survey (MCBS) data for Medicare-eligible individuals and combined 2011-2015 Medical Expenditure Panel Survey (MEPS) data for other insurance types. For those under 65 not covered by Medicare, PD cases were identified based on the presence of any PD diagnosis in MEPS chronic condition files using ICD-9 code 332 (which may include some cases of secondary parkinsonism). The MCBS identified PD cases based on self-reported diagnosis. 2. **Direct Medical Cost Estimation:** Excess healthcare costs attributable to PD were calculated by comparing the average costs of PWPs to matched control groups (without PD, matched on age, gender, race/ethnicity, and insurance). Three data sources were used: Optum de-identified Normative Health Information (dNHI) system for privately insured individuals under 65, Medicare Standard Analytical File (SAF) 5% sample claims data and MCBS for Medicare-eligible individuals. For the small group of individuals under 65 with non-private, non-Medicare insurance, costs were imputed from the privately insured group, adjusting for potential cost differences using a generalized linear model (GLM). PWPs were identified in claims data based on PD diagnosis codes (including related neurodegenerative conditions to minimize false negatives), requiring at least one inpatient claim or at least two non-inpatient claims. 3. **Indirect and Non-medical Cost Estimation:** The Social and Financial Impact of Parkinson's Disease Survey (n=4,548) provided data on indirect costs (lost productivity due to reduced employment, absenteeism, presenteeism, social productivity loss) and non-medical costs (paid daily non-medical care, home modifications, vehicle modifications). Future earnings loss due to premature death was calculated using CDC WONDER data and Medicare SAF data. Survey weights were calculated to ensure representativeness of the U.S. PD population. 4. **Future Projections:** Future PD prevalence was projected by applying the 2017 age- and gender-specific prevalence rates to U.S. Census Bureau population projections (2018-2037). Future economic burden was projected by combining projected prevalence with estimated 2017 per-PWP costs. The projections are in 2017 real dollars, without adjusting for inflation or changes in healthcare utilization or prices.

The study estimated that 1.04 million individuals in the U.S. had been diagnosed with PD in 2017. The total economic burden was estimated at $51.9 billion, comprising $25.4 billion in direct medical costs and $26.5 billion in indirect and non-medical costs. Medicare covered over 90% of the direct medical costs. The largest cost categories within direct medical costs were hospital inpatient care (28.4%), non-acute institutional care (28.2%), and outpatient care (21.7%). Indirect costs included substantial losses in productivity for both PWPs and caregivers. For PWPs, the largest indirect cost categories were future earnings loss due to premature death, earnings loss from reduced employment, and absenteeism. For caregivers, absenteeism and presenteeism represented the largest losses. Non-medical costs were substantial, with paid daily non-medical care and home modifications being the largest components. The study projected that by 2037, the number of PWPs would increase to 1.64 million, with a total economic burden exceeding $79 billion. Direct medical costs and social productivity losses were projected to increase by about 52%, while other cost components (disability income, paid care, other non-medical costs, caregiver productivity loss) were projected to increase by approximately 50%.

The findings of this study provide a significantly higher estimate of the economic burden of PD compared to previous studies. This is largely due to methodological improvements, including the use of more recent and comprehensive data sources and a more robust approach to cost estimation. The higher estimates emphasize the substantial societal burden of PD, impacting patients, caregivers, employers, and payers. The projections underscore the urgent need for interventions that can reduce PD incidence, delay disease progression, and alleviate symptoms, thus mitigating the future economic impact. The study acknowledges limitations related to potential underestimation of undiagnosed cases, potential misclassification of PD cases in the data sources used, and the small sample size for some subgroups, limiting the ability to analyze racial and ethnic differences. The imputation of costs for a small subgroup of PWPs also represents a limitation.

This study offers a comprehensive and updated assessment of the substantial economic burden of Parkinson's disease in the U.S., revealing a significantly higher cost than previously reported. The study's projections highlight the escalating financial strain anticipated in the coming decades. The results underscore the urgent need for research into preventive measures and disease-modifying therapies, alongside policy initiatives to support affected individuals and families. Future research should focus on refining cost estimations by addressing the study's identified limitations, including focusing on undiagnosed cases and more detailed subgroup analyses.

The study acknowledges several limitations: underestimation of undiagnosed PD cases; potential misclassification of PD cases due to reliance on various diagnostic codes and data sources; small sample size for some subgroups, particularly those under 65 with non-Medicare and non-private insurance, limiting detailed analyses of racial and ethnic cost differences; imputation of costs for a subgroup of PWPs under 65 with non-private and non-Medicare insurance, which might introduce uncertainty; potential underestimation of long-term care costs due to data limitations; and the projections assume constant incidence and mortality rates.