To compare differences in health care resource utilization (HcRU) over time between Medicare beneficiaries with and without Parkinson disease (PD).
This retrospective observational study used the Chronic Conditions Data Warehouse (5% Medicare sample) between 2005 and 2015. In a propensity score–matched (age, sex, race, and comorbidity adjusted) sample of beneficiaries with and without PD, we examined all-cause HcRU due to inpatient admissions, emergency department (ED) admissions, skilled nursing facility (SNF) admissions, health care provider encounters, neurologist visits, rehabilitation service visits, and non-PD medication fills. Relative to beneficiaries without PD, we reported adjusted incidence rate ratios (IRRs) and 95% confidence intervals (CIs) for beneficiaries with PD using generalized linear models with log link and negative binomial variance functions.
A total of 467,064 Medicare enrollees (unmatched sample) met the inclusion criteria. Of these, 3.3% had PD. In the matched sample and relative to beneficiaries without PD, beneficiaries with PD displayed higher rates of inpatient admissions (IRR 1.29, 95% CI 1.24–1.34), ED admissions (IRR 1.31, 95% CI 1.27–1.34), SNF admissions (IRR 2.00, 95% CI 1.92–2.09), health care provider encounters (IRR 1.18, 95% CI 1.16–1.20), neurologist visits (IRR 5.57, 95% CI 5.35–5.78), rehabilitation service visits (IRR 1.47, 95% CI 1.41–1.53), and non-PD medication fills (IRR 1.10, 95% CI 1.08–1.11) over time.
These results reflect patterns of medical care among Medicare beneficiaries with PD. The findings can help clinicians, payers, and policy makers make evidence-based decisions for the allocation of scarce health care resources for PD management.
Classification of Evidence
This study provides Class II evidence that Medicare beneficiaries with PD use more health care resources than matched controls without PD.