Comparative outcomes and healthcare costs of three device-aided pharmacotherapies for patients with advanced Parkinson's disease in Sweden: A nationwide registry-based cohort study.

BACKGROUND: Patients with advanced Parkinson's disease often require treatment beyond conventional oral or transdermal therapy. However, real-world evidence across emerging advanced treatment strategies remains limited. OBJECTIVES: To compare patient characteristics, treatment patterns, clinical outcomes, healthcare resource utilization (HCRU), and costs among Parkinson's disease patients treated with levodopa/carbidopa dispersible microtablets (LC-5), levodopa-carbidopa intestinal gel (LCIG)/levodopa-entacapone-carbidopa intestinal gel (LECIG), or continuous subcutaneous apomorphine infusion (CSAI). METHODS: Patients initiating LC-5, LCIG/LECIG, or CSAI between 2014 and 2024 were identified in Swedish national register data. Treatment groups were balanced using inverse probability weighting. Descriptive analyses, cumulative incidence estimates, and Cox proportional hazards models were used to assess outcomes. RESULTS: 426 LC-5, 1112 LCIG/LECIG, and 490 CSAI users were included. 63.6% of LC-5 users were women, whereas men constituted the majority in the other treatment groups. 3.5% of LCIG/LECIG and 6.1% of CSAI users discontinued treatment, compared with 28.6% of LC-5 users. Compared with LCIG/LECIG users, the mortality hazard ratio estimates were lower for LC-5 (HR 0.71, 95% CI 0.45-1.11) and CSAI users (HR 0.80, 95% CI 0.63-1.02), but neither difference was statistically significant. In the 12 months following treatment initiation, weighted estimates of healthcare costs were higher among LCIG/LECIG (€37,613) and CSAI users (€29,214) than among LC-5 users (€17,290). CONCLUSIONS: In this nationwide Swedish cohort, no statistically significant mortality difference was detected. Weighted one-year HCRU and cost estimates were lower among LC-5 users than among infusion-therapy users; however, residual confounding by disease severity, treatment selection, and other unmeasured clinical factors limits causal interpretation.