Longitudinal Study of Treatment Variability for Parkinson's Disease across Specialized Centers.

BACKGROUND: Real-world evidence on treatment practices in Parkinson's disease (PD) has been limited due to the difficulty in collecting comprehensive and generalizable clinical data. OBJECTIVES: We sought to identify treatment patterns and test how treatment changed in response to (1) falling, (2) worsening disease, and (3) worsening quality of life across PD specialized centers. METHODS: We used the Parkinson Outcomes Project data collected from 2010 to 2023 across 31 international PD specialized centers. Demographic and clinical characteristics were collected annually and included medication use, physical therapy referral, psychologist or psychiatrist care, and deep brain stimulation (DBS) surgery. Treatment practice variation was described by center and in response to outcomes (self-reported falls, higher Hoehn and Yahr stage, worse emotional and mobility subscale scores on quality-of-life scale). RESULTS: A total of 12,664 participants were analyzed. Treatment practices varied substantially across centers with the use of levodopa in the first 5 years of disease ranging from 59.3% to 94.6% and physical therapy referral ranging from 13% to 71%. At ≥ 5 years of disease, DBS rates varied from 2% to 41%. After a fall, individuals were more likely to be referred for physical therapy (β: 0.44, 95% confidence interval [CI]: 0.36, 0.52), and mental health services were recommended after a decline in emotional subscores (β: 1.74, 95% CI: 1.50, 1.98). However, there was no change in levodopa-equivalent daily dose after worsening mobility subscores (β: -29.97, 95% CI: -76.67, 16.73). CONCLUSIONS: These results highlight the large variability in PD practice across specialty centers and the importance of establishing best practice guidelines. Understanding the drivers of this variability is an essential next step.

Duke Scholars

Author Sheng Luo Biostatistics & Bioinformatics, Division of Biostatistics