Sleep Medicine Resource Utilization in Individuals With Parkinson Disease: A Population Study of Health Administrative Data.

Abstract

Background and objectives: Individuals with Parkinson disease (PD) may face barriers in obstructive sleep apnea diagnosis/management due to sleep being lower priority and sleep disturbances being poorly recognized. Evidence on sleep-medicine service utilization in the PD population is lacking. We conducted a population-based study to identify discrepancies in sleep-medicine resource use (prevalence rates of polysomnograms [PSG] performed and positive airway pressure [PAP] initiated) over a 10-year period between the PD population and the matched non-PD population.

Methods: We conducted a retrospective longitudinal population-based study using health administrative databases in Ontario from 2012 to 2021 in adults with PD to compare overall and annual prevalence rates of PSGs performed and PAP initiated to 1:1 randomly selected propensity score matched controls from the matched non-PD population based on simultaneous exact matching on age, sex, and calendar year and caliper matched propensity scores from a logistic regression model (based on sociodemographic variables and comorbidities) using validated health administrative definitions to identify PD cases and controls. We hypothesized that the PD population has lower rates of PSGs performed and PAP treatments initiated compared with the similar matched non-PD population. We used Poisson regression to estimate annual prevalence rate ratios to determine the relative change in prevalence over the study period between the groups.

Results: Sixty-five thousand, one hundred sixty-seven patients with PD and 11,460,672 controls met our inclusion criteria. We successfully propensity score matched 64,879 PD cases to controls. From 2012 to 2021, there were a higher prevalence of any PSG performed in the PD population (8.2% vs 6.3%, [RR: 1.30, 95% CI: 1.25-1.35], p < 0.001) and no difference in the rates of any PAP initiated in the PD population vs controls (4.0% vs 4.1%, [RR: 0.93-1.03, 95% CI: 0.93-1.03], p = 0.46). For both groups, annual prevalence rates generally increased over time. There was no difference in the annual prevalence rate ratio of any PSG performed or any PAP initiated in the PD population vs controls (1.07 [95% CI: 1.06-1.07] vs 1.07 [95% CI: 1.07-1.08], p = 0.5; 1.10 [95% CI: 1.09-1.11] vs 1.11 [95% CI: 1.10-1.11], p = 0.18, respectively).

Discussion: Sleep-medicine resource utilization in the PD population is at least similar to the matched non-PD population and follows the increase with time observed in the general population.