Parkinson's prevalence estimated by a state registry.

A solid understanding of the descriptive epidemiology of a disease is essential in etiologic investigations; this includes prevalence and incidence, as well as groups within the larger community who may have noticeably lower or higher rates. We ascertained the usefulness of a non-traditional registry in describing Parkinson's disease (PD) patterns in a community. A passive surveillance PD registry in Nebraska began data collection on 1 January 1997. All physicians were required to report PD diagnosis, pharmacists reported new prescriptions of anti-PD drugs (PD cases were confirmed later with the prescribing physician), and there was a patient self-report mechanism. The overlap of reporting by the sources allowed estimation of the number not reported by any source, using the statistical technique "capture-recapture." As of January 2000, the Nebraska PD Registry had reports of 5,062 PD patients. The number not reported by any Registry reporting source was calculated to be 117, leading to an estimated total of 5,179 cases and a prevalence of 329.3 per 100,000 population. Tabulations of age- and gender-specific prevalence rates, as well as county-level estimates, allow examination of areas of elevated or lowered prevalence. The combination of a passive surveillance system and capture-recapture technique presents a useful method for epidemiologic description, and more traditional survey methods could benefit by including capture-recapture capability. Copyright 2003 Movement Disorder Society