Identifying incident cases of parkinsonism among veterans using a tertiary medical center

Abstract

To better understand the impact of incident Parkinson's disease (PD) on the Veteran's Health Administration (VHA) and to develop methods applicable to future epidemiological research, we performed a medical record review study at a tertiary referral VHA medical center. Searching the local data base, we identified 782 veterans with diagnostic codes for PD or secondary parkinsonism (SP) between 1998 and 2000. Based on structured medical record review, a movement disorders specialist confirmed diagnoses for incident parkinsonism cases. Among the 782, 191 incident parkinsonism cases were identified (100 PD, 75 SP, and 16 Parkinson's Plus). Incident PD cases were older at diagnosis (74.5 vs. 70.4 yr; P < 0.05) and more likely to be white (81% vs. 62; P < 0.07) than incident SP cases. Diagnostic codes were insufficient to distinguish between incident PD and SP (positive predictive value, 57% and 39%, respectively), and VHA sources failed to identify 21% of confirmed deaths among the incident PD cohort by November 2004. Although the large number of incident cases identified suggests PD is an important cause of disability among elderly VHA users, observed limitations of VHA sources for identifying incident PD cases and determining their vital status should be considered when designing future studies.