[The prevalence of Parkinson disease in West Germany--are general practice data a suitable survey instrument?]

Investigations on prevalence rates in idiopathic Parkinson's disease (IPD) have raised fairly different results due to variations in methodology. Door-to-door studies in circumscribed geographical areas revealed lower prevalence rates in China and Japan compared to countries of the Western hemisphere. Due to the health system organization and present legislation on personal data referral such systematic door-to-door epidemiological surveys seem nearly impossible in the Federal Republic of Germany. The present pilot study therefore used doctor's practices as a means to obtain evidence rates on patients diagnosed on and treated for IPD. Standardized questionnaires with appropriate items regarding symptoms, diagnosis and treatment for the disease were distributed by mail or personally in 387 practices of general practitioners, internal specialists, neurologists, and psychiatrists and 3 hospital ambulances in the handestadt Lübeck, the Kreis Stormarn and the Kreis Herzogtum Lauenburg. Investigation time was March, 1 to June, 30, 1987. Each patient cared for IPD was to fill up a personal questionnaire with similar items as well. A second estimation of prevalence was made according to published sales of L-Dopa drugs for treatment of IPD. The third estimation used statistics about disease pattern of the Bundesverband der Allgemeinen Ortskrankenkassen (AOK). Participation of doctors was to be maintained high by varying methods of repeated requirements during the investigation period. 41 of 210 (20%) of the doctors requested by mail and 126 of the 180 visited physicians (70%) participated in the study with 1018 patients altogether. The estimate of prevalence for the Hansestadt Lübeck was 188, for the Kreis Stormarn 206 and for the Kreis Herzogtum Lauenburg 147 patients per 100,000 (183 in toto). According to latest data on L-Dopa drug sale the prevalence rate in the above areas was 144/100,000. According to the statistics of disease pattern of the AOK the prevalence estimate was 185/100,00 assured persons. These results are discussed with special emphasis on the methodological aspects.