Underreporting of malaria incidence in The Netherlands: results from a capture-recapture study.

The aim of this study was to estimate the completeness of notification of malaria by physicians and laboratories in the Netherlands in 1996. We used a capture-recapture (CRC) analysis of three incomplete, partially overlapping registers of malaria cases: a laboratory survey, the Notification Office and the hospital admission registration. The response of the laboratories was 83.2%. In 1996 the laboratories microscopically identified 535 cases of malaria, 330 patients with malaria were admitted to hospital and physicians notified 311 malaria cases. 667 malaria cases were recorded in at least one register. CRC analysis estimated the total number of malaria cases at 774 (95 % CI of 740-821). This implies a completeness of notification of 40.2% for physicians and 69.1% for the laboratories. It can be concluded that laboratory-based notification can considerably increase the number of officially reported malaria cases as compared to notification by physicians. However, possibly one-third of the cases may still go unreported.