Record-linkage between two anonymous databases for a capture-recapture estimation of underreporting of AIDS cases: France 1990-1993

Abstract

Objective To estimate the completeness of the French mandatory AIDS surveillance system (Déclaration Obligatoire DO) over the 1990–1993 period using a capture-recapture approach, by matching the mandatory reports with the AIDS cases present in the French Hospital Database on HIV infection (FHDH). Methods An anonymous record-linkage algorithm was developed to identify those cases common to both anonymous surveillance systems. The linkage was based on sex, date of birth, and infection risk group, all strictly matched, and on the dates of AIDS diagnosis and of death, the places of diagnosis and residence, and the AIDS-defining diseases at diagnosis. The total number of AIDS cases and completeness of both surveillance systems were estimated using a capture-recapture approach, assuming independence of the ascertainment sources. Results The completeness of the mandatory reporting was estimated at 83.6% (95% CI : 82.9–84.3), and that of the FHDH at 47.6% (95% CI : 46.9–48.3) for the surveillance of AIDS cases diagnosed among adults in France between 1990 and 1993. The completeness of the system based on FHDH increased over the study period as more hospitals joined the project, while the completeness of the DO surveillance system remained stable. Conclusion This approach was useful in estimating the underreporting of AIDS cases in France. Regularly performed, it will allow the impact of underreporting to be monitored over time.