The Québec BCG Vaccination Registry (1956–1992): assessing data quality and linkage with administrative health databases

BACKGROUND: Vaccination registries have undoubtedly proven useful for estimating vaccination coverage as well as examining vaccine safety and effectiveness. However- their use for population health research is often limited. The Bacillus Calmette-Gu̩rin (BCG) Vaccination Registry for the Canadian province of Qu̩bec comprises some 4 million vaccination records (1926-1992). This registry represents a unique opportunity to study potential associations between BCG vaccination and various health outcomes. So far- such studies have been hampered by the absence of a computerized version of the registry. We determined the completeness and accuracy of the recently computerized BCG Vaccination Registry- as well as examined its linkability with demographic and administrative medical databases. METHODS: Two systematically selected verification samples- each representing ~{}0.1% of the registry- were used to ascertain accuracy and completeness of the electronic BCG Vaccination Registry. Agreement between the paper listings (n = 4-987 records) and vaccination certificates (n = 4-709 records) and electronic formats was determined along several nominal and BCG-related variables. Linkage feasibility with the Birth Registry (probabilistic approach) and provincial Healthcare Registration File (deterministic approach) was examined using nominal identifiers for a random sample of 3-500 individuals born from 1961 to 1974 and BCG vaccinated between 1970 and 1974. RESULTS: Exact agreement was observed for 99.6% and 81.5% of records upon comparing- respectively- the paper listings and vaccination certificates to their corresponding computerized records. The proportion of successful linkage was 77% with the Birth Registry- 70% with the Healthcare Registration File- 57% with both- and varied by birth year. CONCLUSIONS: Computerization of this Registry yielded excellent results. The registry was complete and accurate- and linkage with administrative databases was highly feasible. This study represents the first step towards assembling large scale population-based epidemiological studies which will enable filling important knowledge gaps on the potential health effects of early life non-specific stimulation of the immune function- as resulting from BCG vaccination.