Advantages and problems with pregnancy registries: observations and surprises throughout the life of the International Lamotrigine Pregnancy Registry.

PURPOSE: The International Lamotrigine Pregnancy Registry monitored for a signal of a substantial increase in the frequency of major congenital malformations associated with lamotrigine exposures in pregnancy over an 18-year period. Key methodological lessons are discussed.
METHODS: The strengths and weaknesses of the Registry were assessed using quantifiable methodological and operational parameters including enrollment, completeness of exposure and outcome data reporting, and lost to follow-up. The choice of comparator groups and stopping rules for registry closure were critically evaluated.
RESULTS: The reliance on voluntary reporting was associated with a clustered geographical distribution of registered pregnancies. The enrollment rate increased over time with new approvals and indications for lamotrigine and publication of interim data. Reporter burden was minimized through a streamlined data collection approach resulting in a high level of completeness of exposure and primary outcome data. Lost to follow-up rates were high (28.5% overall) representing a major limitation; incentives to increase the completeness of reporting failed to reduce rates. A lack of an internal comparator group complicated data interpretation; but external comparisons with multiple external groups allowed an assessment of consistency of outcome data across multiple data sources. A lack of a priori closure criteria prolonged the life of the Registry, and consideration of regulatory guidelines on this subject is encouraged at the time of conception of future registries.
CONCLUSIONS: A successful pregnancy exposure registry requires ongoing flexibility and continuous re-assessment of enrollment, recruitment, and retention methods and the availability of comparison data, throughout its lifecycle.