Abstract
Introduction – The Registry of Congenital Malformations (CM) in Reunion Island (REMACOR) oversees the surveillance of CM for public health and research purposes. This surveillance consists of the exhaustive collec- tion of cases, and temporal analysis of CM. Since 2013, REMACOR has retrospectively geolocalized the data to allow their spatial analysis and the detection of health signals the methodology of which is presented hereunder.Materials and methods – Each case is geolocated according to the mother’s address, using the reference address databases. Birth-related prevalence is then calculated by aggregating cases according to the different administrative scales. The choice of the suitable scale is determined using a Poisson test that estimates the minimum number of births required to perform the analysis in a statistically signi cant manner, and then a compromise between this representativeness of information, data completeness and spatial resolution.Results – 95% of the cases could be geolocalized. Different cluster detection methods are used to identify the most affected areas. Finally, clusters detected at different scales are nally intersected to calculate an index of belonging to 1, 2 or 3 scales.Discussion and conclusion – The spatialized database set up now allows REMACOR to consider the spatial heterogeneity of the distribution of the most frequent CM in order to inform public health stakeholders.