Incorporating data on residential history for Disease Mapping

When studying the relationship between an individual’s location and the acquisition of disease, the location to use is not always clear. When location at exposure is different from that at diagnosis, the latter may not represent the relevant information. While time and

location of exposure are often unknown, residential history of cases can substantially inform a spatial analysis. In spatial surveillance, spatial data on cases are often used to detect and locate subareas of the study region with higher or lower risk of disease. Current literature

has adapted detection methods to incorporate residential history of cases where available.

We extend a disease mapping method to incorporate such data. Through simulations we show that our method is more accurate at identifying a localized increased risk of disease when compared to mapping when only location at diagnosis is considered.