Abstract:
|
Advancements in linking publicly available census records with vital and administrative records have enabled novel investigations in epidemiology and social history. However, linkage may only be successful for a subset of the census cohort due to limited access to unique identifiers and error prone matching variables. In survival analysis, differential ascertainment of event times can negatively impact inference on risk associations. Motivated by this problem, we conduct a simulation study to compare the performance of five analytic approaches to estimating the effects of exposure on survival in the presence of missing event times and equivocal matches. We also explore the effects of different missing data mechanisms and matching success rates on their performance. The methods are applied to a historic cohort of residents in Ambler, PA established using the 1930 US census, from which only 2,440 out of 4,514 individuals (54%) had death records retrievable from publicly available data sources and death certificates. Using this cohort, we examine the effects of occupational and paraoccupational asbestos exposure on survival and disparities in mortality by race and gender.
|