Estimating the number of opiate users in Rotterdam using statistical models for incomplete count data

Abstract

Often, the number of opiate users in a city is not known. This lack of knowledge hampers adequate policy making. However, the unknown size of the opiate using population can be estimated with help of statistics. We review seven statistical models and apply them to a real data-set. All models are based on a single data-set. This dataset is incomplete in the sense that only opiate users who apply for methadone maintenance are registered while all others are not. The only information which is available in this data-set is the number of visits of uniquely identifiable persons who are on methadone maintenance, plus a set of covariates like age, sex and so on for each of these persons. Still, from this single and incomplete data-set inferences about the total population size can be made. The estimates are compared and the pro's and con's of the estimators are evaluated.