APH 2001, 59, 77-100:
of problematic drug use.
Methodological aspects and feasibility in Belgium.
and D. Walckiers
Keywords: cancer, epidemiology, cancer registry, Belgium
|Objective: To assess the feasibility in Belgium of
methods designed to estimate the extent of problematic drug use.
Method: The number of prevalent cases of problematic drug use may be estimated from information including data from police services, data from hospitals and specialised centres for the treatment of drug addicts, data from death certifications, data from the HIV/AIDS register, and data related to methadone consumption. The size of this population, at the national level, may be obtained with five calculation methods including (1) the multiplier method, (2) the capture-recapture technique, (3) the use of the basic relation between prevalence, incidence, and duration of a disease, (4) the use of methadone consumption data and (5) a multivariate method based on a principal component analysis of drug-related indicators. Availability of theses data and feasebility of these methods in Belgium are investigated.
Results: Currently, the Belgian HIV/AIDS register appears to be the only source of data that could allow an estimate of the number of injecting drug users in Belgium. All the other sources of information do not provide enough reliable data. The major reasons for this situation are the law on privacy, which makes difficult to match persons appearing in different lists, lack of standardisation of the data on treament demands recorded by the different monitoring systems operating in the different Regions and/or Communities of the country, and lack of knowledge of the coverage registration rate of these monitoring systems.
Conclusion: A robust and consistent estimate of the prevalence of problematic drug use should ideally be based on estimates obtained by at least two different methods. Such robust estimate could not presently be achieved in Belgium and efforts must consequently be made to bridge this gap. In order to quickly obtain national prevalence estimates, the use of data on treatment demand appears to be the most promising approach in terms of feasibility.