Estimating Problem Drug Use Prevalence at National Level: Comparison of Three Methods

• NCJ Number: 214681
• Journal: Drugs: Education, Prevention and Policy Volume: 13 Issue: 2 Dated: April 2006 Pages: 109-120
• Author(s): Filip Smit; Margriet van Laar; Lucas Wiessing
• Date Published: April 2006
• Length: 12 pages
• Grant Number(s): TSER/DG-12
• Sponsoring Agency: European Cmssn
• Publication Format: Article
• Publication Type: Research (Applied/Empirical)
• Language: English
• Country: United Kingdom

Annotation

This paper presents a comparative analysis of three national methods in the prevalence estimation of drug use in the Netherlands.

Abstract

All three models produced similar results in the prevalence estimation of drug use. The annual prevalence as estimated with the treatment multiplier was 0.30 percent and 0.31 percent for the regression imputation model. The multiple imputation method came up with similar results, 0.32 percent. The results indicate that the prevalence rate in 2001 was between 0.22 percent and 0.43 percent for the population aged 15-64. The use of multiple imputation in conjunction with regression imputation techniques for getting prevalence estimates of problem drug use is encouraged. Utilizing two established methods for prevalence estimation, treatment multiplier and the multivariate indicator method, a cross-validation is conducted on the outcomes from a third method, multiple imputation. The data used in this study were from the Netherlands. Tables, references