Abstract
OBJECTIVE: The purpose of this study is to examine the prevalence of indications of alcohol or drug use disorders in five different national Swedish registers and to investigate the correlation between these registers. Furthermore, the intent is to investigate whether combining data from different registers increases the prevalence of these indications in the population due to the identification of different demographic groups in different registers. DESIGN: Cross-sectional study. SETTING AND PARTICIPANTS: Individuals living in Sweden aged 20-64 years in 2006, n=5 453 616. PRIMARY OUTCOME: National registers included the Registers of Inpatient Care, Outpatient Care, Medications, Social Insurance and Convictions. Demographic variables were sex, age, migrant status, education and civil status. Indications of alcohol or drug use disorders were presented as prevalence in percentage (%), correlation was examined using phi correlation coefficients and differences across demographic factors were studied using logistic regression. RESULTS: The prevalence of an indication of alcohol or drug use disorder varied between registers, meaning that prevalence increased when all registers were considered together. The prevalence of alcohol use disorder increased by 60% and 66% among men and women, respectively, while the prevalence of drug use disorder increased by 45% and 80% among men and women, respectively, when all registers were combined, compared with only using the register with the highest prevalence. Registers contributed different indications of drug and alcohol use disorders. CONCLUSIONS: Accurate estimates of alcohol or drug use disorders are critical for healthcare and rehabilitation. This study shows that using a single register alone underestimates the burden of disease unevenly, while combining a range of registers can provide a more accurate picture.