Potential bias in pharmacoepidemiological studies due to the length of the drug free period: a study on antidepressant drug use in adults in the Netherlands

Purpose The aim of this study was to evaluate the effect of the length of the drug free period on incidence measurements as well as on cohort characteristics in users of antidepressants.

Methods The study population consisted of patients aged 18 years or older who filled a prescription for an antidepressant drug in the Netherlands, between October 2001 and September 2002. One-year incidence of antidepressant drug use was estimated using drug free periods varying in length from I month to 9 years. In addition, we evaluated what effect the drug free period has on cohort characteristics by comparing a cohort of first time antidepressant drug users defined using a 9-year drug free period with cohorts using 6, 12 and 24 months drug free period.

Results When using a 6-month drug free period the measured incidence was about 32 per 1000 individuals (95%CI: 31.3, 32.6) while the measured incidence was 27.5 (95%CI: 26.9,28.1),23.5 (95%CI: 22.9,24.0) and 17.2 (95%CI: 16.7,17.7) per 1000 individuals when using a 12-month, 24 month respectively a 9-year drug free period. Furthermore, the prevalence of characteristics in inception cohort studies changes when using different drug free periods.

Conclusion Altering the drug free period from a short to a longer one results in decreased incidence. Furthermore, for inception cohorts where first time drug use is an inclusion criterion the drug free period can influence the prevalence of cohort characteristics and for short drug free periods give biased estimates. Copyright (c) 2006 John Wiley & Sons, Ltd.