Predictors of the incidence and discontinuation of long-term use of benzodiazepines: a population-based study.

Long-term use of benzodiazepines (BZDs) has been linked with an array of negative health consequences and increased medical costs and social burden. In this study, we sought to investigate the factors accounting for differential risks in the process from incident BZD use to long-term use and discontinuation in the general population. On the basis of a random sample of 187,413 people enrolled in Taiwan's National Health Insurance program on January 1, 2000, data of 2000-2002 healthcare and pharmacological services utilization were retrieved. Long-term use (LTU) was defined by having received BZD prescriptions for 180 or more days within any given calendar year. Multivariate logistic regression analyses were carried out to assess the strength of associations while adjusting for the effects of individual sociodemographics, service providers, and pharmacological agents simultaneously. Results indicated that males, elderly, and those with physical or mental disorders were more likely to become long-term users of BZDs. Having received BZD prescriptions in multiple pharmacological agents, short-acting or mixed-type agents, and hypnotic indication were associated with a roughly 2- to 5-fold increased risk of BZD LTU soon after prescription initiation. With respect to discontinuation, the effects of pharmacological characteristics seem more salient as compared to those of individual and service-provider factors. Future strategies targeting individual factors and modifying service-provider prescription behaviors may be considered to reduce possible negative consequences of BZD LTU.