• benzodiazepines;
  • inappropriate prescribing;
  • incidence;
  • hospitalization;
  • ambulatory care


Initiatives to reduce benzodiazepine use have been largely unsuccessful despite strong associations with adverse outcomes. Curtailing incident use of benzodizepines is an alternate strategy that has yet to be explored. This study aims to determine the source of incident benzodiazepine prescriptions by comparing the risk of receiving a new prescription upon hospital discharge versus after an ambulatory care clinic visit.


Data were derived from 1189 community-dwelling adults aged 65 years naive to benzodiazepine consumption, enrolled in the Étude sur la Santé des Ainés, a prospective 3-year cohort study conducted in Québec, Canada. Health survey questionnaires were linked with provincial administrative databases of prescription and health service claims. Analysis with multivariate Poisson regression models compared the risk of incident benzodiazepine use post-hospitalization versus after an ambulatory care visit. Models were adjusted for sex, age, antidepressant use, and concomitant drugs. Sub-analyses were conducted for chronic prescriptions.


Incident benzodiazepine use was 11% over a 2-year period, with 18.3% of prescriptions leading to chronic use (> 90 days). Hospitalization conferred a 2.7-fold greater risk of incident use than an outpatient visit (OR 2.66, 95% CI 1.78–3.98) and a 4.7-fold (OR 4.74, 95% CI 1.63–13.78) increased risk of chronic use, after adjusting for potential confounders. Despite the increased risk, only 13% of new prescriptions originated post-hospital discharge, with the remainder prescribed during outpatient visits.


Interventions are required to curb incident benzodiazepine prescriptions at their source both in hospitals and in ambulatory care settings. Copyright © 2012 John Wiley & Sons, Ltd.