Effectiveness of safety warnings in atypical antipsychotic drugs: an interrupted time-series analysis in Spain

Abstract

Background: Studies conducted to obtain drug authorization are often of short duration and based on small sample sizes in selected populations. Policies on drug safety rely on the validity of the methods used to achieve rapid and effective communication of new information. No formal evaluation has ever been made of the Spanish communications system, although indirect data have raised questions about its effectiveness. Objective: To evaluate the impact of two safety warnings issued by the Spanish Drug Agency, and of a later prior authorization requirement involving the use of atypical antipsychotic drugs in the elderly. Methods: The study was based on a time-series analysis constructed with data corresponding to monthly invoicing from 2000 to 2006 for olanzapine and risperidone in the Region of Valencia, Spain. Because the safety warnings and the prior authorization policy applied exclusively to prescriptions of these drugs for elderly patients with dementia, we investigated whether these interventions were successful and therefore changed prescription patterns for pensioners receiving low-strength formulations (the available proxy for elderly subjects with dementia), without altering patterns for those receiving the highest-strength formulations (typically used in schizophrenic patients) or for prescriptions for non-pensioners (any strength formulations). These two latter groups were therefore established as the control groups. Results: Defined daily doses (DDDs) for olanzapine in low-strength pharmaceutical forms showed a clear levelling off after the first warning, while that for risperidone showed less pronounced decline. The prior authorization policy had a dramatic effect on the consumption of risperidone, but not on that of olanzapine. DDDs for low-strength formulations between the 12 months prior to the first warning and the 12 months following the prior authorization showed a substantial reduction (22% for risperidone and 33% for olanzapine). In the high-strength forms and in non-pensioners the upward trends in DDDs remained unaltered after both interventions. Conclusion: The safety warnings concerning atypical antipsychotic drugs were effective in reducing the prescribing of risperidone and olanzapine in low-strength doses in pensioner prescriptions, and the implementation of a prior authorization policy had a dramatic effect on the prescribing of risperidone.