Assessment of abuse potential of benzodiazepines from a prescription database using ‘doctor shopping’ as an indicator

Background: Benzodiazepines are widely used for different purposes because of their pharmacological properties, but their abuse potential may represent a limitation to their use. Data suggest that this abuse potential may vary between products and available dosages. Doctor shopping (the simultaneous use of several physicians by a patient) is one of the most important ways in which prescription drugs, in particular benzodiazepines, are diverted.

Objective: To assess the potential for abuse of several benzodiazepines using doctor shopping in a French administrative area as a proxy for abuse.

Methods: All prescriptions reimbursed during the year 2003 in Haute-Garonne, France (one million inhabitants) for benzodiazepines that were available in ambulatory care through community pharmacies as solid oral forms were extracted from a reimbursement database. The benzodiazepines were alprazolam (0.25 mg, 0.50 mg), bromazepam 6 mg, clonazepam 2mg, clorazepate (5mg, 10 mg, 50 mg), diazepam (1 mg, 5 mg, 10 mg), flunitrazepam 1 mg, lorazepam (1 mg, 2.5 mg) and tetrazepam 50 mg. For each patient, the quantities prescribed, dispensed and obtained by doctor shopping (i.e. overlap between prescriptions from different prescribers) were computed. Benzodiazepines were compared using their ‘doctor shopping indicator’ (DSI, the percentage of each drug obtained through doctor shopping among the total reimbursed quantity).

Results: About 128 000 patients received at least one benzodiazepine during the year. Four groups of benzodiazepines were identified according to their abuse potential: very high abuse potential (flunitrazepam, DSI = 42.8%); high abuse potential (diazepam 10 mg, DSI = 3.2%; clorazepate 50 mg, DSI = 2.7%); intermediate abuse potential (alprazolam 0.50 mg, bromazepam, clonazepam, DSI ranging from 1.8% to 1.9%); and low abuse potential (other benzodiazepines and dosages, DSI ranging from 0.3% to 1.1%).

Conclusion: The DSI can be used to assess the relative abuse liability of benzodiazepines and to detect signals of new patterns of abuse in settings where centralized records of prescription or deliveries are available for the great majority of patients.