Effectiveness at long-term of a multidisciplinary intervention in the reduction of overuse of benzodiazepine prescriptions: A cluster controlled trial.

Abstract

Background: The overuse of Benzodiazepines (BZD) has become an area of concern because of its consequences on patients' health and healthcare costs.

Objective: to determine whether a multicomponent intervention based on a multidisciplinary education program is effective in withdrawal from BZD long-term use, compared to usual care.

Methods: Cluster-randomized, parallel-group trial with health centres as units of randomization and patients as units of analysis.

Inclusion criteria: adults with more than 4 weeks of BZD use, without serious mental disorder, non-terminal, without alcohol dependence or dementia.

Intervention: (i) discussion with the patient describing the advantages, disadvantages and alternatives of BZD use accompanied by tapering protocol with educational material, (ii) the offer of a brief consultation, and (iii) a letter addressed to the patient supported by 6 scientific societies. The primary outcome was dispensing of BZD. Socioeconomic variables and others related with the consumption of BZD (Indication, previously attempted to cease and duration of use).

Analysis: The effect of the intervention on the rate of dispensing of BZD using the multivariable generalized-estimating-equations (GEE) extension of logistic regression.

Results: Complete cessation of dispensing BZD at six months after the intervention was achieved in 108 of 333 participants (32,4 %) compared to 27 of 139 controls (19,4 %). This decrease was maintained over 5 years (41,7 % vs. 33,8 %). Adjusted absolute risk difference was -14.3 (-19.6 to -9.5). An adjusted OR 2.11 (95 % CI, 1.24-3.59). Those with high income were more likely to quit using BZD with the intervention as those with a lower income (OR 1.81, IC 95 % 1.13-2.9, p = 0.012). Similarly, those who had previously tried to withdraw from BZD were more likely to stop using with the intervention (OR 1.8, IC 95 % 1.17-2.8, p = 0.007). Those with high income were more likely to quit using BZD with the intervention as those with a lower income (OR 1.81, IC 95 % 1.13-2.9, p = 0.012). Similarly, those who had previously tried to withdraw from BZD were more likely to stop using with the intervention (OR 1.8, IC 95 % 1.17-2.8, p = 0.007).

Conclusions: The intervention was effective at 6 months and its effects persist up to 5 years. The factors favouring interruption BZD seem to be having a high income and a previously tried to withdraw from BZD.