Predictors of relapse after discontinuation of long-term benzodiazepine use by minimal intervention: a 2-year follow-up study.

Background: Long-term results of minimal intervention strategies to cut down benzodiazepine use are not available.
Objective: To evaluate the relapse rate over a two-year period and to search for predictors of relapse among patients who quit benzodiazepine use after receiving a discontinuation letter.
Methods: Baseline assessment and prospective monitoring of the medical records of 109 patients who quit long-term benzodiazepine use after a minimal intervention strategy in general practice.
Results: After 819 +/- 100 days of follow-up, 53 (49%) patients had remained completely abstinent. Two independent predictors of relapse were identified by Cox regression analysis: use of more than 10 mg diazepam equivalent (RR = 2.4 [1.2 - 4.7]) and poor general health perception (RR = 0.98 [0.97 - 0.99]).
Conclusion: Short-term success rates after a minimal intervention were maintained well during long-term follow-up. High-dose users have the highest risk of relapse.