To examine this question, a team of investigators from Glostrup, Denmark, conducted a systematic review using standard Cochrane methods. Data from 2295 patients were extracted from 35 trials. Of 18 comparison interventions, no single intervention was assessed in more than four trials. Furthermore, the authors found that, partly because of the very low quality of evidence for the reported outcomes, it was not possible to draw firm conclusions regarding pharmacologic interventions for facilitating benzodiazepine discontinuation in chronic benzodiazepine users.
So where does this leave us? In short, we have a huge clinical problem. All we know is that weaning patients from benzodiazepines is best done very slowly and is often painful for our patients. In my own practice, I often manage very slow tapers that last between 6 months and a year. It seems that most patients can manage this, but I have no evidence beyond my clinical impression to back up my opinion.
Given the millions of people worldwide who take benzodiazepines long-term, we are desperately short of data and high-quality randomized trials aimed at developing evidence-based withdrawal protocols. These need to occur.