Comment: The quality of evidence is downgraded by study quality (unclear allocation concealment, high dropout rates), inconsistency (heterogeneity in patients and interventions), imprecise results (few studies for each comparison) and indirectness (a high number of studies were very old and did not meet modern standards for clinical trials, short follow-up time).
A Cochrane review [Abstract] 1 included 35 studies with a total of 2295 patients. The trials investigated patients with a varying clinical picture dominated by anxiety. In most trials, the majority were women. In most trials the mean age was around 50 years and mean duration of benzodiazepine use between 5 and 10 years. Trial duration ranged between 1 and 24 weeks, the mean was 9.4 weeks. For benzodiazepine discontinuation, valproate (RR 2.55, 95% CI 1.08 to 6.03; 1 study, n=27) and tricyclic antidepressants were potentially beneficial at end of intervention (RR 2.20, 95% CI 1.27 to 3.82; 1 study, n=47). For benzodiazepine withdrawal symptoms at end of intervention, pregabalin may be effective (MD -3.10 points, 95% CI -3.51 to -2.69; 1 study, n=106), as well as paroxetine (MD -3.57 points, 95% CI -5.34 to -1.80; 2 studies, n=99), tricyclic antidepressants (MD -19.78 points, 95% CI -20.25 to -19.31; 1 study, n=38) and flumazenil (SMD -0.95, 95% CI -1.71 to -0.19; 3 studies, n=58). However, the positive effect of paroxetine did not persist until longest follow-up (MD -0.13 points, 95% CI -4.03 to 3.77; 1 study, n=54). The following treatments reduced symptoms of anxiety at end of intervention: carbamazepine (MD -6.00 points, 95% CI -9.58 to -2.42; 1 study, n=36), pregabalin (MD -4.80 points, 95% CI -5.28 to -4.32; 1 study, n=106), paroxetine (MD -6.75 points, 95% CI -9.64 to -3.86; 2 studies, n=99), and flumazenil (MD -1.30 points, 95% CI -2.28 to -0.32; 1 study, n=18). Valproate reduced the proportion of patients relapsing to benzodiazepine use (RR 0.31, 95% CI 0.11 to 0.90; 1 study, n=27).
Date of latest search:
Primary/Secondary Keywords