Evidence‐based medicine is an important way of allowing the reader to judge clearly whether a treatment has a place in a particular condition, and to see what faults were present in the various trials of its efficacy. It is often rather unsettling to read in a meta‐analysis or in a systematic literature review how poorly constructed many trials are. The authors from Minneapolis have carried out such a study into the use of trazodone in male erectile dysfunction. They draw attention to the poor quality of many of the trials and give their reasons for this observation. They suggest that trazodone may be helpful in men with this condition, possibly at higher doses and in men with psychogenic erectile dysfunction.
OBJECTIVE
To determine the efficacy and safety of trazodone in the treatment of erectile dysfunction (ED) in a meta‐analysis.
METHODS
The data sources used were Medline and the Cochrane Library databases (January 1966 to May 2002), bibliographies of retrieved articles and review articles, and conference proceedings and abstracts. Trials were eligible for inclusion in the review if they included men with ED, compared trazodone with a control, were randomized, of ≥ 7 days’ duration and assessed clinically relevant outcomes. Two reviewers independently evaluated study quality and extracted data in a standardized fashion.
RESULTS
Six trials (comprising 396 men) met the inclusion criteria; they consisted of heterogeneous populations, were small, brief and in some cases methodologically weak. Three of the six trials showed an apparently clinically meaningful benefit of trazodone for ED compared with placebo, the differences being statistically significant in two. In pooled results, trazodone monotherapy appeared more likely than placebo to lead to a ‘positive treatment response’, although this difference was not statistically significant (37% vs 20%; relative benefit increase, 1.6; 95% confidence interval, CI, 0.8–3.3). Subgroup analyses suggested that men with psychogenic ED might be more likely to benefit from trazodone than those with mixed or physiological ED. The efficacy of trazodone also appeared greater at higher doses (150–200 vs 50 mg/day). Men randomized to trazodone were not significantly more likely than those receiving placebo to withdraw for any reason or for an adverse event, or to have specific adverse events, but wide CIs could not exclude a greater risk of these adverse outcomes with trazodone. Specific adverse events with trazodone included dry mouth (19%), sedation (16%), dizziness (16%) and fatigue (15%).
CONCLUSION
Trazodone may be helpful in men with ED, possibly more so at higher doses, and in men with psychogenic ED. Future high‐quality trials should compare trazodone with placebo and other therapies in men with depression and psychogenic ED.