A Cochrane review included 51 RCTs with a total of 3603 patients. Forty-four studies with 3372 patients with physical illness and depression contributed data to the efficacy analyses. The trials were conducted in stroke (n=11), HIV/AIDS (n=7), Parkinson's disease (n=6), cancer (n=4), chronic obstructive pulmonary disease (COPD, n=3), diabetes (n=3), myocardial infarction (n=3), renal failure (n=2) and one in each of rheumatoid arthritis, brain injury, asthma, coronary artery disease, chronic heart failure, epilepsy and chronic prostatitis. There were also 3 trials with mixed diagnoses. Depression included major depression, adjustment disorder and dysthymia. SSRIs and tricyclic antidepressants vs. placebo were used in majority of the trials. Pooled data for the primary outcome, response to treatment, provided an OR of 2.33 (95%CI 1.80 to 3.00; 25 studies, 1674 patients) favouring antidepressants. Antidepressants were also more efficacious at the other time-points. At 6-8 weeks number-needed-to-treat (NNT) for antidepressants was 6. At 4-5 weeks, 9-18 weeks and >18 weeks NNT was 7. At 6-8 weeks, fewer patients receiving placebo dropped out compared to patients treated with an antidepressant. Dry mouth and sexual dysfunction were more common in patients treated with an antidepressant.
Comment: The quality of evidence is downgraded by study quality (inadequate allocation concealment) and inconsistency (heterogeneity in patients and treatments). The quality of evidence is upgraded by large magnitude of effect.