I read with interest the recent article by Wilson et al (Reference Wilson, Mottram and Ashworth2003). On the basis of a randomised, double-blind placebo-controlled trial, they conclude that sertraline is not effective in preventing recurrent episodes of depression. However, I would like to make certain observations.

First, looking at Table 2 (p. 494), we find that the number of patients remaining in the study at 100-week follow-up is 15 in the sertraline group and 12 in the placebo group. These numbers are too small to draw any major conclusions. Also, looking at the same table, we find that at 4-week follow-up there were six recurrences of depression in the placebo group compared with only two in the sertraline group; that is, the sertraline group had significantly fewer recurrences of depression in the first 4 weeks of prophylactic therapy.

Second, I would like to make an observation about statistical v. clinical significance. Again looking at Table 2, we find that the number of cumulative recurrences were fewer in the sertraline group than in the placebo group at all points of the maintenance phase over 2 years. Even though these numbers did not reach statistical significance, they are clinically significant. This opinion is based on two reasons: first, for a physician, prevention of even one case of recurrence is important and satisfying; second, from a community and financial perspective, sertraline prophylaxis has been found to be more cost-effective than treating each new episode of depression with dothiepin (Reference Hatziandreu, Brown and RevickiHatziandreu et al, 1994). If Wilson et al had included an analysis of treatment costs (including the cost of treating episodes of recurrent depression) in both the groups, it might have made interesting reading.

Third, as Wilson et al pointed out, failure to increase the dose of sertraline at the earliest signs of recurrence contributed to the greater number of recurrences observed in their study. There should have been provision to increase the dose of sertraline as and when the clinical situation demanded it. After all, a significant number of patients do require a daily dose in excess of 50 mg sertraline (Reference Suri, Altshuler and RasgonSuri et al, 2000), a dose that was used to treat almost three-quarters of the patients in this study.

In conclusion, the data presented by Wilson et al are insufficient to suggest making any changes in the current practice of prescribing sertraline for treatment and prophylaxis of depression in older people.