Hendrick et al (Reference Hendrick, Fukuchi and Altshuler2001) state that the findings of their study provide no reason to discourage nursing among women taking paroxetine, fluvoxamine or sertraline at standard therapeutic doses. Comparison with previous studies is difficult, owing to the research literature consisting mainly of single case reports or small samples, difference in methods and lack of key information (as reviewed by Reference Yoshida, Smith and KumarYoshida et al, 1999).

While I applaud the effort of studying 50 nursing mother—infant pairs, I disagree with the inclusion of all of them as study subjects for two main reasons.

First, seven were included whose prescribed doses of antidepressant were below the recommended dose (British Medical Association & Royal Pharmaceutical Society of Great Britain, 2001) for the treatment of depression (paroxetine 5 mg (n=1), paroxetine 10 mg (n=2), sertraline 25 mg (n=4)). In the case of sertraline, where 30 pairs were included, exclusion of these subjects would increase the percentage of detection of medication, including metabolites, from 24% (8/30) to 34% (8/26).

Second, Hendrick et al came to the same conclusion regarding the safety of fluvoxamine, sertraline and paroxetine, but according to their Table 1 (p. 164) only one serum sample of the five taken from mother—infant pairs where the mother was taking fluvoxamine should be taken into consideration. Of the remainder, no maternal medication concentration was obtained in three cases, and in the fourth maternal medication concentration was below the detectable range of the assays, raising questions about compliance that will add bias to the results.