The primary efficacy measures from our paper about treatment effect at week 24 (Blomhoff et al, 2001) are reported in the method section of the paper about the follow-up study (Haug et al, 2003). In the pairwise comparisons, combined sertraline and exposure and sertraline alone were significantly superior to placebo, while a non-significant trend towards increased efficacy of exposure alone compared with placebo was reported.

The four study groups had a significant reduction in scores on all social phobia scales from baseline to follow-up. Furthermore, there was no significant difference in scores on primary efficacy measures between the active treatment groups in any of the time-point analyses between week 0 and week 24. In the follow-up analyses we were therefore mainly interested in the changes after cessation of treatment. For the exposure group and the placebo group there was a further improvement in scores on social phobia from week 24 to week 52 and the changes on several of the sub-scales were highly significant. On SF–36, which demonstrates changes in a more global functioning, there was a significant improvement for the exposure alone and the placebo groups, while there was a significant deterioration in both the sertraline-treated groups. Changes in scores on other social phobia scales for the sertraline-treated groups were non-significant, but there was a tendency towards deterioration (Tables 1 and 2, pp. 314–315). We agree that the changes in sertraline-treated groups during the follow-up period were marginal. However, contrasting these minimal changes with the significant improvement in the exposure-treated group, we find it appropriate to conclude that exposure therapy given alone seems to be more beneficial in the long term. Longer follow-up could have added valuable information to this issue. In all groups about 20% of the patients were treated with sertraline during the follow-up period so this could not explain the differences in scores between the groups at week 52.