Response:

We thank the authors of this letter for this interesting and important comment [1]. We fully agree that the RCI is a statistical concept and is not equal to clinical improvement (or deterioration). The authors of this letter are proposing a new approach to estimate the percentage of the patients who have truly scored at or more than their baseline score. We are not in a position to ascertain the appropriateness of this method but we agree that this is an interesting approach, in addition to the approach we used.

Assuming that the proposed method is correct, it has two clinical implications.

  1. (1)

    The estimated percentage represents those whose true score is at or more than the baseline score, and is still different from the percentage of patients who have clinically meaningfully deteriorated.

  2. (2)

    The authors argue in this connection that this percentage would still be meaningful because any slight increase in depression is a serious problem because patients should not become more depressed during the treatment. If we take this position, the more clinically appropriate interpretation of the calculated estimates is that the psychotherapies decrease the percentage of patients scoring at or above the baseline after treatment from 36 to 19%, thus leading to an NNT of 5.9 (in addition to an NNT of 6.2 for increasing response).