We thank Drs Bschor and Baethge for their letterReference Bschor and Baethge¹ regarding our study on the effect of switching between escitalopram and nortriptyline.Reference Köhler-Forsberg, Larsen, Buttenschon, Rietschel, Hauser and Souery² Drs Bschor and Baethge have performed important work in the area of switching between antidepressants,Reference Bschor, Kern, Henssler and Baethge^3, Reference Bschor and Baethge⁴ and we would like to comment on their letter to continue this important discussion.

We agree that depression improves over time and has a large placebo effect, but it has been clearly established that antidepressants have better antidepressant effects compared with placebo.Reference Cipriani, Furukawa, Salanti, Chaimani, Atkinson and Ogawa⁵ Patients in the GENDEP trial, forming the population for our switching study,Reference Köhler-Forsberg, Larsen, Buttenschon, Rietschel, Hauser and Souery² were moderately to severely depressed and assessed by a psychiatrist to be in need of antidepressant treatment. Patients were followed at weekly intervals for up to 12 weeks, and the decision of switching between the two antidepressants (i.e. from escitalopram to nortriptyline or vice versa) was based on an overall evaluation of the treatment effects and side-effects and was taken together with the patient.

As emphasised in our discussionReference Köhler-Forsberg, Larsen, Buttenschon, Rietschel, Hauser and Souery² and in the letter by Drs Bschor and Baethge, the major limitation of our study was the inability to compare the effect of switching antidepressants with a randomised continuation or placebo arm. Nevertheless, we argue that the strength of GENDEP and hence of our paper was the real-world setting and thus the ability to investigate treatments assigned based on real-world and guideline-based clinical decisions. In GENDEP, patients did not switch because of clearly defined cut-offs or randomisation but based on a decision made by the clinician together with the patient as a result of non-response and/or side-effects. Mostly, it was the combination of missing response and side-effects that led to the switching decision, which also often is the challenge in everyday clinical work.

Hence, although not including a comparison group, our study yields interesting and important clinical knowledge suggesting that switching antidepressant may represent one approach to achieve or continue an antidepressant response. Therefore, we do not agree with Drs Bschor and Baethge that our conclusions are unjustified. The clinical relevance of our findings refers to the challenging but frequent clinical situation of a patient in need of antidepressant treatment, but with the patient experiencing side-effects and/or non-response to the antidepressant. For this situation, our findings suggest that switching to an antidepressant from a different class (in our study between an selective serotonin reuptake inhibitor and a tricyclic antidepressant) may result in (continued) antidepressant response. We agree with Drs Bschor and Baethge that more high-quality randomised clinical trials are needed comparing the effect of switching to continuation or placebo.Reference Bschor, Kern, Henssler and Baethge^3–Reference Bschor and Baethge⁴ Nevertheless, since antidepressants improve response and remissionReference Cipriani, Furukawa, Salanti, Chaimani, Atkinson and Ogawa⁵ but rather often are not well tolerated, our findings represent clinical evidence that switching between antidepressants can help to achieve or continue an antidepressant response when treatment with a first antidepressant cannot be continued.