This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

Letter to the Editor

Drs Bschor and Baethge Reply

Tom Bschor, MD and Christopher Baethge, MD

Published: December 27, 2017

This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

See letter by Chen and Mei and article by Bschor et al

Drs Bschor and Baethge Reply

To the Editor: We greatly appreciate the comment by Chen and Mei regarding our systematic review and meta-analysis on switching antidepressants. Their views provide a welcome opportunity to clarify important points.

Chen and Mei criticize some of the studies included in the meta-analysis on methodological grounds. In 3 of the 8 studies selected,1-3 they find fault with the definition or rate of nonresponse. Of note, 2 of those studies1,2 are not part of our prespecified primary analysis, which we labeled "strict analysis," as opposed to our post hoc "broad analysis" (both meta-analyses unequivocally found no advantage of switching the antidepressant in comparison to continuation of the so far insufficient antidepressant).

In addition, Chen and Mei may have missed Petrescu and coworkers’ 2 clear definition of nonresponse as a MADRS score > 20 and a CGI-S score 3 after 6 weeks of selective serotonin reuptake inhibitors (SSRI) treatment. Also, Bose and colleagues’ 1 definition is unusual, rather than wrong, because there is no universally agreed upon definition of nonresponse. What is more, even if one follows Chen and Mei’s argument, one would not expect the switching arm to be statistically and clinically significantly inferior, which indeed is the case.1

As regards the study that we did include in our strict analysis,3 we cannot follow Chen and Mei’s view that the response criterion is artificially high. In fact, at approximately 55%, subjects did respond at a very common rate. Again, this study reported a statistically and clinically significantly superior effect of continuation relative to switching.

To support their view that different doses of antidepressants may have put switching at a disadvantage, Chen and Mei cite a systematic review and meta-analysis by Jakubovski et al4 as showing a dose-response relationship in SSRI treatment. Again, it is important to note that this argument does not apply to our strict analysis with its negative result for switching. However, by design, Jakubovski and coworkers’ study cannot support a dose-response relationship of SSRIs because in the studies included, patients were not randomized to different doses of SSRIs. On the contrary, Jakubovski et al compared different studies that employed different doses—in effect treating the trials as observational studies. Even so, the effect they found was clinically insignificant: a standardized mean difference of 0.08 is less than half of what is usually considered low. Further, we have just finished a systematic review and meta-analysis on randomized controlled trials comparing different doses of antidepressants after failure of an antidepressant monotherapy and found no advantage of higher doses for SSRIs.5

In the same vein, Chen and Mei claim that Corya et al6 used noncomparable doses of venlafaxine and fluoxetine in their study. We cannot agree: For a fair comparison, optimal doses of both antidepressants should be used. This is the case with a mean of 275.4 mg/d in the venlafaxine arm and of 37.5 mg/d in the fluoxetine arm. As detailed above, higher fluoxetine dose are not more effective than standard doses, and 37.5 mg/d can already be considered high-dose.

We do, however, agree with Chen and Mei that placebo effects contribute to antidepressive effects seen in double-blind studies. But it seems illogical to us that such a placebo effect should be at work in continuation arms only.

Chen and Mei emphasize the gap between the results of our meta-analysis and the positive effects seen in patients who switch antidepressants. This is a valid point, but not difficult to explain: Depressed patients show a high placebo response,7 and, naturally, most get better over time.8,9 This is what makes the randomized trial design scientifically so invaluable. Improvement is to be expected when patients switch antidepressants, but, as the randomized controlled trials summarized in our meta-analysis show, not to a larger extent than among patients continuing their initial treatment.


1. Bose A, Tsai J, Li D. Early non-response in patients with severe depression: escitalopram up-titration versus switch to duloxetine. Clin Drug Investig. 2012;32(6):373-385. PubMed doi:10.2165/11631890-000000000-00000

2. Petrescu B, Vasile D, Vasiliu O, et al. SSRI dose escalation versus duloxetine in treatment of major depressive disorder not responding to initial SSRI. Eur Neuropsychopharmacol (suppl). 2014;24:S455-S456. doi:10.1016/S0924-977X(14)70729-1

3. Souery D, Serretti A, Calati R, et al. Citalopram versus desipramine in treatment resistant depression: effect of continuation or switching strategies: a randomized open study. World J Biol Psychiatry. 2011;12(5):364-375. PubMed doi:10.3109/15622975.2011.590225

4. Jakubovski E, Varigonda AL, Freemantle N, et al. Systematic review and meta-analysis: dose-response relationship of selective serotonin reuptake inhibitors in major depressive disorder. Am J Psychiatry. 2016;173(2):174-183. PubMed doi:10.1176/appi.ajp.2015.15030331

5. Rink L, Braun C, Bschor T, et al. Dose increase versus unchanged continuation of antidepressants after initial antidepressant treatment failure in patients with major depressive disorder: a systematic review and meta-analysis. J Clin Psychiatry. In press.

6. Corya SA, Williamson D, Sanger TM, et al. A randomized, double-blind comparison of olanzapine/fluoxetine combination, olanzapine, fluoxetine, and venlafaxine in treatment-resistant depression. Depress Anxiety. 2006;23(6):364-372. PubMed doi:10.1002/da.20130

7. Bschor T, Kilarski LL. Are antidepressants effective? a debate on their efficacy for the treatment of major depression in adults. Expert Rev Neurother. 2016;16(4):367-374. PubMed doi:10.1586/14737175.2016.1155985

8. Henssler J, Kurschus M, Franklin J, et al. Long-term acute-phase treatment with antidepressants, 8 weeks and beyond: a systematic review and meta-analysis of randomized, placebo-controlled trials [published online ahead of print January 3, 2017]. J Clin Psychiatry. doi:10.4088/JCP.15r10545 PubMed

9. Kirsch I, Sapirstein G. Listening to prozac but hearing placebo: a meta-analysis of antidepressant medication. Prevention & Treatment. 1998;1. doi:10.1037/1522-3736.1.1.12a

Tom Bschor, MDa

Christopher Baethge, MDb

aDepartment of Psychiatry, Schlosspark-Hospital, and Department of Psychiatry and Psychotherapy, University Hospital, Technical University of Dresden, Germany

bDepartment of Psychiatry and Psychotherapy, University of Cologne Medical School, Cologne, Germany

Potential conflicts of interest: None.

Funding/support: None.

J Clin Psychiatry 2017;78(9):e1317

© Copyright 2017 Physicians Postgraduate Press, Inc.

Volume: 78

Quick Links: