On 2017 May 01, Doug Berger commented:

LACK OF BLINDING IN THIS STUDY WAS A SERIOUS METHODOLOGIC FLAW. ADDITIONALLY, FACULTY SUPERIOR OF LEAD AUTHOR STEVEN HOLLON, DR.STEPHAN HECKERS, EDITOR OF JAMA PSYCHIATRY, BOTH AT VANDERBILT DEPT. OF PSYCHIATRY REQUIRES REPORTING AS CONFLICT OF INTEREST

This study by Hollon et al. compared an antidepressant medication-only arm with a combined cognitive therapy/antidepressant arm and concluded that the cognitive therapy/antidepressant combination enhanced the recovery rates compared with antidepressant alone, and that the magnitude of this increment nearly doubled for patients with more severe depression.

We opine that for subjects with greater severity, there could have been both antidepressant efficacy, as well as more hope and expectation as bias in the group who knew openly that they had received combined cognitive therapy/medication as a possible treatment. This can lead to an erroneous conclusion of greater efficacy for the combined group. The large subject number in this study could also easily lead to an erroneous finding on statistical testing as a small amount of bias in the subjects adds-up.

In addition, it goes against clinical trial logic to compare the unbilnded-cognitive therapy/medication group to the unblinded medication-only group.This is because, as all the study arms were unblinded, the combined cognitive therapy/medication group has an advantage over the medication-only group. The combined group does not filter any hope or expectation bias that may be lurking in the cognitive therapy arm, while the medication-only group engenders no different hope or expectation than the medication arm in the combined group. It is thus logically invalid to compare the cognitive therapy-medication arm that can have an unfiltered cognitive therapy-positive bias from the unblinded nature of receiving cognitive therapy tasks by cognitive therapy trained therapists vs. the medication-only arm which has the same bias possibility as the medication in the combined group, but lacking any possible positive bias from the combined cognitive therapy arm. Medications are required to show efficacy when compared in a double-blind study that includes a blind-placebo control as these controls are necessary to filter bias of any hope or expectation of efficacy. Neither blind controls nor blinded placebo were used in the design of the Hollon et al. study here.

Dr. Hollon should have also noted the conflict of interest in that the Director of his dept. Dr. Stephan Heckers was also on the Editorial Board of JAMA Psychiatry when this paper was submitted and was the Editor-in-Chief of JAMA Psychiatry when it was published.

The paper was retracted once for multiple errors, and it should be withdrawn completely because of poor clinical trial logic in making claims from unblinded subjects and treaters, and in addition due to conflict of interest in having the Editorial Director (Stephan Heckers) of the publication (JAMA Psychiatry) also as head of the Faculty of the lead author's (Hollon) affiliation at Vanderbilt University.

The full comment on the Hollon et al. study above can be read here;

Double blinding requirement for validity claims in cognitive-behavioral therapy intervention trials for major depressive disorder. Analysis of Hollon S, et al., Effect of cognitive therapy with antidepressant medications vs antidepressants alone on the rate of recovery in major depressive disorder: a randomized clinical trial. By D. Berger. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4863672/

On 2016 Apr 22, Ivan Oransky commented:

This paper has been retracted and replaced: http://retractionwatch.com/2016/04/21/authors-retract-replace-highly-cited-jama-psych-paper-for-pervasive-errors/

On 2016 Apr 22, Ivan Oransky commented:

This paper has been retracted and replaced: http://retractionwatch.com/2016/04/21/authors-retract-replace-highly-cited-jama-psych-paper-for-pervasive-errors/

On 2017 May 01, Doug Berger commented:

LACK OF BLINDING IN THIS STUDY WAS A SERIOUS METHODOLOGIC FLAW. ADDITIONALLY, FACULTY SUPERIOR OF LEAD AUTHOR STEVEN HOLLON, DR.STEPHAN HECKERS, EDITOR OF JAMA PSYCHIATRY, BOTH AT VANDERBILT DEPT. OF PSYCHIATRY REQUIRES REPORTING AS CONFLICT OF INTEREST

This study by Hollon et al. compared an antidepressant medication-only arm with a combined cognitive therapy/antidepressant arm and concluded that the cognitive therapy/antidepressant combination enhanced the recovery rates compared with antidepressant alone, and that the magnitude of this increment nearly doubled for patients with more severe depression.

We opine that for subjects with greater severity, there could have been both antidepressant efficacy, as well as more hope and expectation as bias in the group who knew openly that they had received combined cognitive therapy/medication as a possible treatment. This can lead to an erroneous conclusion of greater efficacy for the combined group. The large subject number in this study could also easily lead to an erroneous finding on statistical testing as a small amount of bias in the subjects adds-up.

In addition, it goes against clinical trial logic to compare the unbilnded-cognitive therapy/medication group to the unblinded medication-only group.This is because, as all the study arms were unblinded, the combined cognitive therapy/medication group has an advantage over the medication-only group. The combined group does not filter any hope or expectation bias that may be lurking in the cognitive therapy arm, while the medication-only group engenders no different hope or expectation than the medication arm in the combined group. It is thus logically invalid to compare the cognitive therapy-medication arm that can have an unfiltered cognitive therapy-positive bias from the unblinded nature of receiving cognitive therapy tasks by cognitive therapy trained therapists vs. the medication-only arm which has the same bias possibility as the medication in the combined group, but lacking any possible positive bias from the combined cognitive therapy arm. Medications are required to show efficacy when compared in a double-blind study that includes a blind-placebo control as these controls are necessary to filter bias of any hope or expectation of efficacy. Neither blind controls nor blinded placebo were used in the design of the Hollon et al. study here.

Dr. Hollon should have also noted the conflict of interest in that the Director of his dept. Dr. Stephan Heckers was also on the Editorial Board of JAMA Psychiatry when this paper was submitted and was the Editor-in-Chief of JAMA Psychiatry when it was published.

The paper was retracted once for multiple errors, and it should be withdrawn completely because of poor clinical trial logic in making claims from unblinded subjects and treaters, and in addition due to conflict of interest in having the Editorial Director (Stephan Heckers) of the publication (JAMA Psychiatry) also as head of the Faculty of the lead author's (Hollon) affiliation at Vanderbilt University.

The full comment on the Hollon et al. study above can be read here;

Double blinding requirement for validity claims in cognitive-behavioral therapy intervention trials for major depressive disorder. Analysis of Hollon S, et al., Effect of cognitive therapy with antidepressant medications vs antidepressants alone on the rate of recovery in major depressive disorder: a randomized clinical trial. By D. Berger. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4863672/