On 2014 Apr 22, James C Coyne commented:

This is a badly done meta analysis that seems to jump to a conclusion to which the authors seemed already committed: The authors systematically searched the literature, but found too few studies to justify their sweeping conclusion: that strength of evidence for psychotherapy for depressive symptoms among cancer patients warrants widespread dissemination of existing treatments and implementation routine care.

They claim to have identified only 6 effect sizes from RCTs for psychotherapy. But three are from collaborative care studies in which patients assigned to the intervention group did not necessarily get psychotherapy and many patients got medication with or without therapy. Patients assigned to the control group in two of these studies had to pay for any treatment whereas it was free for patients in the intervention group. Two effect sizes came from ta single study, violating the assumption of independent effect sizes, and a support group was counted as psychotherapy. It is usually considered a control condition. If it had served as such for the CBT treatment, CBT would have had a negative effect size.

You can read more at my 2 blog posts.

http://blogs.plos.org/mindthebrain/2014/04/15/meta-analyses-conducted-professional-organizations-trustworthy/

http://jcoynester.wordpress.com/2014/04/22/does-psychotherapy-work-for-depressive-symptoms-in-cancer-patients/

On 2014 Apr 22, James C Coyne commented:

This is a badly done meta analysis that seems to jump to a conclusion to which the authors seemed already committed: The authors systematically searched the literature, but found too few studies to justify their sweeping conclusion: that strength of evidence for psychotherapy for depressive symptoms among cancer patients warrants widespread dissemination of existing treatments and implementation routine care.

They claim to have identified only 6 effect sizes from RCTs for psychotherapy. But three are from collaborative care studies in which patients assigned to the intervention group did not necessarily get psychotherapy and many patients got medication with or without therapy. Patients assigned to the control group in two of these studies had to pay for any treatment whereas it was free for patients in the intervention group. Two effect sizes came from ta single study, violating the assumption of independent effect sizes, and a support group was counted as psychotherapy. It is usually considered a control condition. If it had served as such for the CBT treatment, CBT would have had a negative effect size.

You can read more at my 2 blog posts.

http://blogs.plos.org/mindthebrain/2014/04/15/meta-analyses-conducted-professional-organizations-trustworthy/

http://jcoynester.wordpress.com/2014/04/22/does-psychotherapy-work-for-depressive-symptoms-in-cancer-patients/