Read on Twitter
A widely shared review of eCBT (Cognitive Behavior Therapy provided via computer/smartphone) published in June found that eCBT is more effective than face-to-face CBT for depression. Problem is, the study results are (probably) diametrically opposed to what they should be. 1/6
Their results, “eCBT was more effective than face-to-face CBT at reducing depression symptom severity (Standardized mean difference: −1.73; 95% CI: −2.72, −0.74; GRADE: moderate quality of evidence)” should, if anything, state the opposite. 2/6
Source: https://doi.org/10.1016/j.eclinm.2020.100442
Source: https://doi.org/10.1016/j.eclinm.2020.100442
It was my colleague Ulf Jonsson @UppsalaUni who found this interesting thing a couple of weeks ago: The forest plot in this meta-analysis seemed to show results that were reversed as when compared to the results in the original studies. 3/6
After asking me and others to corroborate his finding, Dr. Jonsson wrote to @Claudia_Editor at the Journal, "EClinicalMedicine – Published by THE LANCET". 4/6
…https://els-jbs-prod-cdn.jbs.elsevierhealth.com/pb/assets/raw/Lancet/ecm/ecm-logo.png
…https://els-jbs-prod-cdn.jbs.elsevierhealth.com/pb/assets/raw/Lancet/ecm/ecm-logo.png
Unfortunately, although the error can be quickly confirmed (check the original papers), the Journal has not yet flagged the paper to inform readers about this issue. A review can take time and it does suggest that the Journal is very careful in their evaluation, but... 5/6
it is very unfortunate that this kind of mistake survived past all of the twenty authors, and the reviewers, and is still unrecognized after media coverage and, particularly, during these times when many therapists are going online. Hence this twitter thread. 6/6
Mentioning some people off the top of my head who might find this interesting/relevant - sorry if I am mistaken. @PerCarlbring @profGerhardA @christianruck @GustavNilsonne @JoeHilgard @ivanoransky
