Overcoming the difficulties of meta-analysis in psychotherapy

Tags: Poster
Lopez PL1, Ciapponi A2
1Institute of Cognitive and Translational Neuroscience (INCyT), INECO Foundation, Favaloro University and CONICET, 2Institute for Clinical Effectiveness and Health Policy (IECS-CONICET)


Although the number of systematic reviews (SRs) and meta-analysis (MAs) that evaluate the efficacy of psychotherapy has increased in recent years, the evidence is still scarce compared to studies of medical treatments. This limitation could be explained by the low number of randomized controlled trials and lack of knowledge about methodology. Additionally, psychotherapy presents particularities that hinder MAs with standard proceedings like high heterogeneity of assessment tools, or different modalities of treatments with the same denomination but different components and session frequencies.


To evaluate the trends in SRs/MAs in psychotherapy in the Cochrane Library compared to PubMed and to assess the frequency of considerable heterogeneity in primary outcomes of a sample of recent Cochrane Reviews. We will discuss alternative proceeding adapted to the heterogeneity of the SRs/MAs.


We reviewed publications of SRs/MAs about psychotherapy in PubMed and the Cochrane Library from the last 10 years and heterogeneity issues in Cochrane MAs in the last year.


From 27 February 2018 there were 557 results from 10219 records (5%) when we searched on 'psychotherapy' in Cochrane Reviews and 4927 SRs or MAs (filtered by publication type) in PubMed. In the last 10 years SRs/MAs in PubMed and the Cochrane Library showed divergent trends: +37 yearly (95% confidence interval (CI) 30 to 45; P < 0.0001) and -3 yearly (95% CI -5 to -2; P = 0.002) respectively (Figure 1).

We found that 20/27 (74%) of Cochrane Reviews published in 2017 which specifically involved psychotherapy, presented considerable heterogeneity (I2 > 60%) in the primary outcomes.


Cochrane MAs in psychotherapy do not follow the growing up trend. Heterogeneity of outcome measures is a common problem. It is necessary that psychologists promote a consensus about the assessment tools and treatment modalities to facilitate and increase MAs and to reduce the heterogeneity.

Also, the standards and requirements from Cochrane editorial boards would need to be rethought to consider the complexity of psychotherapy research.

Patient or healthcare consumer involvement: