The generalizability of psychotherapy efficacy trials in major depressive disorder: an analysis of the influence of patient selection in efficacy trials on symptom outcome in daily practice

We performed an extensive literature search in PubMed, PsycInfo and additional databases for PETs for MDD. From these, we identified commonly used exclusion criteria. We investigated the influence of exclusion criteria on symptom outcome by multivariate regression models in a sample of patients suffering from MDD according to the MINIplus from a routine clinical practice setting (n=598). Data on routine clinical practice patients were gathered through Routine Outcome Monitoring.We selected 20 PETs and identified the following commonly used exclusion criteria: ‘a baseline severity threshold of HAM-D≤14’, ‘current or past abuse or dependence of alcohol and/or drugs’ and ‘previous use of medication or ECT’. In our routine clinical practice sample of patients suffering from MDD (n=598), presence of ‘current or past abuse of or dependence on alcohol and/or drugs’ had no significant influence on outcome.‘Meeting a baseline severity threshold of HAM-D≤14’ and ‘previous use of medication or ECT’ were associated with better outcome, but the explained variance of the models was very small (R2=2-11%).The most consistently used exclusion criteria are not a major threat to the generalizability of results found in PETs. However, PETs do somewhat improve their results by exclusion of patients with minor depression and patients who used antidepressants prior to psychotherapy.In the development of guidelines, randomised controlled trials (RCTs) and meta-analyses thereof are considered the most reliable source of evidence. However, it is unknown to what extent the results of these RCTs are generalizable to routine clinical practice. In RCTs, much effort is put in optimising the internal validity, i.e. the possibility to determine to what extent the observed efficacy is reproducible and attributable to the investigated treatment. The internal validity of trials is improved by the use of strict criteria for patient selection. While this is very important for methodological and ethical