Dear Editor:We thank the authors of 2 letters inquiring how ratings were established with regards to short-term psychodynamic psychotherapy (STPP) for major depressive disorder in the Canadian Network for Mood and Anxiety Treatments (CANMAT) guidelines.1 The second-line treatment recommendation for STPP captures the emerging nature of the evidence and clinical utility of STPP. Briefly, evidence from meta-analyses can result in the highest level of evidence, but we did not rely on abstracts for conclusions. 2 We examined key factors of each study from any meta-analysis, including whether at least 2 studies used the same manual, choice of control conditions, replication by another research group, overall sample size, and finally the multiple subanalyses conducted. When there are multiple outcomes, divergent findings weaken conclusions. Good scientists, and good clinicians, may reasonably disagree both on which subanalyses should have primacy, as well as the relative importance of study characteristics.Leichsenring and Steinert rightly mention that there are 2 studies of the de Jonghe et al.3 model of STPP-but one is a study designed to compare STPP alone to STPP with pharmacotherapy. 4 There was no control group for the psychotherapy, and in the results, the authors state, ''The advantage of combining antidepressants with STPP were equivocal. Neither the treating clinicians nor the independent observers were able to ascertain them.'' The second de Jonghe et al. model of STPP study, conducted by Driessen et al., 5 was a noninferiority trial comparing STPP to cognitive-behavioural therapy (CBT), and its authors conclude that noninferiority was demonstrated at posttreatment but could not be demonstrated for remission rates and follow-up measures. Both trials lend support but together fail to establish level 1 evidence for efficacy of that model of STPP. They subsequently cite the Connolly Gibbons et al.6 study, published after the review period of the CAN-MAT article, invoking the Luborsky 7 model of STPP. This, too, is a noninferiority trial without a control condition. As a single trial from a single research group, it does not satisfy the criterion of level 1 evidence. Furthermore, noninferiority was demonstrated on a primary outcome but not on 4 secondary outcomes, and both treatments had low remission rates.Town et al. cite various subanalyses from the Driessen et al. meta-analysis, but the issues of multiple divergent findings in that article and the use of many small and methodologically challenged studies do not allow for a clear, unequivocal conclusion as would be needed for a first-line treatment recommendation. Additionally, they state that other therapies may use different treatment manuals. While this is true, each of the other first-line treatments had evidence from at least 2 studies using the same specific type of therapy. Behavioural activation and mindfulness-based CBT each share many features with CBT, but each therapy has its own trials and evidence. Randomised controlled trials are needed in whic...