In Reply The issues identified by Dr Lee and colleagues are commonly raised in trials of psychotherapeutic interventions. In pragmatic trials1 carried out in real-life settings, it is difficult to show which elements of the intervention are most efficacious.2 In addition, a greater dose of therapy is not necessarily better. A metaregression analysis of 70 studies (92 comparisons) with 5403 patients in which individual psychotherapy was compared with a control group (eg, waiting list, usual care) found that contact time and duration of therapy do not predict outcomes when other characteristics, such as quality of the study, are taken into account.3 Larger studies that measure many variables in tightly controlled experimental conditions will be needed to determine the mediators, moderators, and other elements of the therapeutic encounter that affect outcomes in psychological interventions.
Atif Rahman, Mark van Ommeren, Saeed Farooq. Interventions for Anxiety and Depression in Conflict-Affected Areas—Reply. JAMA. 2017;317(13):1376–1377. doi:10.1001/jama.2017.1848