Who are we studying? Reporting practices and sample characteristics in psychotherapy research.

Psychotherapy research informs clinical practice and mental health policy. However, limited and inconsistent reporting of sample and contextual characteristics makes it difficult to determine who is represented in psychotherapy research and to assess the generalizability of findings.

Objectives: To assess the reporting of PTTC characteristics (patients, therapists, treatment, and context) in psychotherapy research and to describe who is being studied.

Method: Descriptive review of 100 psychotherapy studies using a standardized codebook-later used to inform a PTTC questionnaire-with independent reviewers extracting PTTC data.

Results: Patients' age and gender were reported in 83.8% and 95% of studies, respectively, whereas education was reported in 46% and race in 50%. Sexual orientation (12%), SES (12%), immigration (5%), rurality (3%), and disability (2%) were rarely reported. Therapist demographics were reported infrequently, including gender (32%) and profession (35%), while years of experience were reported in 15% of studies. Treatment delivery characteristics were moderately reported (modality 56%, session frequency 50%, number of sessions 58%), whereas contextual indicators (care system 59%, level of care 24%, setting 46%) were reported inconsistently. When characterized, patients were primarily adults (97.9%), female (67.7%), urban (81.1%), and tertiary educated (79.4%); therapists were mostly female (81.0%) psychologists (46.0%) working under supervision (67.0%), with an average of 9 years of experience. The context favored individual (66.5%) outpatient treatment (78.3%), most commonly delivered in tertiary-level settings (52.1%) in Europe (46%) and North America (33%).

Conclusions: Psychotherapy research overrepresents adults, women, urban residents, and tertiary-educated participants from higher-income settings. Current reporting practices obscure for whom, by whom, and under what conditions psychotherapy is most effective, limiting generalizability and clinical applicability.