What predicts improvement when ketamine is paired with therapy for PTSD

The question

When ketamine is combined with psychotherapy for post-traumatic stress disorder (PTSD), which patient and treatment factors are linked to greater symptom improvement?

What they did

Researchers searched for studies in adults with PTSD that combined ketamine with psychotherapy, then pooled the data on individual participants rather than just study averages. Twelve studies with 533 participants were included; nine were rated poor quality. The main outcome was change on the PCL-5, a standard self-report checklist of PTSD symptoms. Statistical models tested age, sex, baseline symptom severity, number of ketamine sessions, number of psychotherapy sessions, how the ketamine was given, and how long treatment lasted.

What they found

Greater symptom improvement was associated with more psychotherapy sessions (β = 1.03 per session), more ketamine sessions (β = 1.15 per session), higher symptom severity at the start (β = 0.50 per PCL-5 point), and shorter overall treatment duration (β = -0.58 per week). Differences between studies were not trivial (intraclass correlation coefficient about 0.15). When men and women were analyzed separately, number of psychotherapy sessions, baseline severity and shorter treatment duration stayed linked to improvement in both groups. In a sensitivity analysis limited to better-quality data (63 people), baseline severity was the only factor that remained statistically significant.

The fine print

Most of the included studies were rated poor quality, and the higher-quality subset covered only 63 people. These are associations, not proof that adding sessions causes better outcomes, and people who start with the most severe scores tend to show larger drops for statistical reasons alone. Studies differed from one another, follow-up length and protocols were not standardized, and this summary is based on the abstract only.

What it means

Across the available studies of ketamine combined with psychotherapy for PTSD, starting with more severe symptoms was the most consistent predictor of larger improvement, while claims about session numbers and timing rest on weak evidence and need testing in properly designed trials.

Evidence grade: Low. A pooled analysis of predictors drawn mostly from poor-quality studies, which cannot show cause and effect even though individual participant data were used.

This summary is based on the study’s published abstract.