Horvath · Psychotherapy (Chicago, Ill.) 2011 · Meta-analysis · n=190 studies (>14,000 treatments)

Alliance in individual psychotherapy.

Cited 2038 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and meta-analysis of 190 independent data sources representing >14,000 treatments.

PubMed 21401269 · doi:10.1037/a0022186 · record verified 2026-08-26

What was done

A random-effects meta-analysis evaluated the relationship between therapeutic alliance and outcomes in individual adult psychotherapy. Included studies required genuine (non-analogue) treatments with 5 or more adult participants where therapeutic, working, or helping alliance was measured as an independent variable. The synthesis analyzed over 200 reports across 190 independent data sources representing more than 14,000 treatments, assessing moderators such as rater perspective, timing, and assessment method.

What was found

The aggregate correlation between therapeutic alliance and treatment outcome was r = .275 (95% CI: .25 to .30, p < .0001; k = 190 independent sources), adjusted for sample size and outcome measure non-independence. The underlying dataset showed substantial heterogeneity.

Why it matters

This review establishes a robust, statistically significant baseline correlation between the therapeutic alliance and positive psychotherapy outcomes across thousands of patient treatments.

Limits

The included evidence is correlational and cannot establish causality or exclude reverse causation (e.g., early symptom improvement driving better alliance). The abstract notes high heterogeneity across studies and does not report specific numerical findings for the tested moderators.

Cited by