Clinicians have several therapeutic relationships and patients only one: The effect on their assessments of relationships.

Objectives Little attention has been given to the common assessment problem that clinicians assess outcomes of several patients and may rate them in comparison to one another, whereas patients assess only their own outcomes without any comparison. We explored empirically whether this would lead to a greater variability of clinician ratings as compared to patient ratings. Methods Data from two independent samples in which clinicians and patients, using consistent instruments, rated their therapeutic relationships. We present descriptive statistics of variability and intracluster correlation coefficients. Results The Helping Alliance Scale was completed at baseline and follow‐up by 20 clinicians and 103 patients in an observational study and by 88 clinicians and 431 patients in a trial. Patients tended to rate their relationship 5–10% more highly than their clinicians, but with 50–100% more variability. Intraclinician Helping Alliance Scale ratings were more correlated than those by patients (intracluster correlation coefficients 0.3–0.7 vs. 0.0–0.2). Conclusion Contrary to our assumption, clinicians' ratings of therapeutic relationships were in both samples less variable than those of their patients. When clinicians rate outcomes of several patients, a cluster effect of ratings may have to be considered in the design and analysis.