It is becoming increasingly more relevant for health care providers to also participate in external quality assurance measures. A few systematic approaches that also integrate outcome quality and include a comparison of different clinics have only recently been developed for the area of the treatment of patients with mental disorders. In this context, a key issue is the appropriate consideration of confounders, i.e., of unevenly distributed patient characteristics that are causally related to the treatment results. Using a sample of consecutive patients from four psychotherapeutic clinics, we exemplarily examined the significance of the confounders for the comparison of the clinics with regard to short-term outcome. The results show that an adjustment is principally advisable and is generally associated with a reduction of the differences in treatment results. The discussion asserts that the use of regression weights, which were ascertained using an independent, representative sample, allows for a more valid risk adjustment. Finally, reference will be made to the relevance of giving consideration to varying resource usage in the assessment of the risk-adjusted clinic comparisons.
