[Anchor-based ascertaining of meaningful changes in depressive symptoms using the example of the German short form of...

Depression is frequently measured on the 15 item version of the Center for Epidemiological Studies Depression Scale (CES-D-15). Up to now, there are no data based estimates for minimal clinically important differences (MCID) in depression scales.

Data on 5241 subjects with depressive symptoms from a German psychosomatic hospital were used. The changes in the CES-D-15 from admission to discharge were compared to the clinically global impression of change (CGIC), i. e., the therapists' perception. Categories of "slightly improved" and "much improved" were used as indicators of a clinically important difference. The relation between these ratings and the CES-D-15 was explored using mean change scores and sensitivity/specificity analyses.

On average, a reduction of approximately nine points or 30 percent in the CES-D-15 or an individual effect size of 0.9 represented a minimal clinically important difference. A consistent relationship between the changes in the CES-D-15 and the CGIC was demonstrated irrespective of age, gender, education level, type of treatment, or first diagnosis. However, higher baseline scores in CES-D-15 required larger changes of raw values to represent a clinically important difference.

Our results show anchor-based values for change in CES-D-15 that best represent a minimal clinically important difference. Reporting the percentage of patients who have reached the MCID improves the evaluation of therapeutical processes. The estimated MCID could also be used as an indicator for relevant changes in clinical practice. A further examination in other patient populations will be needed.