To Be or Not to Be Improved: Patients' Perception of Symptom Improvement - Linking the SCL-90-R to Patient-Rated Global Improvement in a Large Real-World Treatment Sample.

Introduction: From both a clinical and research perspective, it is important to determine what constitutes a perceivable change in commonly used outcome measures. Objective: We aimed to do so for the Symptom Checklist-90-Revised (SCL-90-R). Methods: Patients from a large real-world sample treated with inpatient psychotherapy (n = 4,791) rated improvements in symptoms on a global 5-point Likert scale at discharge. These ratings were related to pre-post changes in the Global Severity Index (GSI) of the SCL-90-R by use of equipercentile linking. Results: A patient rating of 5 ("clearly improved") was found to be equivalent to an absolute pre-post difference in the GSI of 0.67 or to a percentage improvement of 54%, with the latter corresponding to the common definition of response as a 50% reduction in symptoms. A rating of 1 ("clearly worse") was equivalent to an increase in the GSI >0.50 and to a percentage worsening >55%. "Slightly improved" or "slightly worse" (ratings of 4 or 2) corresponded to pre-post changes in the GSI of 0.07 and 0.50. For severely disordered patients, larger changes were required for ratings of improvement, and for less severely disordered patients, larger changes were required for ratings of worsening. Results for depressive, anxiety, and personality disorders were widely consistent with those of the total sample. Conclusions: This study is the first to link patient ratings of improvement or worsening to changes in the SCL-90-R. Results are relevant to both the interpretation of changes in individual patients and of effect sizes in outcome research. Results require replication.