Reliable and clinically significant change based on the Health of the Nation Outcome Scales.

To evaluate treatment outcomes of individual patients based on clinician-rated instruments, the assessment of reliable and clinically significant change (RCSC) is essential. In heterogeneous samples, RCSC underestimates treatment outcome. Therefore, the Reliable Change Index (RCI) was adjusted by a stratification into subsamples. This method was tested on the Health of the Nation Outcome Scales for Children and Adolescents (HoNOSCA) ratings (n = 12,547) at admission and discharge of youths (age 6-18 years) treated in ten psychiatric institutions. Based on the test-retest reliability of a subsample (n = 397), the RCI was calculated for three severity groups ("very severe," "moderately severe," and "subclinical/mild"). Individuals who accomplished reliable change during treatment and moved to a subclinical score were classified as recovered. Using the traditional RCSC calculation, the large majority (75.1%) of the sample would be considered as unchanged, 2.9% as deteriorated, 18.9% as improved, and 3.1% as recovered. Using RCI cutoff points based on the severity ratings at admission results in a more representative distribution of outcome groups, where 54.6% of the sample was stable, 7.5% worsened, 21.6% improved, and 16.3% recovered. This methodological framework for calculating RCSC for heterogeneous populations is applicable for all HoNOS instruments, making it very useful for mental health professionals.