Background: Many nonbiological variables are reported to predict treatment response for major depression; however, there is little agreement about which variables are most predictive.
Method: Inpatient subjects (N = 59) diagnosed with current DSM-IV major depressive disorder completed weekly depressive symptom ratings with the Hamilton Rating Scale for Depression (HAM-D-17) and Beck Depression Inventory (BDI), and weekly health-related quality-of-life (HRQL) ratings with the Quality of Well-Being Scale (QWB). Acute responders were identified by a 50% decrease in HAM-D-17 score from baseline within 4 weeks of medication treatment. Predictor variables were initially chosen from a literature review and then tested for their association with acute treatment response.
Results: An initial predictive model including age at first depression, admission BDI score, and melancholia predicted acute treatment response with 69% accuracy and was designated as the benchmark model. Adding the admission QWB index score to the benchmark model did not improve the prediction rate; however, adding the admission QWB subscales for physical and social activity to the benchmark model significantly improved acute treatment response prediction to 86% accuracy (p = .001).
Conclusion: In addition to being designed for use in cost-effectiveness analyses, the QWB subscales appear to be useful HRQL variables for predicting acute inpatient depression treatment response.