Aim: To assess the ability of ENterprising SElective Multi-instrument BLend for hEterogeneity analysis (ENSEMBLE) Minimum Dataset instrument dimensions to discriminate among subgroups of patients expected to have differential outcomes.
Materials & methods: Patients with Type 2 diabetes, knee osteoarthritis, ischemic heart disease or heart failure completed a survey designed to represent three dimensions (health, personality and behavior). Health-related outcomes and utilization were investigated using claims data. Discriminant validity and associations between the dimensions and outcomes were assessed.
Results: A total of 2625 patients completed the survey. The dimensions discriminated 50-100% of the outcome levels across disease cohorts; behavior dimension scores did not differ significantly among the healthcare utilization level subgroups in any disease cohort.
Conclusion: ENSEMBLE Minimum Dataset dimensions discriminated health-related outcome levels among patients with varied diseases.
Keywords: healthcare utilization; heterogeneity of treatment effect; patient-reported health status; questionnaires.