This study used a multidimensional categorical model to concurrently estimate individual’s self-efficacy for managing their chronic conditions across five related domains measured with the Patient-Reported Outcomes Measurement Information System Self-Efficacy Measure for managing chronic conditions (PROMIS-SE). A total of 1087 individuals with chronic conditions was analyzed in this study. A Diagnostic Classification Model (DCM) was applied to PROMIS-SE’s 4-item short forms measuring five behavioral domains (daily activities, emotions, medications and treatments, social interactions, and symptoms) to provide patient multidimensional categorical outcomes (high, transition, or low self-efficacy). Psychometric properties were examined using classification consistency, model fit, entropy value, domain and item-level information, and patient profiles. DCM PROMIS-SE showed adequate classification consistency, fit, and high entropy values. Five domains demonstrated different average probabilities of having high self-efficacy for patients with chronic conditions from 42.0% (emotions) to 70% (medications and treatments). Rating scale analysis indicated the rating 5 (very confident) most critically discriminated patients with high or low self-efficacy for managing chronic conditions across all domains. Only four common patient profile groups contained more than 5% of the sample. Acceptable psychometric properties indicate that DCM PROMIS-SE satisfactorily classified patients with chronic conditions. This study demonstrates a feasible approach for other existing multidimensional measures to classify patients’ conditions and support clinical judgment.
- And multidimensionality
- Patient-reported outcome measure
ASJC Scopus subject areas
- Life-span and Life-course Studies