Development and Initial Validation of the PROMIS®-Plus-HF Profile Measure

Autor: Elliot S. Fisher, Jill M. Gelow, Michael A. Kallen, Susan Yount, Kathleen L. Carluzzo, Stephen E. Kimmel, Faraz S. Ahmad, Karen E. Schifferdecker, Peter A. McCullough, David Cella
Rok vydání: 2019
Předmět:
Zdroj: Circ Heart Fail
ISSN: 1941-3297
1941-3289
Popis: Background: Bringing together generic and heart failure (HF)–specific items in a publicly available, patient-reported outcome measure may facilitate routine health status assessment for improving clinical care and shared decision-making, assessing quality of care, evaluating new interventions, and comparing groups with different conditions. Methods and Results: We performed a mixed-methods study to develop and validate the PROMIS®-Plus-HF (Patient-Reported Outcomes Measurement Information System®-Plus-Heart Failure) profile measure—a HF-specific instrument based on the generic PROMIS. We conducted 8 focus groups with 61 patients with HF and phone interviews with 10 HF clinicians. The measure was developed via an iterative process of reviewing existing PROMIS items and developing and testing new HF items. In a 600-patient sample, we estimated reliability (internal consistency; test-retest, with n=100 participants). We conducted validity analyses using Pearson r and Spearman ρ correlations with Kansas City Cardiomyopathy Questionnaire subscores. In a longitudinal sample, we performed responsiveness testing (paired t tests) with 75 patients with HF receiving interventions with expected health status improvement. The PROMIS-Plus-HF measure comprises 86 items (64 existing; 22 new) across 18 domains. Internal consistency reliability (Cronbach α) coefficients ranged from 0.52 to 0.96, with α≥0.70 in 12 of 17 domains. Test-retest intraclass correlation coefficients were ≥0.90. Correlations with Kansas City Cardiomyopathy Questionnaire subscores supported expected convergent ( r /ρ>0.60) and divergent validity ( r /ρP Conclusions: The PROMIS-Plus-HF profile measure—a complete assessment of physical, mental, and social health—exhibited good psychometric characteristics and may facilitate patient-centered care and research. Subsets of domains and items can be used depending on the clinical or research purpose.
Databáze: OpenAIRE