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

• Published in: Circulation. Heart failure Vol. 12; no. 6; p. e005751
• Main Authors: Ahmad, Faraz S, Kallen, Michael A, Schifferdecker, Karen E, Carluzzo, Kathleen L, Yount, Susan E, Gelow, Jill M, McCullough, Peter A, Kimmel, Stephen E, Fisher, Elliot S, Cella, David
• Format: Journal Article
• Language: English
• Published: United States 01.06.2019
• Subjects: Adolescent; Adult; Female; Health Status; Heart Failure - rehabilitation; Heart Failure - therapy; Humans; Male; Middle Aged; Patient Reported Outcome Measures; Psychometrics; Quality of Life; Reproducibility of Results; Surveys and Questionnaires; Young Adult; health status; heart failure; quality of life; decision-making; patient-reported outcome measures
• ISSN: 1941-3289; 1941-3297
• DOI: 10.1161/CIRCHEARTFAILURE.118.005751

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/ρ<0.30). In the longitudinal sample, 10 of 18 domains had improved ( P<0.05) scores from baseline to follow-up. 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.