Psychometric Analysis of the Heart Failure Somatic Perception Scale as a Measure of Patient Symptom Perception

Background: Symptoms are known to predict survival among patients with heart failure (HF), but discrepancies exist between patients' and health providers' perceptions of HF symptom burden. Objective: The purpose of this study is to quantify the internal consistency, validity, and prognostic value of patient perception of a broad range of HF symptoms using an HF-specific physical symptom measure, the 18-item HF Somatic Perception Scale v. 3. Methods: Factor analysis of the HF Somatic Perception Scale was conducted in a convenience sample of 378 patients with chronic HF. Convergent validity was examined using the Physical Limitation subscale of the Kansas City Cardiomyopathy Questionnaire. Divergent validity was examined using the Self-care of HF Index self-care management score. One-year survival based on HF Somatic Perception Scale scores was quantified using Cox regression controlling for Seattle HF Model scores to account for clinical status, therapeutics, and lab values. Results: The sample was 63% male, 85% white, 67% functionally compromised (New York Heart Association class III-IV) with a mean (SD) age of 63 (12.8) years. Internal consistency of the HF Somatic Perception Scale was alpha =.90. Convergent (r = -0.54, P < .0001) and divergent (r = 0.18, P >.05) validities were supported. Controlling for Seattle HF scores, HF Somatic Perception Scale was a significant predictor of 1-year survival, with those most symptomatic having worse survival (hazard ratio, 1.012; 95% confidence interval, 1.001-1.024; P =.038). Conclusions: Perception of HF symptom burden as measured by the HF Somatic Perception Scale is a significant predictor of survival, contributing additional prognostic value over and above objective Seattle HF Risk Model scores. This analysis suggests that assessment of a broad range of HF symptoms, or those related to dyspnea or early and subtle symptoms, may be useful in evaluating therapeutic outcomes and predicting event-free survival.