Reliability and validity of dyspnea sensory quality descriptors in heart failure patients treated in an emergency department

BACKGROUND: Sensory qualities of dyspnea are known to differ by diagnosis. Less is known about whether sensory qualities vary with changes in health status in a given diagnosis. PURPOSE: The goal of this study was to evaluate the reliability, validity, and factor structure of dyspnea sensory quality descriptors in patients with heart failure (HF) treated in an emergency department (ED) and to investigate whether change in sensory quality influences HF patients to seek care in an ED. METHODS: HF patients (N = 57) treated in an ED were interviewed retrospectively. Open-ended characterization of dyspnea at the time of the ED visit were analyzed qualitatively. A subset of subjects (n = 34) rated the intensity of 13 dyspnea descriptors (0 = not endorsed; 1 = very mild; 10 = very severe) as the descriptors applied to the time at which they decided to come to the ED (Decision) and a week before the visit (Week Before). Descriptor ratings were analyzed for congruence with open-ended characterizations, endorsement frequency, internal consistency, factor structure, and correlations (by descriptor and within subjects) between the 2 time frames. RESULTS: Open-ended characterizations of dyspnea provided support for the content validity of most descriptors. Internal consistency of numerical ratings was high (α <0.90) in both recalled time frames. Factor analysis of descriptor ratings was unifactorial for Week Before, but suggested multiple sensory quality factors at Decision (suffocation, air hunger, effort/impedance, and, possibly, rate). Within-subject concordance and descriptor-by-descriptor correlations across time frames were mostly low, suggesting change in sensory quality from Week Before to Decision. Correlations in descriptor ratings were lowest among subjects who reported duration of dyspnea (as severe as at Decision) of 3 days or less. Subjects who recalled a duration of 6 days or more gave highly concordant ratings across both time frames. CONCLUSIONS: Sensory quality descriptor-based ratings were internally consistent and content valid. Low correlations in rating of sensory quality for most subjects across recalled time frames suggest that change in sensory quality may be an aspect of perceived increases in dyspnea severity before an ED visit. Results require replication and extension with larger samples and other diagnoses.