Prognostic value of health-related quality of life in patients with acute dyspnea

Abstract Background Previous studies have shown the prognostic value of health-related quality of life (HRQL) in stable and ambulatory chronic heart failure patients. However, it is unknown whether HRQL can predict all-cause mortality in patients presenting to the emergency department (ED) after acute onset of symptoms. In order to address this unmet need, the aim of this study was to assess the prognostic value of HRQL in patients with acute dyspnea caused by acute heart failure (AHF) and other dyspnea aetiologies for 360-day mortality. Purpose To assess prognostic value of HRQL using the generic EQ-5D and visual analogue scale (EQ VAS) in patients with acute dyspnea. Methods Basics in Acute Shortness of Breath EvaLuation (BASEL V) is a prospective, multicenter, diagnostic study enrolling adult patients presenting with acute dyspnea to the ED. For this analysis, only patients with a complete set of variables necessary for calculation of EQ-5D (range 0–10; with higher score indicating worse HRQL) and EQ VAS (range 0–100; with 100 being the best imaginable health state) at baseline were included. The endpoint was the prognostic value of EQ-5D and EQ VAS at 360 days of follow-up regarding all-cause death. Prognostic accuracy was calculated using c-statistics. In a cox regression analysis EQ-5D was treated as both, a continuous and categorical variable. Adjustments were made for clinically relevant covariates (age, sex, orthopnoea, edema, level of N-terminal pro-B-type natriuretic peptide (NT-proBNP) at presentation, history of coronary artery disease and chronic obstructive pulmonary disease, diuretics, β-blockers and ACE-inhibitors at discharge). Results Among 2605 patients enrolled, 1141 (43,8%) had a complete set of variables allowing the calculation of EQ-5D and EQ VAS. Of these patients 594 (52.1%) had an adjudicated final diagnosis of AHF. 211 (18.5%) patients died within 360 days of follow-up. Median EQ-5D was 3 (interquartile range (IQR) 1.5–5) and median EQ VAS was 50 (IQR 40–70). The prognostic accuracy for 360-day mortality was 0.65 (95% confidence interval ((CI) 0.61–0.69) and 0.58 (95% CI 0.54–0.62) for EQ-5D and EQ VAS, respectively (p=0.002). After combining EQ-5D and EQ VAS in a logistic regression model c-statistics regarding all-cause mortality within 360 days did not improve. The prognostic accuracy of EQ-5D was comparable to that of NT-proBNP (c-statistics 0.69, p=0.385). In an adjusted cox regression analysis the hazard ratio for patient