Validation and derivation of short-term prognostic risk score in acute decompensated heart failure in China

Background Few prognostic risk scores (PRSs) have been routinely used in acute decompensated heart failure (ADHF). We, therefore, externally validated three published PRSs (3A3B, AHEAD, and OPTIME-CHF) and derived a new PRS to predict the short-term prognosis in ADHF. Methods A total of 4550 patients from the Heb-ADHF registry in China were randomly divided into the derivation and validation cohorts (3:2). Discrimination of each PRS was assessed by the area under the receiver operating characteristic curve (AUROC). Logistic regression was exploited to select the predictors and create the new PRS. The Hosmer-Lemeshow goodness-of-fit test was used to assess the calibration of the new PRS. Results The AUROCs of the 3A3B, AHEAD, and OPTIME-CHF score in the derivation cohort were 0.55 (95% CI 0.53-0.57), 0.54 (95% CI 0.53-0.56), and 0.56 (95% CI 0.54-0.57), respectively. After logistic regression analysis, the new PRS computed as 1 x (diastolic blood pressure < 80 mmHg) + 2 x (lymphocyte > 1.11 x 10(9)/L) + 1 x (creatinine > 80 mu mol/L) + 2 x (blood urea nitrogen > 21 mg/dL) + 1 x [BNP 500 to < 1500 pg/mL (NT-proBNP 2500 to < 7500 pg/mL)] or 3 x [BNP >= 1500 (NT-proBNP >= 7500) pg/mL] + 3 x (QRS fraction of electrocardiogram < 55%) + 4 x (ACEI/ARB not used) + 1 x (rhBNP used), with a better AUROC of 0.67 (95% CI 0.64-0.70) and a good calibration (Hosmer-Lemeshow chi(2) = 3.366, P = 0.186). The results in validation cohort verified these findings. Conclusions The short-term prognostic values of 3A3B, AHEAD, and OPTIME-CHF score in ADHF patients were all poor, while the new PRS exhibited potential predictive ability. We demonstrated the QRS fraction of electrocardiogram as a novel predictor for the short-term outcomes of ADHF for the first time. Our findings might help to recognize high-risk ADHF patients.