Félix A. Vega, Servicio de Cardiología, Sanatorio Güemes, Buenos Aires, Argentina
Joaquín Perea, Servicio de Cardiología, Sanatorio Güemes, Buenos Aires, Argentina
Victoria Rizzo-Cismondo, Servicio de Cardiología, Sanatorio Güemes, Buenos Aires, Argentina
Miguel De Larrechea, Servicio de Cardiología, Sanatorio Güemes, Buenos Aires, Argentina
Jorge Aguirre-de la Camara, Servicio de Cardiología, Sanatorio Güemes, Buenos Aires, Argentina
David Castañeda, Servicio de Cardiología, Sanatorio Güemes, Buenos Aires, Argentina
Cristian Silva, Servicio de Cardiología, Sanatorio Güemes, Buenos Aires, Argentina
María I. Sosa-Liprandi, Servicio de Insuficiencia Cardíaca, Miocardiopatías e Hipertensión Pulmonar, Sanatorio Güemes, Buenos Aires, Argentina
Marina Pérez-Bec, Servicio de Insuficiencia Cardíaca, Miocardiopatías e Hipertensión Pulmonar, Sanatorio Güemes, Buenos Aires, Argentina
Melisa Elfman, Servicio de Insuficiencia Cardíaca, Miocardiopatías e Hipertensión Pulmonar, Sanatorio Güemes, Buenos Aires, Argentina
Introduction: Estimating the mortality risk in patients hospitalized for acute decompensated heart failure (ADHF) is essential to guide clinical management. However, currently available prognostic stratification models during hospitalization have limitations in their predictive performance. In this context, we assessed the performance of the ADHERE risk score for predicting in-hospital mortality, and evaluated whether the addition of biomarkers improves its predictive ability. Method: A cohort of adult patients hospitalized for ADHF was analyzed. The ADHERE score was calculated using systolic blood pressure, serum creatinine, and blood urea nitrogen (BUN). The cohort was randomly divided into a training set (70%) and a testing set (30%). In the training set, a logistic regression model including the ADHERE score was fitted, along with an extended model incorporating B-type natriuretic peptide (BNP) and troponin. Both models were evaluated in the testing set using the area under the receiver operating characteristic curve (AUC), the DeLong test for comparison between curves, and net reclassification improvement (NRI). Results: A total of 300 patients were included, with an in-hospital mortality rate of 16.7%. The ADHERE score was significantly associated with in-hospital mortality (odds ratio [OR]: 1.88; 95% confidence interval [95% CI]: 1.31-2.73), although its discriminative ability was limited (AUC: 0.58; 95% CI: 0.40-0.75). The addition of BNP and troponin, both independently associated with higher mortality risk, allowed the development of an extended model with a significant improvement in discrimination (AUC: 0.70; 95% CI: 0.56-0.85; p = 0.04). Net reclassification analysis showed improved identification of patients who died (positive NRI: 0.73; 95% CI: 0.29-1.00), at the expense of increased misclassification among survivors (negative NRI: –0.35; 95% CI: –0.54 to –0.17). Internal calibration was adequate. Conclusions: In this cohort, the ADHERE score was significantly associated with in-hospital mortality, although with limited discriminative performance. The addition of BNP and troponin improved the model’s discriminative ability and identification of high-risk patients, suggesting an incremental value of biomarkers in the prognostic stratification of acute heart failure.
Keywords: Heart failure. ADHERE. B-type natriuretic peptide. Troponin.