Sensitivity and specificity of different exercise oscillatory ventilation definitions to predict 2-year major adverse cardiovascular outcomes in chronic heart failure patients

Background: Exercise oscillatory ventilation (EOV) shows a four-fold greater risk of adverse events. This study aims to analyze the sensitivity and specificity of three EOV diagnostic definitions to predict adverse outcomes at a 2-year follow-up and to compare its EOV prevalence and relations with the patient's profile.Methods: Cardiopulmonary exercise tests from 233 heart failure patients were analyzed. Two blinded reviewers used a semiautomated software to identify EOV cases pattern according to the definitions of Ben-Dov, Corr`a, and Leite. Data were grouped in EOV-positive or EOV-negative according to each definition. Baseline characteristics, EOV prevalence, relative risk, sensitivity, and specificity to predict 2-years of major adverse cardiovascular outcomes were analyzed.Results: The Corra definition led to the best prediction of 2-year major cardiovascular adverse outcomes (HR 2.46 [1.16 to 5.25]; p = 0.019, AUC = 0.618; p = 0.007). EOV prevalence was 17.2%, 17.2%, and 9.4% applying BenDov, Corr`a, and Leite definition, respectively. The main clinical differences between EOV-positive and EOVnegative patients were: MECKI score and VE/VCO2 slope (all definitions), and BNP levels (Ben-Dov and Leite). BNP levels were correlated with amplitude (rho = 0.255; p = 0.033) and cycle length (rho = 0.388; p = 0.002).Conclusion: Corra definition was the only one that exhibited the capacity to predict major adverse cardiovascular outcomes at a 2-year follow-up. Regardless of its definition, EOV was more often prevalent in patients with a greater MECKI score and VE/VCO2 slope values.