Impact of different diagnostic criteria on the prevalence and prognostic significance of exertional oscillatory ventilation in patients with chronic heart failure

Introduction Exertional oscillatory ventilation (EOV) occurs in many patients with chronic heart failure. Two different definitions of EOV have been proposed by Corra and Leite. We aimed to compare the prevalence of EOV and its prognostic significance in patients with chronic heart failure using the two diagnostic approaches. Methods Patients underwent a symptom-limited, treadmill-based exercise test with metabolic gas exchange measurements using the modified Bruce protocol. EOV was defined (i) as cyclic fluctuations in ventilation lasting for more than 60% of exercise duration, with an amplitude of greater than 15% of the average amplitude of cyclic fluctuations at rest (Corra) and/or (ii) as three or more regular oscillations with regularity defined if the standard deviation of three consecutive cycle lengths was within 20% of the average coupled with a minimal average amplitude of ventilatory oscillation of 51 (Leite). Results Two hundred and forty patients (mean age 59 +/- 13 years; 73% males; left ventricular ejection fraction 34 +/- 6%; peak VO2 21.0 +/- 4.6 ml/kg per min; VE/VCO2 slope 35 +/- 9) were included in the study. The prevalence of EOV was 25% using the Corra method and 31% using the Leite method. Fifty percent of patients diagnosed with EOV by the Corra criteria and 58% diagnosed by the Leite criteria had died at 12-month follow-up. EOV (Corra) was a predictor of mortality independent of peak VO2, VE/VCO2 slope, left ventricular ejection fraction, age, and 6-min walk test distance. A hazard ratio (H R) of 6.3 (P< 0.0001; 95% confidence interval = 1.6-25.2) was evident for the Corra method, which was higher than for the Leite method (HR=4.9; P<0.0001; 95% confidence interval=2.6-18.2). Conclusion The prevalence of EOV was between 25 and 31% depending on the criteria used to define it. The presence of EOV was a powerful predictor of adverse outcome, and diagnosed with the Corra criteria was associated with a higher H R than the Leite criteria. Eur J Cardiovasc Prev Rehabil 16:451-456 (C) 2009 The European Society of Cardiology