RESUMEN
INTRODUCTION: The Innocor® device uses an insoluble gas (SF 6 ) to estimate lung volume and the rate of disappearance of a soluble gas (nitrous oxide) to measure pulmonary blood flow (PBF), which approximates cardiac output assuming no shunt. We sought to identify error in the measurement of the insoluble gas in an effort to reduce variation in Innocor® measurement. METHODS: We enrolled 28 participants from the Dallas Heart Study (mean age, 63 yr; 57% men; 43% White). Stroke volume was measured at rest and at submaximal (20 and 40 W) exercise using both echocardiography (Philips iE33) and the Innocor® device. We defined a priori peak and equilibrium SF 6 measurement errors as greater or less than 20% of the mean observed value. Three Innocor measurements were obtained at rest ( n = 27) for a total of 81 measurements. Of these, 22% had SF 6 measurements that fell outside of the a priori range. RESULTS: Resting Innocor® stroke volume measures with peak SF 6 measured above a priori range (>0.12%) was associated with larger stroke volumes compared with stroke volume measures without peak SF 6 error (101.4 [26.8] vs 64.9 [8.7] mL; P = 0.006) and overestimated stroke volume when compared with stroke volume by echo (101.4 [26.8] vs 59.9 [16.3] mL; P = 0.017). A similar pattern was observed at submaximal exercise. In contrast, there was no consistent association between variation in equilibrium SF 6 concentrations and measured stroke volume. CONCLUSIONS: Variability in peak SF 6 concentration is common while using the Innocor® device and results in overestimated stroke volume. These findings have implications for research protocols using this device.