A simplified echocardiographic formula to estimate cardiac index in the intensive care unit

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Measurement of cardiac index (CI) is crucial in the hemodynamic assessment of critically ill patients. Pulmonary artery catheterization (PAC) remains the gold standard for CI evaluation, but it's invasive. Doppler trans-thoracic echocardiography (TTE) has been proposed as an accurate bedside non-invasive alternative to PAC for CI estimation, based on measurement of left ventricular outflow tract velocity time integral (LVOT-VTI) and diameter (LVOT-d). However, intrinsic limits of geometrical assumptions and practical limitations related to the intensive care unit (ICU) setting may lead to inaccurate CI calculation, mainly due to errors in LVOT-d. A body surface area (BSA)-based formula to predict LVOT-d has been proposed and validated. Including the predicted LVOT-d in the CI equation we obtained a new "simplified formula" for TTE CI estimation: CI=2,17*LVOT-VTI*HR/1000. Purpose This study aimed to assess the accuracy of a new TTE-based simplified formula, only requiring LVOT-VTI and HR, for TTE estimation of CI. Methods We prospectively enrolled all consecutive patients admitted to the ICU requiring PAC from January to December 2021 (n=74). We measured TTE-based CI method using both the "traditional formula" (traditional CI TTE) and the new "simplified formula" (simplified CI TTE), and compared the accuracy of both with the PAC-derived CI (CI PAC) as the reference method. Results As shown in Figure 1a, there was a moderate correlation (r = 0.70, p < 0.001) between traditional CI TTE and the CI PAC. A plot of the differences between traditional CI TTE and the CI PAC is illustrated in Figure 2a, resulting in a mean difference of -0.18 ± 0.58 L/min/m2, and 95% limits of agreement of -1.32 to +0.96 L/min/m2. The correlation between simplified CI TTE and CI PAC was strong (r = 0.81, p < 0.001) (Figure 1b) and resulted significantly greater than the traditional CI TTE and CI PAC correlation (p < 0.05). On Bland-Altman plot analysis (Figure 2b), the mean difference between simplified CI TTE and CI PAC was +0.19 ± 0.48 L/min/m2 with 95% limits of agreement of -0.74 to +1.13 L/min/m2. Conclusion In this study, we validated an easy-to-use and practical approach, leveraging on TTE LVOT-VTI and HR only, for non-invasive estimation of CI in ICU patients. In this peculiar setting, our formula was more reliable than the traditional CI TTE formula. However, despite the strong correlation with CI PAC, our formula introduces a slight but systematic overestimation of CI (+0.19 L/min/m2), as illustrated in the Bland-Altman plot. Therefore, this formula may be particularly useful in longitudinal CI assessment rather than precisely quantifying CI on a single measurement. This approach may thus be apt to quickly assess the hemodynamic response to pharmacological and mechanical supports during longitudinal assessment.