Local temperature control improves the accuracy of cardiac output estimation using lung-to-finger circulation time after breath holding

As timely measurement of the cardiac index (CI) is one of the key elements in heart failure management, a noninvasive, simple, and inexpensive method of estimating CI is keenly needed. We attempted to develop a new device that can estimate CI from the data of lung-to-finger circulation time (LFCT) obtained after a brief breath hold in the awake state. First, we attempted to estimate CI from the LFCT value by utilizing the correlation between 1/LFCT and CI estimated with MRI. Although we could obtain LFCT from 45 of 53 patients with cardiovascular diseases, we could not find the anticipated relation between 1/LFCT and CI. However, we realized that when we adopted only LFCT from patients with a finger temperature of ≥31°C, we could obtain a consistent and clear correlation with CI (correlation coefficient, r =.81). Thus, we next measured LFCT before and after warming the forearm. We found that LFCT decreased after the local temperature increased (from 27.5 ± 13.6 to 18.4 ± 5.3 s, p < 0.01). The correlation between the inverse of LFCT and CI improved after warming (1/LFCT vs. CI, from r =.69 to r =.82). The final Bland–Altman analysis between the measured and estimated CI values revealed that the bias and precision were −0.05 and 0.37 L min⁻¹ m⁻², respectively, and the percentage error was 34.3%. This study clarified that estimating CI using a simple measurement of LFCT is feasible in most patients and a low fingertip temperature strongly affects the CI-1/LFCT relationship, causing an error that can be corrected by proper local warming.