A Comparison of Photoplethysmography Versus Esophageal Doppler for the Assessment of Cardiac Index During Major Noncardiac Surgery.

Abstract

In this prospective observational study, we compared changes in cardiac index (CI) during fluid challenge using photoplethysmography (PPG; Nexfin™) (CIPPG) versus esophageal Doppler (ED) (CIED) in major noncardiac surgery patients. Measurements were obtained when the attending anesthesiologist decided to perform a fluid challenge. Correlations with linear regression, Bland-Altman analysis, and analysis of covariance were performed. Trending ability was studied using 2 different methods: a 4-quadrant plot and a polar plot. Forty-three patients were analyzed with a total of 111 fluid challenges. There was a significant linear relationship between CI PPG and CI ED (r2 = 0.34; P < 0.001). The bias between the ED and the PPG measurements of CI was -0.114 (95% confidence interval [CI95], -1.9 to 1.7) L/min/m2, with a mean percentage error of 55%. The correlation between the changes in CI during a fluid challenge was significant (r2 = 0.25; P = 0.002). The concordance rate of directional changes (increase or decrease) of CI PPG and CI ED during fluid challenge was 67% (CI95, 57-75) for the whole data set and 85% (CI95, 70-94) with an exclusion zone of 15%. When considering ED as a reference, the sensitivity and specificity to give an additional bolus with PPG (increase in CI PPG ≥ 15%) were 35% (CI95, 19-55) and 90% (CI95, 81-96), respectively, with a positive predictive value of 58% (CI95, 33-80) and a negative predictive value of 78% (CI95, 68-86). In major noncardiac surgery patients, the evaluation of CI using PPG is not interchangeable with the evaluation of CI using ED. When considering the ED as an accurate device to assess changes in CI, PPG is not appropriate to assess the need for additional fluid administration. These results clearly indicate the limitations of PPG as an accurate device to track changes in CI compared with ED.