Accuracy of predictive equations for resting energy expenditure estimation in mechanically ventilated Thai patients

Abstract

Abstract Background: Indirect calorimetry (IC) is the most precise approach for estimating calorie demand in critically ill patients. Despite this, owing to unaffordable devices, it is rarely used in practice. Predictive equations are the alternatives. Objectives: To assess the accuracy of 14 predictive resting energy expenditure(REE) equations in ventilated Thai patients. Methods: We compared the accuracy and agreement of 14 equations. The equations included the American College of Chest Physicians(ACCP) equation, Harris–Benedict equation(HBE), 1.2×HBE, 1.5×HBE, Mifflin–St. Jeor(MSJ), Ireton-Jones 1992 and 2002, Penn State 2003(HBE and MSJ) and 2010, Swinamer 1990, Faisy, Brandi 1999, and 25 kcal/kg equation. An equation was ascertained as accurate if the calculated values fell within ±10% of the measured REEs. Spearman correlation coefficient, Bland–Altman method, and intraclass correlation coefficient were used to analysis. Results: We obtained data from 24 ventilated patients undergoing REE measurement by IC. Fifty percent of them were male with a median age of 64.5 years, a median height of 160 cm, and a median body mass index of 22.95 kg/m2. The predictive precision of all equations was poor, with largely different accuracies from 6.7% to 48.1%. The most reliable equation was Penn State 2010. The ACCP, HBE, MSJ, and Penn State 2003(HBE) tended to underestimate calorie need. Contrastingly, the other equations tended to overestimate REEs. Despite a moderate degree of correlations, the Bland–Altman plots demonstrated clinically unacceptable discrepancies between measured REE and REE calculated by each equation. Conclusions: In ventilated Thai patients, there were no precise equations for determining REE.