Resting energy expenditure in severely burned children: Analysis of agreement between indirect calorimetry and prediction equations using the Bland–Altman method

Abstract

Introduction Knowledge of a child's resting energy expenditure (REE) is essential in optimizing nutritional support for severely burned children. The provision of adequate nutritional support is vital in order to avoid the consequences of malnutrition or overfeeding. Nutritional requirements for severely burned children are often based on equations for estimates of REE. The accuracy of the predictive equations of REE has been questioned and many authors have advocated the measurement of REE. This study tests the hypothesis that estimates of REE vary significantly from measured REE (MREE) in a population of severely burned children, and are not accurate for determining nutritional requirements. Methods In 91 severely burned children aged between 3 and 18 years, REE was measured by indirect calorimetry (MREE) at the height of the hypermetabolic response and compared with predicted equations (PREE) from the Food and Agriculture/World Health Organization/United Nations University (FAO/WHO/UNU), Schofield- HW and Harris–Benedict. Agreement between indirect calorimetry and predicted equations was assessed following the Bland–Altman method. Results In the entire cohort group, predicted REE from all three equations were significantly lower compared to MREE ( p < 0.05). There was poor agreement between the MREE and predicted using all three equations. The Schofield- HW equation showed the lowest mean MREE–PREE difference: 635 ± 526 kcal/day (limits of agreement −608 and 1878 kcal/day; 95% confidence interval for the bias 525–745 kcal/day). Additionally, all three equations under predicted REE and were not significantly different from one another ( p = 0.98). Conclusions Until more accurate predicted equations are developed, we recommend indirect calorimetry measurements for determining resting energy expenditure in severely burned children.