Purpose
An accurate determination of energy expenditure is needed to prevent postoperative complications and provide effective medical care for surgical patients. The aim of this study was to compare measured resting energy expenditure (REE) in patients undergoing pancreaticoduodenectomy (PD) using an indirect calorimetry (IC) with the help of five predictive equations. Methods: The subjects chosen for the study were 18 patients (12 males, 6 females) who underwent PD at Chonbuk National University Hospital between March 2010 and October 2011. REE was measured by IC (ICREE) and calculated with the following five predictive equations - (eqREE)-Harris-Benedict (HB), Mifflin St. Jeor (MI), American College of Chest Physician (AC), Ireton- Jones (IJ) and Cunningham (CU) equation - on postoperative day 3. Results: The mean measured REE by IC was 1,117.9 ±180.2, whereas the mean predicted REE by HB, MI, AC, IJ and CU equations were 1,241.3±156.1, 1,226±153.0, 1,401.7±188.0, 1,355.1±342.7, and 1,324.8±188.7 kcal/ day, respectively. The mean predicted REE by HB, MI, AC, IJ, and CU were found to be significantly different from the measured ICREE. The intraclass correlation coefficient (ICC), which describes the degree of agreement with the ICREE, was significantly highest in CUeqREE (r= 0.679). The rate of accuracy of HBeqREE, MIeqREE, ACeqREE, IJeqREE, and CUeqREE for ICREE was 27.8%, 50.0%, 27.8%, 22.2%, and 23.5%, respectively, in individual subjects. Conclusion: The intraclass correlation coefficient (r= 0.124∼0.679) and accuracy rate (22.2∼50.0%) determined from the results of REE measured by IC and that calculated by five predictive equations were low in overall. The low accuracy rate of energy expenditure by equations, which apply uniform indicators, can be explained by ongoing metabolic changes of the patient. Thus, the use of IC is recommended to assess energy requirements in PD; when this method is not practical in clinical settings, nutrition intervention is needed through regular monitoring because of the low accuracy rate of the predictive methods and ICREE that can cause inappropriate energy management, leading to metabolic disorders. (JKSPEN 2013;5(1):24-30)