Indirect calorimetry compared with predictive equations for resting energy expenditure in patients with intestinal failure

Background Accurate assessment of energy requirements is essential for patients with intestinal failure (IF) to avoid complications associated with parenteral nutrition. Predictive equations are commonly used in clinical practice to estimate resting energy expenditure (REE), however, evidence supporting their accuracy in IF patients is limited. Indirect calorimetry (IC) is the gold standard for measuring REE but is not routinely available. This study evaluated the agreement between routinely used predictive methods in the UK and IC-measured REE in patients with IF.Material and methods A service evaluation was conducted at a tertiary IF centre. Adult inpatients with IF had REE estimated using predictive equations based on PENG and ESPEN guidance, followed by IC measurement. Agreement between predicted and measured REE was assessed using mean bias, Bland-Altman analysis, and proportion of estimates within ±10% of IC values. Associations between prediction accuracy and IF aetiology, age, sex, and body mass index (BMI) were explored using ANOVA/linear regression.Results and conclusion Indirect calorimetry was performed in 91 patients (mean age 55 ± 17.4 years, mean BMI 21.6 ± 5.8 kg/m2). Mean measured REE was 1449 ± 297 kcal/day. Predictive equations showed poor agreement with IC at an individual level, with wide limits of agreement. Mean bias was + 18.6 kcal/day (-585 to +622 kcal/day) for PENG and + 94.9 kcal/day (-620 to +809 kcal/day) for ESPEN. Only 33.7 (PENG) and 30.4% (ESPEN) of estimates were within ±10% of measured REE. Predictive equations are inaccurate in IF patients, supporting routine use of IC to enable precise, individualised energy prescription.