Demographic data and baseline characteristics.

ObjectivesThis study aimed to assess the accuracy of bedside ultrasound in predicting resting energy expenditure (REE) in critically ill patients.MethodsWe studied critically ill patients admitted to our hospital’s ICU between November 2021 and March 2023 who underwent REE, cardiac ultrasound, and muscle ultrasound evaluations. General demographic information and ultrasound data (including cardiac output, biceps brachii and quadriceps femoris thickness) were collected to estimate REE (REE-US). Simultaneously, REE was measured using indirect calorimetry (REE-IC). Correlations between REE-US and established equations (Harris-Benedict, Penn State University (PSU), Mifflin, ASPEN standard) as well as REE-IC were evaluated. Additionally, the feasibility and application of ultrasound for REE prediction across different disease conditions in critically ill patients were analysed.ResultsNinety-seven critically ill patients with 124 ultrasound measurements were included. The Penn State University formula showed the highest correlation with REE-IC (r = 0.779, p 2) (r = 0.521, p = 0.009) but comparable to the PSU formula in normal and high BMI groups (BMI 20–30 kg/m2: r = 0.682 vs. r = 0.714, p = 0.5743; BMI > 30 kg/m2: r = 0.712 vs. r = 0.882, p = 0.1294). In subgroup analysis, REE-US performed similarly to the PSU formula in the sepsis subgroup (r = 0.612 vs r = 0.661, p = 0.6852) and ICU patients in the late period of the acute phase (r = 0.675 vs r = 0.751, p = 0.2762).ConclusionsThe Penn State University formula demonstrated the strongest correlation with REE-IC in critically ill patients. Ultrasound may replace the PSU formula in non-mechanically ventilated patients with unavailable gas measurement parameters. However, ultrasound-derived REE is less predictive in patients with low BMI or during the early acute phase of critical illness. Further research is warranted to refine ultrasound application in these populations.