ECMO patient energy requirements: A descriptive, retrospective cohort study.

Background: Indirect calorimetry (IC) in patients receiving extracorporeal membrane oxygenation (ECMO) is complicated. This study presents a novel IC method for this population and compares energy expenditure (EE) from IC with predictive equations.

Methods: IC was performed at the native lung using a Q-NRG+ indirect calorimeter. A CO2 sensor connected to the ECMO oxygenator primary exhalation port collected FeCO2 during IC studies. ECMO-VO2 and ECMO-VCO2 were calculated from sweep gas oxygen concentration, FeCO2, and sweep flow. EE was calculated from the combined readings. EE was compared with 25 kcal/kg, 30 kcal/kg, and Mifflin St Jeor. Subanalysis compared EE over time (ECMO days 1-3, 4-10, 11-21, and >21) and between venoarterial and venovenous patients.

Results: In total, 90 assessments in 52 patients were analyzed. The cohort was 67.3% male with a median age of 54 years, and median ECMO duration of 207 h. EE was 1523 ± 432 kcal/day (18.9 ± 6.9 kcal/kg/day). Energy needs did not vary significantly over time (P = 0.24); however, readings from days 11 to 21 were higher than days 1-3 (P = 0.0497). No significant differences between cannulation types were observed. EE was significantly lower than all predicted results (P < 0.001). Mean difference between EE and predicted energy ranged from 413 to 1099 kcal/day. No equation was strongly correlated with EE (rs = 0.15-0.61) overall or after stratification by cannulation type.

Conclusions: This study presents a viable method for incorporating IC in patients receiving ECMO. Using this method, EE in patients receiving ECMO was significantly lower than predicted. Using IC may help prevent overfeeding.