AB138. SOH26AB_0039. Avoiding under and over feeding perioperative patients using indirect calorimetry in the intensive care unit

Background: Both over and under nutrition can negatively impact patient outcomes. Nutritional requirements were calculated by the dietitian using predictive equations to estimate resting energy expenditure (REE). These equations can be inaccurate. The gold standard for assessing REE is indirect calorimetry (IC). Following introduction of IC, our aim was to compare predictive energy equations with IC measurements.

Methods: Data was retrospectively collected over the first year of IC use. Data were collected from the intensive care unit (ICU) computer system and from the indirect calorimeter. Excel was used for data analysis. IC measurements were compared to the Penn State equation, the Faisy equation, and 25 kcal/kg.

Results: Out of 125 tests, perioperative accounted for 28% (n=125). Of the 125 tests, 42% of patients’ Penn State were within 10% of the IC measurement, and 66% of patients’ Penn State were within 20% of the IC measurement. In total, 23% of patients’ Faisy requirements (n=22) were within 90–110% of the IC measurement, and 59% were within 20% of the IC measurement. In total, 39% of patients’ 25 kcal/kg (n=75) were within 90–110% of the IC measurement, and 57% were within 20% of the IC measurement.

Conclusions: Under and over feeding can lead to increased length of stay, increased ventilation duration, and infection rates in perioperative patients. On average, more than one-third of the predictive equations were outside the range of 80–120% accuracy of the IC measurements. Out of all the predictive equations, the least accurate was 25 kcal/kg.

Keywords: Indirect calorimetry (IC); perioperative; predictive equations; nutritional requirements; resting energy expenditure (REE)