Nutrition Assessment With Indirect Calorimetry in Patients Evaluated for Left Ventricular Assist Device Implantation

Background: Malnutrition is known to negatively impact the clinical course of advanced heart failure and is associated with increased mortality following left ventricular assist device (LVAD) implantation. Appropriate assessment of nutrition requirements in these patients is critical in their clinical care, yet there has been little discussion on how to best determine resting energy expenditure (REE) in the hospital setting. We investigated the use of indirect calorimetry in a group of patients with advanced heart failure. Materials and Methods: Results from preoperative indirect calorimetry testing in 98 patients undergoing evaluation for LVAD candidacy were collected. REE was compared with 10 predictive equations that estimated caloric need based on a range of patient-specific demographic and clinical variables. Results: This study enrolled 22 female and 76 male patients with a mean age of 59.4 ± 12.5 years, body mass index of 29.6 ± 6.0 kg/m², and ejection fraction of 19.4 ± 6.6%. The average REE by indirect calorimetry in this group was 1610.0 ± 612.7 kcal/d. All predictive equations significantly overestimated REE. However, those equations intended for use in the critically ill demonstrated the greatest accuracy, with the Brandi equation achieving both the highest correlation (r = 0.605, P <.001) and the lowest standard error of the estimate (504.8 kcal/d). Conclusions: Indirect calorimetry may be reliably and safely used to determine caloric requirements in patients with advanced heart failure. The use of predictive equations based on demographic and clinical parameters appears to generate inaccurate estimations of REE in these patients. However, equations designed for use in critically ill patients better estimate nutrition requirements than those designed for healthy individuals.