Predicted vs measured energy expenditure in critically ill, underweight patients.

A retrospective analysis was conducted to compare 4 energy-prediction equations against measured resting energy expenditure (MREE) determined via indirect calorimetry. Data from a heterogeneous group of 42 critically ill, severely underweight (59.50 +/- 17.30 kg; 77.1 +/- 9.7% ideal body weight [IBW]) male patients were assessed. The Hamwi formula was used to determine IBW. The Harris-Benedict (HB) equation was calculated for patients <90% IBW using both current body weight (CBW) and IBW. Energy needs were also estimated with an Ireton-Jones formula for all mechanically ventilated patients (n = 37). For patients <85% IBW (n = 31), an adjusted body weight was determined ([CBW + IBW]/2) and used in the HB formula. The HB formula using the IBW, CBW, and adjusted body weight was significantly different (p < .05) than MREE. The Ireton-Jones equation was not significantly different (p > .05) from MREE but tended to overestimate energy needs (109.3% +/- 16.8% MREE). Conversely, using the CBW or IBW in the HB underestimated the patient's energy needs; 77.0% +/- 11.6% MREE and 90.9 +/- 16.1% MREE, respectively. For patients <85% IBW, use of the adjusted body weight in the HB represented 84.2% +/- 13.9% MREE. The average caloric need was 31.2 +/- 6.0 kcal/kg CBW. Indirect calorimetry remains the best method of determining a patient's energy needs. Until a large prospective trial is conducted, a combination of prediction equations tempered with clinical judgment and monitoring the appropriateness of the nutrition prescription remains the best approach to quality patient care.