Resting energy expenditure in injured, septic, and malnourished adult patients on intravenous diets.

Energy expenditures of 237 adult patients and 37 normal subjects receiving all nutrition intravenously were analysed retrospectively. Patients were classified as nutritionally depleted (67), post-operative (96), injured (43), or septic/depleted (31). Groups were further divided into those receiving either: (1) only 5% dextrose (D5W); (2) hypocaloric regimens including glucose and amino-acids; and (3) eucaloric or hypercaloric total parenteral nutrition (TPN) which also included fat. Resting energy expenditures (REE) of normal subjects on D5W were only 85% of predicted basal values based on either the Aub-Du Bois or Harris-Benedict equations. During D5W infusions, increases for the patient groups, above these values for normal subjects, varied depending on whether they were based on absolute values or ratios to predicted values. They were: (1) 1-11% for depleted; 1-21% for post-operative; 28-30% for injured; and 18-30% for septic/depleted patients. The average increase in REE with TPN was 10%. Variability within the patient groups was high, reducing the utility of these values as a basis for estimating energy requirements of patients needing artificial nutrition. Coefficients of variation averaged 15% across patient groups when the data were expressed in kJ/kg, and were reduced only slightly, to 12%, when data were expressed as ratios to predicted values. Thus, 1 3 of the patients would differ by more than 12% from mean values, and 1 out of 20 by more than 24%. Properly performed measurements of individual energy expenditure are therefore superior to values predicted from equations or average values previously obtained from patient groups and should be used wherever possible, particularly in the very sick.