Energy expenditure and balance following pediatric intensive care unit admission: a longitudinal study of critically ill children

OBJECTIVE: Longitudinal comparison of prescribed energy, actually administered energy, and energy expenditure (EE) predicted by Schofield's equations to actual EE, as determined by daily indirect calorimetry measurements in critically ill children during the first 7 days following admission.

DESIGN: Observational study.

SETTING: Pediatric intensive care unit, high and medium care wards, in a university hospital.

PATIENTS: Forty-six mechanically ventilated and spontaneously breathing infants and children (0-18 yrs) who were admitted with sepsis or following major abdominal or thoracic surgery or trauma.

INTERVENTIONS: Daily indirect calorimetry measurements and assessment of energy balance.

MEASUREMENTS AND MAIN RESULTS: Energy balance studies were performed for a total of 298 admission days in 13 sepsis, 27 surgery, and 6 trauma patients. Indirect calorimetry measurements were performed on 89% of the days. Mean measured EE was 44.6 +/- 15 kcal/kg.d and equaled predicted EE (44.2 +/- 12 kcal/kg.d; p = .56). Measured EE did not change over time, neither overall nor in diagnostic subgroups. Overall, median (range) administered energy was 31.1 (0-119) kcal/kg.d, which was significantly lower than measured EE (p < .001) and predicted EE (p < .001). Patients were underfed on 60% of days and overfed on 28% of days. Administered energy rose significantly in the course of admission, independently of diagnostic category, and did not differ from prescribed energy (p = .42). Energy intake was significantly higher in sepsis patients than in surgery and trauma patients during the whole course of the study (p < .01). The cumulative energy balance was positive only in sepsis patients. The administration of parenteral feeding was the single significant factor determining energy intake in mixed-effect modeling.

CONCLUSIONS: Measured EE was stable and not significantly different from predicted values over the course of hospitalization. Underfeeding was frequently present and mainly due to prescription and administration of energy amounts inferior to measured EE values in enterally fed patients.