Accuracy of 30-minute indirect calorimetry studies in predicting 24-hour energy expenditure in mechanically ventilated, critically ill patients.

BACKGROUND

There is no consensus regarding the optimal duration of measurement or time of day to perform indirect calorimetry (IC). Energy expenditure (EE) varies at different times of day and with different activity levels. We sought to assess the variability of EE in mechanically ventilated patients over a 24-hour period and the accuracy of 30-minute IC studies in predicting the 24-hour energy expenditure (EE24).

METHODS

The study was a prospective comparison between the resting EE obtained by 30-minute measurement of IC and EE values obtained from 24-hour measurements. Tests were performed in the Medical Intensive Care Unit (MICU) of a tertiary care, university hospital. Oxygen consumption (VO2) and carbon dioxide production (VCO2) were measured for 24 hours in eight ventilated patients. Measurements were made every 3 minutes and used to calculate 30-minute and 24-hour oxygen consumption values. EE24 was calculated using the modified Weir equation. Each 30-minute interval was compared with the value obtained from the 24-hour measurement.

RESULTS

Three hundred forty-one of 384 30-minute intervals remained for analysis. Average EE24 measured was 1490 +/- 486 kcal/d. Average EE24 predicted by extrapolation from 30-minute studies was 1501 +/- 503 kcal/d, with a mean difference of 0 +/- 209 kcal/d from the measured 24-hour values (range: -1068 to +585 kcal/d). Thirty-minute studies were within 20% of 24-hour measurements for 89% of intervals. The difference between 24-hour and 30-minute studies correlated with changes in minute ventilation (VE), heart rate, systolic blood pressure, and breath rate from their 24-hour means (p < .001). The mean error of EE estimate was greatest between 3 and 11 PM (p < .001).

CONCLUSIONS

We conclude the following: (1) EE in MICU patients is variable; (2) 30-minute IC studies predict measured EE24 acceptably well for clinical purposes; and (3) accuracy is maximized if a 30-minute study is performed between 11 PM and 3 PM, and when Ve, heart rate, systolic blood pressure, and breath rate are near the day's average.