Energy Expenditure in the ICU

Abstract

In this issue of Nutrition in Clinical Practice, Malone provides an overview of assessing energy expenditure in the ICU. Malone’s article complements two previous reviews in this journal regarding indirect calorimetry. The topics discussed include basic concepts in indirect calorimetry, appropriate measurement conditions, application of resting energy expenditure toward estimation of total energy expenditure, pitfalls and potential applications of the Fick equation, use of various predictive formulas, and indications for indirect calorimetry. It would be impractical to write an exhaustive review on indirect calorimetry as there are entire books dedicated to this subject. However, these three reviews from the journal would provide the reader with a solid foundation in the understanding of the use of indirect calorimetry in clinical practice. I felt compelled to use this editorial to embellish on or provide an alternative viewpoint regarding some of the issues discussed in Malone’s review. Sometimes in our eagerness to implement indirect calorimetry into our practice, we inadvertently omit several essential elements of the procedure that must be done before measuring patients. These elements are not often described in reviews. One of the most important of these elements is the validation of measurement conditions and techniques using healthy subjects. This is particularly pertinent if you plan to compare your resting energy expenditure measurements with normal or predicted values. For example, if you plan to use the Harris-Benedict equations as your normal or predicted values, then the vast majority of healthy subjects should fall within 10% or 15% of these values. Before measuring patients at the hospital, we measured 10 healthy subjects who were within 10% of ideal body weight. Nine subjects were within 10% of predicted values by the Harris-Benedict equations and the tenth was within 13% of predicted values. As a result, we felt more comfortable using the Harris-Benedict equation as our reference, or normal or predicted, energy expenditure when measuring patients at the hospital. Another area for further elaboration is the use of indirect calorimetry in patients who are ventilator dependent. Malone appropriately identifies some pitfalls in this area. In addition to the parameters identified in the review, the stability of the FiO2 should be monitored before the indirect calorimetry measurement. Certain ventilators tend to have more fluctuation in oxygen concentration delivery than others, resulting in additional errors in oxygen consumption measurements. We use oxygen consumption error sensitivity to provide a more reliable indicator of measurement error than respiratory quotient (RQ). This marker takes into account inspired oxygen concentration, concentration variability, and metabolic demand. Examining the data for a nonphysiologic RQ can only detect gross errors. A small fluctuation in inspired oxygen content (eg, 60.0% versus 60.5%) can result in approximately a 24% error in oxygen consumption sensitivity in an average-sized, unstressed individual. The error would be further magnified in a stressed, critically ill patient. It is also important to use a metabolic cart with a sensitive oxygen sensor for hospitalized ventilator-dependent patients. Our system uses a differential paramagnetic oxygen sensor that is sensitive to 0.01% (eg, it can detect differences from 50.01% to 50.02% in inspired oxygen content). This is in contrast to some other systems that may only be able to detect to the 0.1% level and may be more likely to be associated with greater error in patients receiving higher FiO2 concentrations. Use of predictive methods is always a concern given the variability of energy expenditure for a given disease. Nevertheless, I agree with Malone’s philosophy in giving guidance to those clinicians without the capability to measure energy expenditure. It was clearly pointed out that the “older estimates” grossly overestimate measured energy expenditure. The study by Foster et al in 100 consecutive patients requiring parenteral nutrition clearly illustrates that point. Of 191 published guidelines, approximately 50% of patient case-formula matches overestimated energy expenditure by up to 1000 kcal/d, and more than 45% of the matches Correspondence: Roland N. Dickerson, PharmD, University of Tennessee Health Science Center, 26 S. Dunlap St., Memphis, TN 38163.