ENERGY NEEDS ARE POORLY PREDICTED IN CRITICALLY ILL ELDERLY

Changes in energy expenditure with age have been described, but this p hysiology is not routinely considered when managing critically ill eld erly patients. To allow us to avoid the potential problems associated with underfeeding or overfeeding the critically ill elderly population , with approval of the human studies committee and appropriate consent from legal guardians, 25 critically ill patients over 65 years of age requiring mechanical ventilation underwent expired gas indirect calor imetry. If they had a pulmonary artery catheter in place far clinical reasons, reverse-Fick indirect calorimetry was also performed. Data ob tained by indirect calorimetry was compared with commonly applied equa tions for predicting energy expenditure by statistical methods of corr elation and limits of agreement. These 25 patients had an average age of 74 +/- 1.23 (standard error of the mean) and an average APACHE II s core of 15. Predictive equations correlated poorly with measured resti ng energy expenditure, and although they showed reasonable bias, they were imprecise in their estimation of resting energy expenditure. Thes e data suggest that energy expenditure in critically ill, mechanically ventilated elderly patients is highly variable. Although generally ov erestimating energy needs, currently available equations for predictin g energy expenditure in this population are associated with significan t bias and imprecision, which may lead to both overfeeding and underfe eding. Although these equations may be suitable as a basis of initiati ng nutritional support, energy provisions should ideally be guided by indirect calorimetry.