Predictive equations for energy needs for the critically ill

Renee N. Walker, Roschelle A. Heuberger

Research output: Contribution to journalReview articlepeer-review

106 Scopus citations


Nutrition may affect clinical outcomes in critically ill patients, and providing either more or fewer calories than the patient needs can adversely affect outcomes. Calorie need fluctuates substantially over the course of critical illness, and nutrition delivery is often influenced by: the risk of refeeding syndrome; a hypocaloric feeding regimen; lack of feeding access; intolerance of feeding; and feeding-delay for procedures. Lean body mass is the strongest determinant of resting energy expenditure, but age, sex, medications, and metabolic stress also influence the calorie requirement. Indirect calorimetry is the accepted standard for determining calorie requirement, but is unavailable or unaffordable in many centers. Moreover, indirect calorimetry is not infallible and care must be taken when interpreting the results. In the absence of calorimetry, clinicians use equations and clinical judgment to estimate calorie need. We reviewed 7 equations (American College of Chest Physicians, Harris-Benedict, Ireton-Jones 1992 and 1997, Penn State 1998 and 2003, Swinamer 1990) and their prediction accuracy. Understanding an equation's reference population and using the equation with similar patients are essential for the equation to perform similarly. Prediction accuracy among equations is rarely within 10% of the measured energy expenditure; however, in the absence of indirect calorimetry, a prediction equation is the best alternative.

Original languageEnglish
Pages (from-to)509-521
Number of pages13
JournalRespiratory Care
Issue number4
StatePublished - Apr 2009


  • Critical illness
  • Energy needs
  • Indirect calorimetry
  • Mechanical ventilation
  • Nutrition support
  • Prediction equations
  • Review


Dive into the research topics of 'Predictive equations for energy needs for the critically ill'. Together they form a unique fingerprint.

Cite this