Formulaic methods of estimating calorie requirements in mechanically ventilated obese patients: a reappraisal.

A variety of formulae are used by clinicians to estimate calorie requirements in hospitalized obese patients. Many of these formulae have been validated by determining their correlation to calorie requirements as determined by indirect calorimetry (IC). The use of correlation coefficients (CC) in this manner may be misleading. A more appropriate statistical methods, limits of agreement (LA), for comparing clinical measurements has recently been described by Bland and Altman. We obtained a total of 188 IC measurements on 113 obese (> 120% IBW), mechanically ventilated patients. These were compared by CC and LA to seven formulaic methods of determining calorie requirements. All formulae correlate significantly with IC. Analysis by LA indicated acceptable bias in several formulae (eg, 21 kcal/kg, Ireton-Jones obese patients (IJ), Ireton-Jones ventilatory patients, (VEE); however, the uniform lack of precision demonstrated by all formulae renders their clinical utility questionable.