Increase Font Size  Decrease Font Size  View as PDF   Print  

In critically ill patients, what is the relationship between resting metabolic rate (RMR) and RMR predicted by the Ireton-Jones 1992 equations?

Conclusion Statement

Seven studies comparing RMR and the Ireton-Jones, 1992 equations report similar mean values, however, for an individual, energy predictions may be different by as much as 500 kcals (52% of non-obese subjects predicted within 10% of RMR). Further research regarding the accuracy of the Ireton-Jones, 1992 equation is warranted.

Evidence Summary

Statistical Background

The agreement between a measurement of energy expenditure and a prediction of energy expenditure using an equation can be evaluated by looking at the percentage of subjects falling within an acceptable range, correlation, bias, and precision. Since group mean data mask larger individual errors and also are not applicable to practice decisions for individual subjects, group mean data were not used to determine the conclusion statements.

In the results that follow, prediction accuracy is defined as the percentage of individuals in the study group whose RMR was predicted to within +/- 10% of measured RMR. When investigators reported the percentage of individual whose prediction was within 15% or within 20% of measured, these data are also given.

Relationships measured by correlation between measured and predicted RMR values can have a high r-value yet still systematically overpredict or underpredict the relationship.

The term bias is used statistically to describe the mean difference between the predicted and measured RMR, with 95% CIs. When the 95% CIs of the mean bias cross zero (i.e. the prediction is not consistently high or low), a prediction is considered "unbiased." Another approach to describing the extent of difference between predicted and measured RMR is the Bland Altman analysis, where the differences between predicted and measured RMR are plotted (with a line superimposed) against the mean difference with 95% CIs. If this line has a slope, bias is implied. The most accurate measures have a mean bias near zero, and 95% CI that are clinically acceptable.

The term precision is used statistically in indirect calorimetry research to describe the root mean square error (square root of the squared difference between predicted vs. measured) as a percentage of the measured RMR. When the 95% CI of the root mean square error is within 15%, a prediction is considered "precise."

Ireton-Jones, 1992 Equations

Spontaneously breathing IJEE (s) = 629 - 11 (A) + 25 (W) - 609 (O)

Ventilator dependent IJEE (v) = 1925 - 10 (A) + 5 (W) + 281 (S) + 292 (T) + 851 (B)

Equations use age (A) in years, body weight (W) in kilograms (kg), sex (S, male = 1, female = 0), diagnosis of trauma (T, present = 1, absent = 0), diagnosis of burn (B, present = 1, absent = 0), obesity >30% above initial body weight from 1959 Metropolitan Life Insurance tables or BMI > 27 (present = 1, absent = 0).

Evidence Summary

Seven studies (five positive-quality, two neutral-quality) compared measurements of energy expenditure with the original Ireton-Jones equations published in 1992. Using the Ireton-Jones, 1992 equations for ventilator-dependent and spontaneously breathing patients generally reports statistically significant correlation between the two methods and similar mean values.

In a positive quality cross-sectional study of 76 critically ill patients with a mean APACHE II score of 12.6 +/- 7.5, MacDonald and Hildebrandt, 2003 reported that the Ireton-Jones equation correlated with 24-hour measured RMR with an r = 0.49 and an r2 = 0.23 (P < 0.0001). The Ireton-Jones equation predicted RMR within 20% of indirect calorimetry values 63% of the time for the entire population studied.

Frankenfield et al, 2004 completed a positive quality cross-sectional study of 47 mechanically ventilated surgical, trauma and medical patients. The Ireton-Jones, 1992 equations correlated with RMR with an R2 of 0.57 (p < 0.05). The Ireton-Jones, 1992 equation was unbiased but not precise: 95% CI bias of –90 to 96 kcals and a precision of 10 – 16% of measured RMR.

Cheng et al, 2002 completed a positive quality cross-sectional study of 46 mechanically ventilated critically ill patients receiving EN, TPN or both. Mean APACHE II score was 18.0 +/- 4.2 and did not differ statistically between groups. Without an added stress factor, Ireton-Jones significantly overestimated the measured EE by 11.9% (1824.8 +/- 232.9 vs. 1631.3 +/- 345.3 kcal/day, P < 0.05), while the correlation was 0.672.

In a positive quality cross-sectional study of 36 mechanically ventilated patients on TPN, Flancbaum et al, 1999 reported that compared to RMR of 2005 +/- 464 kcal/day, the Ireton-Jones equation calculated a mean of 2156 +/- 332 kcal/day, underestimating RMR in 89% of patients. Although predicted RMR did not differ significantly from measured RMR, the correlation coefficient was 0.26 and the mean absolute difference was 386 kcals. The average difference was 150 kcals, with a lower limit of -67 kcals and an upper limit of 368 kcals.

Dickerson et al, 2002 completed a positive quality cross-sectional study of 24 thermally injured patients with >20% body surface burn and a mean Tobiasen Burn Severity Index of 7.3 +/- 2.0. For ventilator-dependent patients, the bias was –67 to 546 kcals/day, and the precision was 458 +/- 356 kcals/day (20% +/- 20% of RMR). For spontaneously breathing patients, the bias was –804 to 846 kcals/day, and the precision was 823 +/- 598 kcals/day (30% +/- 29% of RMR).

Campbell et al, 2005 completed a neutral quality cross-sectional study of 42 underweight (<90% IBW) critically ill males, 37 of them were mechanically ventilated. Measured RMR was 1790.1 +/- 365.2 kcals, while that measured using the Ireton-Jones equation was 1900.2 +/- 207.1 kcals, overestimating the patients’ energy needs (109.3% +/- 16.8%). The mean prediction error was 110.1 kcals/day, which was significantly correlated with measured RMR (r = 0.61, p < 0.05).

In a neutral quality cross-sectional study of 15 burn patients with obesity and 15 nonobese burn patients, Ireton-Jones, 1997 reported that measured EE and that predicted by the Ireton-Jones equations were not significantly different (nonobese: 2491 +/- 604 kcals/day vs. 2723 +/- 489 kcals/day, p = NS).

Overview Table

Article (worksheet) Quality Rating of Article Sample Description Subjects within +/- 10% Subjects within +/- 15% Subjects within +/- 20% Bias Precision Difference r2
MacDonald et al 2003 Positive 76 critically ill patients 28% 43% 63% Not reported Not reported 2343 +/- 517 kcals (RMR) vs 1933 +/- 251 kcals (IJ) 0.23
Frankenfield et al 2004 Positive 47 mechanically ventilated surgical, trauma and medical patients, both obese and non-obese 60% 68% Not reported 95% CI bias of –90 to 96 kcals 10 – 16% 3 +/- 316 kcals, absolute difference 257 +/- 197 kcals 0.57
Frankenfield et al 2004   29 non-obese subjects 52% 66% Not reported Not reported Not reported Not reported Not reported
Cheng et al 2002 Positive 46 mechanically ventilated patients Not reported Not reported Not reported Not reported Not reported 1631.3 +/- 345.3 kcals (RMR) vs 1824.8 +/- 232.9 kcals (IJ) 0.45
Flancbaum et al 1999 Positive 36 mechanically ventilated patients Not reported Not reported Not reported Average difference of 150 kcals, mean absolute difference of 386 kcals Upper limit of 368 kcals, lower limit of -67 kcals 2005 +/- 464 kcals (RMR) vs 2156 +/- 332 kcals (IJ) 0.07
Dickerson et al 2002 Positive 24 thermally injured patients Not reported Not reported Not reported –67 to 546 kcals for IJ(v), -804 to 846 kcals for IJ(s) 458 +/- 356 kcals/day (20% +/- 20%) for IJ(v), 823 +/- 598 kcals/day (30% +/- 29%) for IJ(s) Not reported Not reported
Campbell et al 2005 Neutral 42 underweight critically ill males Not reported Not reported Not reported 110.1 kcals Not reported 1790.1 +/- 365.2 kcals (RMR) vs 1900.2 +/- 207.1 kcals (IJ) 0.37
Ireton-Jones 1997 Neutral 30 burned patients Not reported Not reported Not reported Not reported Not reported nonobese: 2491 +/- 604 kcals/day (RMR) vs 2723 +/- 489 kcals/day (IJ) Not reported

Quality Rating Summary
For a summary of the Quality Rating results, click here.
Your Opinion Counts
Was the information on this page helpful to you?
Yes No
* This is a user feedback tool only. If you have specific questions requiring
a response, please use the contact link at the top of this page.
Copyright 2014 © Academy of Nutrition and Dietetics (A.N.D.), All Rights Reserved  |  Disclaimer