Determining the Accuracy and Reliability of Indirect Calorimeters Utilizing the Methanol Combustion Technique

Objective There are several indirect calorimetry (IC) instruments commercially available but validity and reliability data is lacking. Site‐to‐site inconsistencies in protocols and subject characteristics, and comparisons to a “gold standard” instrument or method which may no longer be accurate enough, have put restraints on drawing conclusions about instruments' performance. Purpose To compare the accuracy and reliability of different metabolic carts using the methanol combustion technique as the criterion measure. Methods A total of eight, 20‐minute methanol burn trials were completed on 12 metabolic carts (2 Cosmed Quark CPET, 2 DeltaTrac II, 2 Parvo Medics TrueOne 2400, 2 Iaeger Oxycon Pros, 2 Omnicals, a Vmax Encore, and a Max‐II Metabolic System) at 11 international study sites. Methanol tests were performed at 0700, 1000, 1300, and 1600 hours on 2 consecutive days. Respiratory Exchange Ratio (RER) and percent (%) recovery of O 2 and CO 2 were calculated after each test. Results Accuracy – 1 Omnical, Cosmed, and Parvo were accurate in measuring RER and % recovery O 2 , while 1 DeltaTrac was also accurate for % recovery O 2 . The same Cosmed and Parvo, and the other DeltaTrac were accurate in measuring % recovery CO 2 . Reliability – 8 instruments were shown to be reliable with the two Omnicals ranking best based on the smallest coefficient of variation (CV) (all CV(s) 1.26%). Both Cosmeds, Parvos, DeltaTracs and 1 Vmax were the reliable instruments for at least one variable (CV(s) 3%). Multiple Regression– Humidity, amount of methanol combusted, and temperature were tested as predictors of IC outcomes. Humidity and amount of combusted methanol were significant predictors of RER (F (2, 60) =10.91, p<0.001, R 2 =0.33). Temperature and amount of burned methanol were significant predictors of % recovery O 2 (F (2, 60) =8.32, p<0.001, R 2 =0.18) while only humidity was a predictor for % recovery CO 2 (F (1, 61) =21.10, p<0.001, R 2 =0.15). Conclusion Omnical, Cosmed, and Parvo showed superior accuracy and reliability; however, accuracy was only found at one of two study sites. Exogenous factors such as humidity and temperature may be influencing instrument performance and could be modified in the lab to optimize IC conditions. Support or Funding Information There were no sources of funding for this manuscript, nor were there any sponsors. 1 Instument AccuracyInstruments RER Instruments O 2 Recovery (%) Instruments CO 2 Recovery (%)X̄ ±SD %RE Rank X̄ ±SD %RE Rank X̄ ±SD %RE RankOmnical2 0.667±0.002 0.001 * 1 Omnical 1 100.1±0.8 0.094 * 1 Parvo2 100.1±3.2 0.144 * 1 Cosmed2 0.673±0.005 0.927 * 2 Cosmed2 99.8±2.1 −0.250 * 2 Cosmed2 100.7±1.7 0.666 * 2 Parvo2 0.660±0.005 −1.034 * ‡ 3 Delta Trac2 100.8±2.4 0.794 * 3 Delta Trac1 100.7±1.5 0.702 * 3 Omnical1 0.677±0.003 1.546 4 Parvo2 101.2±2.9 1.185 * 4 Vmax Encore 98.4±2.1 −1.615 4 Parvol 0.678±0.006 1.685 5 Vmax Encore 101.8±5.1 1.804 5 Omnical1 101.6±0.8 1.640 5 Max‐II 0.652±0.022 −2.226 6 Parvol 102.0±13.9 2.026 6 Delta Trac2 98.1±3.8 −1.874 6 Delta Trac2 0.649±0.020 −2.649 7 Oxycon2 97.9±7.2 −2.051 7 Omnical2 103.6±1.3 3.647 7 Oxycon1 0.649±0.007 −2.705 8 Cosmed1 102.5±2.2 2.469 8 Parvo1 104.0±14.4 3.976 8 Vmax Encore 0.645±0.023 −3.221 9 Delta Trac1 97.4±1.7 −2.619 9 Cosmed1 95.7±1.2 −4.293 9 Delta Trac1 0.689±0.008 3.420 10 Omnical2 103.6±1.3 3.647 10 Oxycon2 104.4±6.7 4.368 10 Cosmed1 0.626±0.014 −6.089 11 Max‐II 107.0±3.9 6.970 11 Max‐II 104.5±2.2 4.493 11 Oxycon2 0.712±0.040 6.755 12 Oxycon1 91.4±3.1 −8.615 12 Oxycon1 88.9±2.6 −11.103 12X̄ Average of measured values by the instruments throughout the 8 burns, SD: Standard Deviation, %RE: Percent of Relative Error, RER: Respiratory Exchange Ratio. ‡ Negative %RE values on this table indicate that the calculated averages for the corresponding instruments were less than the theoretical value. Instruments rankings (best to worst) are based on % RE   ( % RE =( A v e r a g e   o f   m e a s u r e d   v a l u e s ‐ T h e o r e t i c a l   v a l u e ) T h e o r e t i c a l   V a l u e× 100 ) ; which explains the difference of the average values measured by the instruments from theoretical. Theoretical values are 0.667 (for RER), and 100% (1.00 for O 2 and CO 2 recoveries). * denotes accurate instrument based on a %RE≤1.5%.2 Instument ReliabilityInstruments RER Instruments % Recovery O 2 Instruments % Recovery CO 2CV (%) Rank CV (%) Rank CV (%) RankOmnical2 0.23 * 1 Omnical1 0.75 * 1 Omnical1 0.76 * 1 Omnical1 0.49 * 2 Omnical2 1.26 * 2 Omnical2 1.21 * 2 Cosmed2 0.68 * 3 Delta Trac1 1.76 * 3 Cosmed1 1.22 * 3 Parvo2 0.75 * 4 Cosmed2 2.10 * 4 Delta Trac1 1.47 * 4 Parvo1 0.85 * 5 Cosmed1 2.18 * 5 Cosmcd2 1.65 * 5 Oxycon1 1.05 * 6 Delta Trac2 2.35 * 6 Vmax Encore 2.09 * 6 Delta Trac1 1.13 * 7 Parvo2 2.89 * 7 Max‐II 2.14 * 7 Cosmed1 2.28 * 8 Oxycon1 3.42 8 Oxycon1 2.97 * 8 DeltaTrac2 3.01 9 Max‐II 3.69 9 Parvo2 3.18 9 Max‐II 3.42 10 Vmax Encore 4.97 10 Delta Trac2 3.84 10 Vmax Encore 3.50 11 Oxycon2 7.31 11 Oxycon2 6.46 11 Oxycon2 5.49 12 Parvo1 13.46 12 Parvo1 13.86 12CV: Coefficient of Variation, RER: Respiratory Exchange Ratio. Instruments rankings (best to worst) are based on CV which was used to assess reliability of measurements by each instruments. * denotes reliable instrument based on a CV ≤ 3%.