|
Metabolic Fingerprint-MetaCheck
Accuracy
MetaCheck Accuracy
Validation Report
May 14, 2002
Joseph Orr, Ph.D.
Introduction:
The MetaCheck (Korr Medical
Technologies, Salt Lake City, Utah) is an instrument
designed to measure Resting Metabolic Rate (RMR) using
indirect calorimetry. Indirect calorimetry is a method
of calculating metabolic rate from the measured the
amount of oxygen consumed by the body. Using the
MetaCheck mouthpiece, the individual being tested
breathes in room air and the gas the person breathes
out, is conveyed to the MetaCheck through the breathing
hose. The MetaCheck analyzes the volumetric flow and
oxygen concentration of the expired gas to determine the
amount of oxygen consumed by the body due to metabolism.
The accuracy of the MetaCheck device was tested using
the nitrogen injection method as described in the
medical and physiology literature. The references listed
at the end of the report are examples of publications
where this test method is described and/or applied to
validate various indirect calorimetry systems.
Methods and Materials:
The accuracy of the MetaCheck was
analyzed using the nitrogen injection method. In this
method, a motorized piston or other device simulates
patient breathing. A precisely measured flow of pure
nitrogen (N2) is added to the gas that is pumped into
the MetaCheck. Injecting nitrogen simulates expired air,
which has a lower concentration of oxygen than fresh
air. By exactly measuring and controlling the flow of
nitrogen, the amount of oxygen consumed can be exactly
controlled and known. In these tests, breathing was
simulated using a motorized dual piston ventilator
(Model 608, Harvard Apparatus, South Natick, MA). Pure
nitrogen was obtained from a tank of compressed
nitrogen. The flow rate of the nitrogen was verified
using a precision flow standard (VT Plus, BioTek
Instruments, Winooski, VT) Since the MetaCheck reports
values in Standard Temperature and Pressure Dry (STPD)
conditions, results were multiplied by the STPD factor
as calculated by the MetaCheck. The MetaCheck barometric
pressure, ambient temperature, and relative humidity
measurements were verified using independent laboratory
instruments to ensure that these reported values were
sufficiently similar to actual ambient conditions. A
wide range of breath rates were simulated by the
motorized piston to ensure that the accuracy of the
MetaCheck was not limited to a narrow range of breathing
patterns. The simulated oxygen consumption rate relative
to the ventilation was also varied over a wide range to
ensure that accuracy was not limited to a narrow range
of oxygen concentrations.
Results:
For each measurement the RMR measured
by the MetaCheck and the simulated RMR was recorded. The
percent difference as calculated as:
%Difference=(RMRMetaCheck
- RMRSimulated) x 100%
_____________________________________________
RMRSimulated
The average of the percent difference
across all measurements was 0.84% (13.2 cal./day). The
standard deviation of the error as 1.3% (25.8
cal./day). The Plot below shows the measured value
plotted against the actual values. Regression analysis
shows the correlation between the actual and measured
values was r2 = 0.9992 and the factor relating the two
values were near a perfect 1.0 at 1.0088.
%Difference=(RMRMetaCheck
- RMRSimulated) x 100%
_____________________________________________
RMRSimulated
The plot below shows the percent
difference of the measurements vs. the minute volume.
Note that the percent difference is similarly low
regardless of the simulated minute volume (minute volume
is amount of air breathed in one minute).

The next plot shows the percent
difference plotted against the concentration of oxygen
in the expired air. Note that the error is consistently
low even at the extremes of oxygen concentration.

Inter-device Variability:
A set of five MetaCheck
systems were tested to assess the variability of the
results between systems. In this test, the N2 dilution
technique was set up using two standard oxygen
consumption and minute volume conditions. The table
below shows the measured results:

The average error across all units
was 0.6% (9 calories per day) with a standard deviation
of the error of 1.3% (24 calories per day). There is no
significant difference between measurements made with
different MetaCheck systems.
Long Term Stability
The MetaCheck was also tested to
assess the stability of the measurements over multiple
days. Simulate oxygen uptake rates were simulated at
approximately the same in 10 separate tests distributed
over a 22-day period. Simulations were done using the
nitrogen dilution technique as discussed above. The
table below lists the date of each test along with the
simulated and measured oxygen consumption values.


The average error was 0.9 ml/min
(0.3%) with a standard deviation of the error of 3
ml/min (0.9%). Over the entire period of the tests, the
worst case error was 1.4% of reading.
Discussion:
As can seen in the plots above,
measurements reported by the MetaCheck are consistently
within 2% of simulated values. This accuracy is as good
as, or better than, accuracies reported in the medical
literature when testing more complex indirect
calorimetry devices that are used in clinical settings
(see references below). The data further show that
accuracy is not limited to a narrow range of breathing
patterns or oxygen concentrations. The accuracy is also
consistent over multiple different MetaCheck systems.
Testing over multiple days showed no degradation in
performance. It appears that the MetaCheck maintains
excellent accuracy over time.
References:
M.C Damask., C. Weissman, J. Askanazi,
A.I. Hyman, S.H. Rosenbaum, and J.M.
Kinney. A systematic method of
validation of gas exchange measurements.
Anesthesiology 57:213-218,
1982
C.T. Kappagoda and R.J. Linden. A
critical assessment of an open circuit technique for
measuring oxygen consumption.,
Cardiovascular Research. 6:589-597, 1972
G. Lister Jr., J.I.E. Hoffman and
A.M. Rudolph. Measurement of oxygen consumption:
assessing the accuracy of a method.
Journal of Applied Physiology. 43:916-917, 1977.
J.A. Orr, D.R. Westenskow, A. Bauer,
A prototype gas exchange monitor for exercise
stress testing aboard NASA space
station., Journal of Applied Physiology, 66(1)
492-497,
1989
MetaCheck Clinical Accuracy Testing Using the Douglas
Bag Method
April 22, 2002
Korr Medical Technologies, Inc.
Introduction
The MetaCheck is an indirect
calorimetry system that is designed to measure resting
metabolic rate (RMR). The system operates by measuring
the volume of oxygen consumed by the patient. Since
every calorie consumed by the body requires a fixed
amount of oxygen, oxygen consumed relates directly to
calories burned. The MetaCheck uses an oxygen sensor and
a gas flow sensor to measure oxygen consumption. The
MetaCheck system auto-calibrates these sensors before
each use. The microcomputer in the MetaCheck device
integrates the flow and oxygen signals to calculate the
rate at which oxygen is consumed. The “Douglas bag” is
the “gold standard” method of validating the accuracy of
oxygen consumption measuring devices, such as the
MetaCheck. The Douglas Bag method uses a large,
non-porous bag to collect all of the gas expired by
he individual being tested. After the gas is collected,
the volume, and oxygen concentration of the gas
collected in the bag are analyzed. This analysis gives
the total volume of oxygen in the bag. Based on the bag
contents and amount of time over which the bag was
filled, the rate at which oxygen was consumed can be
calculated.
Methods
The MetaCheck system was calibrated
using the standard automatic calibration before each
test. Following auto-calibration, subjects breathed
through a standard MetaBreather disposable airway
adaptor connected to the MetaCheck system. Breathing was
allowed to stabilize for at least 1 minute before data
collection began. After stabilization, expired gas
exiting the MetaCheck was collected in a 100 Liter
Douglas Bag (Hans Rudolph P/N 112377, Hans Rudolph inc,
Kansas City, MO). Oxygen consumption (VO2) for each
breath along with the breath rate measured by the
MetaCheck were stored digitally for each breath during
the test. Expired gas was collected for at least 2
minutes and at least 20 Liters of gas was collected for
each test. After the gas was collected, the bag was
sealed. Average oxygen consumption for all of the
breaths measured by the MetaCheck during data collection
was calculated. The total time of data collection was
recorded as well. The volume and ontents of the
Douglas bag were analyzed following each individual's
data collection. The volume was measured by drawing gas
out of the bag by a vacuum pump at a fixed rate of 8
liters per minute through a precision flow measurement
system (P/N RSS100-HR with Neonatal Fixed-orifice Flow
Sensor P/N 101110, Hans Rudolph inc, Kansas City, MO).
The gas was drawn out of the bag until a vacuum of –3 cm
H2O was observed indicating that the bag had been
completely emptied. The flow measurement system was set
to measure expired air at room temperature and ambient
barometric pressure. The average oxygen concentration of
the air in the bag was measured using a gas flow
analyzer (VT Plus, BioTek Instruments, Winooski, VT).
The average CO2 concentration was measured using a CO2
analyzer (model 8200 Cosmo+, Novametrix Medical Systems,
Wallingford, CT). The volume of oxygen inspired by the
subject was calculated using the measured ambient
relative humidity and temperature. Further compensation
was made to account for the difference in the rate of
oxygen consumption to carbon dioxide production. The
oxygen consumed during the test is the difference
between oxygen consumed by the subject and the volume of
oxygen that was collected in the bag. The total oxygen
consumed divided by the collection time gives the rate
or oxygen consumption measured by the Douglas Bag
method. This rate of oxygen consumption can then be
compared to the average oxygen consumption rate measured
simultaneously by the MetaCheck. Oxygen consumption
rates were converted to metabolic rates for purposes of
presenting the data.
Results
A total of 13 comparisons were made
using 8 subjects. Tests were repeated in some subjects
at various levels of physical activity to produce a
wider range of test conditions. Measured metabolic rates
ranged from 959 to 3795 calories per day. The average
difference between the MetaCheck and the Douglas bag
method was –3.22% (-53 calories/day). The standard
deviation of the error was 3.4% (46 calories/day). The
data plots below show the relationship between Oxygen
consumption and metabolic rates tested using the Douglas
bag and the MetaCheck. The line relating MetaCheck
oxygen consumption measurements to the corresponding
Douglas bag values has a slope of 1.0081 with an offset
of –10 ml/min. The correlation coefficient between the
two methods was R2 = 0.9969.


Discussion
This data shows very good agreement
between the MetaCheck and the Douglas Bag method. The
number of data points is small and further testing will
be done in the future. However, the accuracy of this
data and the range over which it was gathered, indicates
that further testing should only reinforce these
results. Linear regression shows a near perfect
correlation between MetaCheck and Douglas bag
measurements of R2 = 0.9969 and a near perfect slope
factor of 1.0081. The accuracy of the MetaCheck system
can be partially attributed the accuracy of the sensors.
At the average expired oxygen level of 15.65% Oxygen,
the “worst case” error attributable to oxygen sensor
error would be 3.64% error. Since the oxygen sensor in
the MetaCheck is specified to be better than 0.2%, the
errors we observed may be attributable to oxygen sensor
error. Competing devices claim an oxygen sensor accuracy
of ±0.8%.
Similar analysis of errors attributable to this level of
oxygen sensor accuracy shows expected errors of 14.6%
error in competing products. Another source of error is
gas flow measurement. Since the MetaCheck only measures
flow in one direction, errors in flow measurement do not
translate into significant errors. For example, in the
MetaCheck, a 2% error in flow measurement can cause an
error in oxygen consumption of no greater than 2%. In
some competing devices, both inspired and expired gas
flow is measured. In this method, even a 1% error in
flow measurement may cause errors as high as 6.7% for
the competing devices.
|