33
Technical information
Device Performance
SpO
2
Accuracy
SpO
2
accuracy testing is performed by
in vivo
accuracy testing under
laboratory conditions on healthy adult subjects with varying skin
pigmentation in an independent research laboratory through induced
hypoxia studies. Analysis of bias
1
was performed vs. Hemoximeter data.
The limits of agreement shown are calculated per:
Bland JM, Altman D.
(2007) Agreement between methods of measurement with multiple
observations per individual. Journal of Biopharmaceutical Statistics 17,
571
– 582.
Root mean square error (RMS error) is calculated as follows:
RMS Error =
√
∑(SpO
2
−SaO
2
)
2
n
1
Bias is defined as the monitor under test reading minus the hemoximeter
reading.
Note: Because pulse oximeter equipment measurements are statistically
distributed, only about two-thirds of Aulisa GA1001 Digital Vital Sign
Monitoring System measurements can be expected to fall within
±A
rms
of the value measured by a co-oximeter.
A
rms
from the clinical study
Sensor Accuracy
(A
rms
)
90%-100%
1.82
80%-90%
2.66
70%-80%
3.19
The graph below shows the error (SpO
2
– SaO
2
) plots of each subject
measured by the Aulisa GA1001 System with upper and lower 95%
limits of agreement. Each sample data point is from a clinical study in
healthy adult volunteers.
