61
https://www.atsjournals.org/doi/pdf/10.1164/rccm.201310-
1863PP#:~:text=Lung%20Function%20(FEV1)&text=The%20American%20Thoracic%20Societ
y%2FEuropean,not%20clinically%20important%20(8)
Analytical Performance Data
Linearity
Four FeNObreath™ devices were tested as part of a linearity study. 2 devices were tested in ‘adult
mode and 2 devices were tested in ‘child mode’. Nitric oxide was mixed in a balance gas of air to
produce NO concentration levels between <5 and >500 ppb. The results are listed in the table below.
Device
Mode (child
or adult)
Slope
y-intercept
R
2
1
Adult
0.9751
1.8037
0.9993
2
Adult
1.0199
-1.8566
0.9997
3
Child
1.0092
-0.6022
0.9992
4
Child
1.0308
1.0465
0.9999
Precision/ Reproducibility and Accuracy
Precision/ Reproducibility
The repeatability and within-device precision was analysed using samples across a range of 10, 25,
75, 200, 350 & 500ppb concentrations. The measurements were performed over 5 days, 2 sessions
per day, 2 measurements per session across 15 separate devices.
Repeatability is an estimation of variation within 1 test run in 1 day. Within-device precision is an
estimate of variation between test runs and days.
FeNObreath™’s precision/reproducibility is claimed to be a standard deviation of
≤
5ppb for readings
≤
50ppb and a CV of
≤
10% for readings >50ppb. The precision/reproducibility testing performed
confirmed that the precision/reproducibility performance of the FeNObreath™ fell within these
specifications.
Limit of Detection
The limit of detection was determined based on CLSI EP17-A2. Ten devices were tested at 3 ppb (50
replicates) and 5 ppb (50 replicates) over three days in both child and adult modes. Nitric oxide
samples were mixed in a balance gas of simulated breath. The limit of detection was calculated using
the parametric option in CLSI EP17-
A2, using the following formulas: LoB=μB+1.645 σB and
LoD=LoB+1.645 σS. The results of the limit of detection analysis support the claimed detection limit
of 5 ppb. The results are presented in the below two tables.
