Long term performance characteristics of an electrochemical nitric oxide analyser.

Background: The NIOX MINO(®) is a nitric oxide (FE(NO)) analyser based on electrochemical technology. It includes a replaceable sensor. Quality control procedures are recommended, but regular calibration is not possible. We aimed to evaluate the performance characteristics of the NIOX MINO(®) to identify if reproducibility changed over time, or with different sensors. Also, there are reports that reproducibility of FE(NO) may be reduced in patients with high FE(NO): our secondary aim was to address this issue.
Methods: Reproducibility in 24 separate sensor-analyser units was calculated on three occasions over two months in 17 patients. These included 9 patients whose FE(NO) was high (mean 80 ppb) and 8 in whom FE(NO) was low (mean 16 ppb).
Results: One device failed quality control testing. For the remaining 23 sensor-analyser combinations, the mean coefficient of variation was 4.0% (range 1.2-7.2%) at baseline, 3.6% (range 2.0-7.0) at one month, and 3.6% (range 1.6-7.6%) at two months. The 95% C.I. for the mean limits of agreement for FE(NO) was ± 4.2 ppb (range 0.9-9.6 ppb), ± 3.8 ppb (range 1.6-6.9 ppb) and ± 3.2 ppb (range 1.2-6.8 ppb) respectively (NS). The limits of agreement exceeded the manufacturer's specifications (± 5 ppb) in 0 devices at baseline, 3 (13%) at one month, and 5 (22%) at two months.
Conclusions: Reproducibility of FE(NO) using the NIOX MINO(®) was within clinically acceptable limits (± 10 ppb) and was generally stable. However, with time, a proportion of individual sensor-analyser combinations yielded variability outside the manufacturer's specifications.
(Copyright © 2010. Published by Elsevier Ltd.)