RT Journal Article
SR Electronic
T1 Inter-instrument variability in exhaled nitric oxide measurement and implications on test interpretation
JF European Respiratory Journal
JO Eur Respir J
FD European Respiratory Society
SP P1007
VO 44
IS Suppl 58
A1 Jeffrey Pretto
A1 Brenton Eckert
A1 Jodie Simpson
A1 Heather Powell
A1 John Upham
A1 Peter Gibson
YR 2014
UL http://erj.ersjournals.com/content/44/Suppl_58/P1007.abstract
AB Differences in measurement of fractional exhaled nitric oxide (FeNO) between instruments is well documented, but the clinical implications of these differences using the 2011 ATS FeNO interpretation guidelines (AJRCCM, 2011; 184:602) has not been described. Aims: To document inter-instrument variability in FeNO measurements from 3 devices and to assess the implications of this variability on how results are interpreted in an adult clinical cohort. Methods: Comparisons between a handheld device (NIOX MINO) and 2 chemoluminescence devices (Ecomedics CLD88-comparison A, and Sievers NOA 280i-comparison B) were made. Using ATS interpretative cut-point criteria of 25ppb and 50ppb, the diagnostic performance of the MINO in identifying abnormal FeNO results was assessed using the chemoluminescence systems as standards. Results: Mean (SD) difference for comparison A in 94 adults (18M:76F) across the 5-98ppb range was small at 0.2 (8.8)ppb. Interpretation agreement between devices was 94.7% using both cut-points. MINO tended to read lower for comparison B in 73 adults (33M:40F) across the 5-131ppb range with difference of 8.8 (9.6)ppb. Interpretation agreement was 87.7%. For both comparisons there were tight relationships with simple linear modelling enabling precise correction of inter-instrument differences (A: Y= 1.16x-3.81. R =0.94, p<0.01. B: Y= 1.25x-0.733. R = 0.98, p<0.01). Conclusions: FeNO measurements from the hand-held MINO device compare favourably with more expensive chemiluminescence analysers with simple corrections enabling comparable instrument readings. Inter-instrument differences resulted in 7% misclassification rate for clinical interpretation (12 of 167 subjects).