The accuracy of nasal nitric oxide testing in PCD diagnostics is population specific

Background: Nasal nitric oxide (nNO) is usually low in primary ciliary dyskinesia (PCD) and measurement is part of a complex diagnostic work-up. Previous research has failed to define the accuracy of nNO testing in differing patient populations. Aims: Use accuracy data of nNO testing in patients referred to a UK national diagnostic service to calculate the predictive value of nNO testing in children with persistent wet cough (estimated 5% PCD prevalence) and in the general population (1:10,000 prevalence). Methods: From June 2006 to January 2014, 282 consecutive referrals (aged 6-79 years) had nNO measured by breath hold technique on the chemiluminescent NIOX Flex analyser. PCD was diagnosed by high resolution video microscopy and quantitative electron microscopy. Sensitivity and specificity was calculated and used to derive positive and negative predictive values (PPV & NPV) for each population. Results: 31 patients were PCD positive. Using a published cut-off of 77 nl/min (256 parts per billion on NIOX Flex), sensitivity was 93.6% (95% CI 78.5-99.0) and specificity 84.1% (95% CI 78.9-88.4). PPV was 42% (30.2-54.5%) and NPV 99.1% (96.6-99.9%) for referrals (prevalence of PCD 11%). PPV was 23.6% for persistent wet cough and Conclusion: PPV is poor and diminishes rapidly with lower PCD prevalence therefore screening should be targeted at those with high clinical suspicion of PCD. Whilst NPV is excellent, it must be interpreted alongside clinical history given published cases of PCD with normal nNO. These findings highlight the importance of using clinical features to select those who undergo nNO testing.