Long-term variability of exhaled nitric oxide measurements

Abstract

Measurement of fractional exhaled nitric oxide (FENO) is recommended as an adjunct in the diagnosis of eosinophilic airway inflammation, to determine the likelihood of corticosteroid responsiveness, to monitor therapy, and to assess adherence to inhaled corticosteroid therapy. The American Thoracic Society Guidelines for the interpretation of FENO concentrations identify cut points in children of greater than 35 parts per billion (ppb) for eosinophilic airway inflammation and less than 20 ppb for no inflammation. For adults, the guidelines recommend greater than 50 ppb and less than 25 ppb, respectively. The NIOX MINO (Aerocrine AB, Solna, Sweden) is a commonly used portable method of analyzing FENO that does not require calibration. The accuracy and precision of this device were largely determined by extensive in vitro testing for US Food and Drug Administration clearance. The same is true for its successor, the NIOX VERO. The clinical study in the MINO FDA submission was a single measurement before and after beginning inhaled corticosteroids, which demonstrated a decrease in FENO. Other published clinical studies are single visit comparisons with other methods or multiple technicians. Only one study compared 2 measurements on different days. Because there are no published reports on the long-term reproducibility of devices used to measure FENO, we undertook the following study. We analyzed 451 quality control measurements obtained over a 2-year period from 5 devices. There were 5 nonasthmatic technicians (quality control subjects) who performed these measurements on each day that the device was used for a clinical study or patient care (see Table E1 in this article’s Online Repository at www.jaci-inpractice.org for more details). All control subjects met the manufacturer’s criteria, which included >18 years old, no ongoing cold or known airway disorder, nonsmoker, expected stable exhaled nitric oxide values between 5 and 40 ppb during the qualification period, and preferably no allergies (except seasonal) or asthma. To obtain a control FENO measurement, each technician was required to empty his or her lungs by slowly exhaling and then forming a tight seal around the mouthpiece, making sure not to let the air leak out. Then, they inhaled to total lung capacity and exhaled forcefully for approximately 10 seconds through the mouthpiece while responding to sound and light cues on the LCD to guide the rate of exhalation. The general linear model procedure was used to determine whether there was a device*subject interaction. Because the interaction was significant (P .05). Measurements for subjects 3-5 who used device 5 were 31.79 (7.2), 11.51 (4.2), and 15.0 (4.1) ppb, respectively (Figure 2). Measurements for subject 3 were significantly higher than those for subjects 4 and 5 (P < .0001). Eighty-two percent of her values were above 25 ppb. The NIOX MINO portable device employing an electrochemical methodology was cleared by FDA as “substantially equivalent” to the chemiluminescence stationary method (NIOX), largely based on extensive in vitro testing using certified nitric oxide in nitrogen calibration gas. Similarly, the new NIOX VERO was cleared in the same way as “substantially