Converting FENO by different flows to standard flow FENO
Full text
Converting F ENO by different flows to standard flow F ENO Paul G. Lassmann-Klee 1 , Lauri Lehtim€ aki 2 , Tuula Lindholm 3 , Leo Pekka Malmberg 4 , Anssi R.A. Sovij€ arvi 1 and P€ aivi Liisa Piiril€ a 1 1 Unit of Clinical Physiology, Helsinki University Central Hospital and University of Helsinki, Helsinki, Finland, 2 Allergy Centre, Tampere University Hospital, Faculty of Medicine and Health Technology, University of Tampere, Tampere, Finland, 3 Department of Clinical Physiology, Finnish Institute of Occupational Health,Helsinki, Finland, and 4 Laboratory of Clinical Physiology, Skin and Allergy Hospital, Helsinki University Hospital, Helsinki, Finland Abstract Correspondence Paul G. Lassmann-Klee, Unit of Clinical Physiology, Helsinki University Central Hospital, PO Box 340, 00029 HUS/Helsinki, Finland. E-mail: [email protected] Accepted for publication Received 5 June 2018; accepted 30 April 2019 Key words adults; alveolitis; asthma; children; COPD; fractional exhaled nitric oxide (F ENO) ; mouthwash; multiple-flow In clinical practice, assessment of expiratory nitric oxide (F ENO ) may reveal eosinophilic airway inflammation in asthmatic and other pulmonary diseases. Currently, measuring of F ENO is standardized to exhaled flow level of 50 ml s 1 , since the expiratory flow rate affects the F ENO results. To enable the comparison of F ENO measured with different expiratory flows, we firstly aimed to establish a conversion model to estimate F ENO at the standard flow level, and secondly, validate it in five external populations. F ENO measurements were obtained from 30 volunteers (mixed adult population) at the following multiple expiratory flow rates: 50, 30, 100 and 300 ml s 1 , after different mouthwash settings, and a conversion model was developed. We tested the conversion model in five populations: healthy adults, healthy children, and patients with COPD, asthma and alveolitis. F ENO conversions in the mixed adult population, in healthy adults and in children, showed the lowest deviation between estimated ^ FENO from 100 ml s 1 and measured F ENO at 50 mL s 1 :028 ppb, 044 ppb and 027 ppb, respectively. In patients with COPD, asthma and alveolitis, the deviation was 116 ppb, 168 ppb and 147 ppb, respectively. We proposed a valid model to convert F ENO in healthy or mixed populations, as well as in subjects with obstructive pulmonary diseases and found it suitable for converting F ENO measured with different expiratory flows to the standard flow in large epidemiological data, but not on individual level. In conclusion, a model to convert F ENO from different flows to the standard flow was established and validated. Introduction Chronic bronchial inflammation of the respiratory mucosa can lead to bronchial hyperreactivity and airway obstruction. Clinicians often employ fractional exhaled nitric oxide (F ENO )to evaluate bronchial eosinophilic inflammation (NICE, 2017). F ENO values are flow-dependent, and an expiratory flow rate of 50 ml s 1 mirrors the bronchial nitric oxide (NO) production and not the NO with peripheral origin (Tsoukias & George, 1998; H€ogman et al., 2000). For this reason, F ENO measurement is currently standardized at the expiratory flow rate of 50 ml s 1 (ATS/ERS, 2005, Horvath et al., 2017). Prior to the standardization, F ENO was acquired in Northern Europe with expiratory flow rates of 50-300 ml s 1 (H€ogman et al., 1997; Ekroos et al., 2002; Rouhos et al., 2008) and a previous guideline endorsed the use of flow rates between 167 and 250 ml s 1 (Kharitonov et al., 1997). Many pioneers in F ENO investigation adopted a flow rate of 100 ml s 1 (Kharitonov & Barnes, 2001). Unfortunately, data measured at different flow levels have been difficult to compare, since F ENO values are affected by the flow rate used and represent NO from anatomically different lung parts. Therefore, a conversion method to interpolate F ENO values to equivalent F ENO values at diverse flows was needed. Since the lowering effect of mouthwashes on F ENO values is well documented (Lassmann-Klee et al., 2018a,b), the conversion method should address also the mouthwashes. The aim of this study was to establish a method for converting F ENO , measured at different expiratory flow levels, to the standard F ENO measured at 50 ml s 1 and validate this method. Further on, we aimed to determine the need of considering the mouthwashes in the conversion method. Glossary F ENO , Fractional exhaled nitric oxide ^ FENO, Estimated fractional exhaled nitric oxide _ V, Expiratory flow rate NO, Nitric Oxide Clin Physiol Funct Imaging (2019) 39, pp315–321 doi: 10.1111/cpf.12574 315©2019 The Authors. Clinical Physiology and Functional Imaging published by John Wiley & Sons Ltd on behalf of Scandinavian Society of Clinical Physiology and Nuclear Medicine 39, 5, 315–321 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
Methods Data acquisition We recruited 30 healthy or asthmatic adults as volunteers (henceforth referred as ‘mixed adult population’) to develop a conversion method. We have previously described this population (Lassmann-Klee et al., 2018b). The volunteers were adult patients (n=9) or healthcare workers (n=21). The patients invited were previously referred for F ENO assessment to the Laboratory of Clinical Physiology or to the Skin and Allergy Hospital at the Helsinki University Central Hospital area. The healthcare employees were included in the study without exclusions. The patients enrolled had respiratory symptoms or a chronic respiratory disease, including asthma (n=4), eosinophilic bronchitis (n=1), building-related respiratory symptoms (n=3) and Sj€ogren’s syndrome (n=1). Spirometric data (n=25) were analysed, and none of the participants had actual bronchodilator reversibility (Pellegrino et al., 2005). F ENO measurements were performed at the Finnish Institute of Occupational Health and at the Skin and Allergy Hospital with CLD 88 sp chemiluminescence NO analysers and EXHALIZERâ’s D devices using SPIROWARE â software (Eco Medics AG, Switzerland). The devices were calibrated in compliance with the producer’s specifications: use of certified span gas (AGA Gas BV, Amsterdam, Netherlands) and a zero-air filtering system (DENOX 88 unit). Additionally, a calibration syringe (Hans Rudolph Inc., USA) was used to calibrate the ultrasonic flow sensor. We complied with all advices from the ATS/ERS statement (ATS/ERS, 2005). We performed F ENO measurements in our mixed adult population (n=30) from September 2016 until May 2017, and the tests for each volunteer were scheduled on 2 consecutive days. All the 30 volunteers followed a mouthwash protocol with tap water and carbonated water. Detailed description of the mouthwashes’ protocol is available in our recent study (Lassmann-Klee et al., 2018b). Briefly, the F ENO measurements were performed after a mouthwash with 100 ml of tap water at each flow level. After 15 min, all measurements were repeated after a mouthwash with 100 ml of carbonated water at each flow level. The mouthwashes’ effect, duration and chemical composition are well documented (Lassmann-Klee et al., 2018a,b). Secondly, we selected 10 healthcare workers from the aforementioned volunteers to perform an additional measurement phase. The selection criterion was inclusion only of those employed at the Skin and Allergy Hospital. In the third appointments, the 10 healthcare workers performed the measurements without a mouthwash. F ENO was acquired from all participants at the following multiple expiratory flow rates: 50, 30, 100 and 300 ml s 1 . At least two measurements of F ENO were obtained at each flow level. The values were accepted, if its variation was less than 2 ppb. Validation For validating our conversion method, 5 different datasets of previously published articles acquired at the Tampere University Hospital were available. They contained multiple-flow data from 69 healthy adults (Lehtim€aki et al., 2010a,b), 66 healthy children (Sepponen et al., 2008), 74 steroid-naive adults with COPD (Lehtim€aki et al., 2010a), 40 steroid-naive adults with asthma (Lehtim€aki et al., 2001) and 17 subjects with untreated alveolitis (Lehtim€aki et al., 2001). The validation process is explained in the statistical section. This study followed the ethical principles of the declaration of Helsinki (World Medical Association Declaration of Helsinki, 2013) and received approval from an ethical committee (99/13/03/00/15). All participants signed an informed consent. Statistics Modelling the conversion method Analyses were performed using RSTUDIOâversion 11383 frontend to the R statistics language (R Core Team, 2018). We agreed on a significance level of a=005 as significant. We calculated the arithmetic mean from individual F ENO values obtained at each flow level. The mean values were plotted against the expiratory flow rate _ Vin a double logarithmic scale, and we performed a non-linear regression. We obtained a slope and intercept and analysed the regression line to develop our conversion model. To further refine the model, we acquired a non-linear least squares estimation of the nonlinear model parameters. This model was used to estimate ^ FENO values from F ENO values measured at different flow rates. Validation To test the validity of our model, we converted F ENO values measured at 30, 100 and 300 ml s 1 to estimated ^ FENO values for a standard flow rate of 50 ml s 1 . Afterwards, we compared the estimated ^ FENO values to the actual F ENO measured at 50 ml s 1 . To assess the agreement between estimated ^ FENO and measured F ENO , we performed an analysis (see below) according to Bland & Altman (2010). Further on, the correlation coefficient rho was obtained with Spearman’s formula to investigate linearity. To validate our conversion model in different external populations, we compared the estimated ^ FENO converted from 100 ml s 1 with F ENO measured at 50 or 40 ml s 1 . For this external validation, a method described by Bland & Altman (2010) was employed. Accordingly, we obtained the individual differences of F ENO , the mean of differences (bias) and the 196 standard deviations of the mean (95% limits of agreement). Additionally, we performed a linear regression analysis (glm) between F ENO values measured at 50 ml s 1 after the tap water and carbonated water mouthwashes, to obtain a ©2019 The Authors. Clinical Physiology and Functional Imaging published by John Wiley & Sons Ltd on behalf of Scandinavian Society of Clinical Physiology and Nuclear Medicine 39, 5, 315–321 FENO conversion model, P. G. Lassmann-Klee et al.316
relation between the mouthwashes and to provide an additional equation to convert measurements with these two mouthwashes to the standard flow level (50 ml s 1 ). When necessary, raw data were examined for outliers using the absolute deviation around the median (3 deviations as threshold). If cases were omitted, the conversion was repeated and the differences and level of agreements adjusted (Leys et al., 2013). Results Conversion model We plotted the mean F ENO values against the expiratory flow rate _ Vand performed a non-linear regression. Acquiring nonlinear least squares parameter estimates resulted in a slope of 08416 SE(03192) for carbonated water, a slope of 084 SE(02989) for tap water and a slope of 083111 SE (005424) in the absence of a mouthwash. In the latter case, the equation model can be further defined as: ^ FENO ¼k_ V083111 ð1Þ Plotting our model with Eq. using measured F ENO and _ V, as well as calculated values for k, resulted in Fig. 1. The linear regression of F ENO at 50 ml s 1 after a tap water mouthwash in relation to carbonated water resulted in a slope coefficient of 1055 ppb and intercept of 0354 ppb (P<0001). When employing the different estimating slopes for the ^ FENO conversions with tap water and carbonated water mouthwashes, the mean estimated ^ FENO for the carbonated water mouthwash was ca. 45% lower than the mean estimated ^ FENO for tap water at all flow levels (unadjusted). Validation results in mixed adult population Using Eq. 1, we calculated the values for ^ FENO (flow level 50 ml s 1 ) interpolated from data obtained at 100 ml s 1 . Applying the (Bland & Altman, 2010) method resulted in mean (SD) differences between the estimated ^ FENO (flow level 50 ml s 1 ) and the measured F ENO (flow level 50 ml s 1 )of -045(244) ppb, upper 95% limit of agreement of 434 ppb and lower 95% limit of agreement of 523 ppb. The measured F ENO and the estimated ^ FENO had a good correlation (Spearman’s q=087; P<00001). We also estimated ^ FENO (50 ml s 1 ) from values measured at all flow levels and mouthwash settings. All differences with the (Bland & Altman, 2010) method showed a good agreement, and the total unadjusted mean of the absolute deviation of ^ FENO from F ENO was 072 ppb. All estimated values were highly correlated with corresponding measured values. Table 1 summarizes these results. Figure 2 exemplifies the unadjusted mean differences of ^ FENO and F ENO after applying Eq. 1 (conversion with carbonated water mouthwash from flow of 100 ml s 1 ). After adjusting measured F ENO by removing outliers and performing a new estimation, a better agreement was found between estimated ^ FENO and measured F ENO , and total mean of the absolute deviations of ^ FENO from F ENO was 066 ppb. The adjusted results after controlling for outliers can be also found in Table 1. Validation results in external populations With the same approach, we converted F ENO data obtained at 100 ml s 1 (Lauri Lehtim€aki et al., 2001; Sepponen et al., 2008; Lehtim€aki et al., 2010a,b) to estimated ^ FENO (flow level 50 or 40 ml s 1 ) without a mouthwash (Eq. ). The mean difference between estimated ^ FENO and measured F ENO was lowest (027 ppb) in the healthy children group, followed by the healthy adult group (044 ppb), as shown in Fig. 3. The mean difference illustrated in Fig. 2 of steroid-naive adults with asthma was 168 ppb. In Fig. 4, the mean difference shown is 116 ppb in steroid-naive adults with COPD, and 147 in the untreated alveolitis population. The healthy groups had narrow limits of agreement, in contrast to the groups with diseases. Table 2 synthesizes these results. Additionally, Fig. 5 demonstrates the distribution of the differences in all populations. Table 3 contains the correlation between the measured and estimated F ENO values and provides information concerning the linearity between the values. Discussion Conversion model We found that using a non-linear regression yielded a simple model to convert F ENO values measured at different flows to estimated ^ FENO at 50 ml s 1 . To prove the feasibility of the equation, we compared estimated ^ FENO levels at the standard 10 20 30 40 100 200 300 Flow (mL/s) FENO (ppb) Figure 1 F ENO as a function of expiratory flow (without mouthwash), n= 10. Curve shows the equation^ FENO ¼k_ V083111. ©2019 The Authors. Clinical Physiology and Functional Imaging published by John Wiley & Sons Ltd on behalf of Scandinavian Society of Clinical Physiology and Nuclear Medicine 39, 5, 315–321 FENO conversion model, P. G. Lassmann-Klee et al.317
flow (50 ml s 1 ) from all flow levels (30, 100 and 300 ml s 1 ), with F ENO acquired at 50 ml s 1 and found a good mean agreement between the estimated and measured values. The limits of agreement between estimated ^ FENO and F ENO were reasonable. Validation Assessment of the conversion in external datasets, including data of a wide range of pulmonary diseases and multiple-flow F ENO values, confirmed these previous findings. The conversion model developed showed the lowest deviation in F ENO conversions in healthy children, healthy adults and in our 0 5 20 40 60 Mean of estimated F ^ENO and measured FENO (ppb) Difference between F ^ENO and FENO (ppb) Figure 3 Bland–Altman plot with mean of F ENO measured at 50 ml s 1 and estimated ^ FENO from 100 ml s 1 in healthy children (grey dots, n= 66) and in healthy adults (black dots, n= 69), plotted against the differences in F ENO . In healthy children: mean differences (grey dotted line), 196 standard deviations (grey dot-slashed line). In healthy adults: mean differences (black solid line), 196 standard deviation (black slashed line). Table 1 Bland–Altman statistics in our mixed healthy and asthmatic adult population (n= 30) and in healthcare workers (n= 10) with mean, bias a , levels of agreement and standard deviation (SD) of the differences between estimated ^ FENO from different flow levels and mouthwashes, and measured F ENO at 50 ml s 1 (tap water: 2727 ppb; carbonated water: 2551 ppb; no mouthwash: 2205) Mean estimated ^ FENO (ppb) at 50 ml s 1 from flow level and mouthwash Bias a Adjusted values Level of agreement Level of agreement Lower Upper SD bias a Lower Upper SD rho b 30 ml s 1 ; tap 2524 203 1117 710 466 123 544 30215 096 3 100 ml s 1 ; tap 2699 028 742 686 364 011 367 344 181 098 3 300 ml s 1 ; tap 2627 100 1902 1701 919 074 579 727 333 095 2 30 ml s 1 ; carbonated 2423 128 492 236 186 150 490 190 173 099 3 100 ml s 1 ; carbonated 2565 013 428 455 225 008 332 316 165 099 4 300 ml s 1 ; carbonated 2507 044 1332 1243 657 099 469 667 290 095 4 30 ml s 1 ; no mouthwash 2164 041 589 506 279 041 589 506 279 084 0 100 ml s 1 ; no mouthwash 2160 045 523 434 244 045 523 434 244 087 0 300 ml s 1 ; no mouthwash 2162 043 567 482 268 043 567 482 268 082 0 Raw data and adjusted values for outliers. Rho according to Spearman’s test. a average of the differences between estimated ^ FENO and measured F ENO . b Number of observations excluded with the adjustment. –10 0 10 0 50 100 150 Mean of estimated F ^ENO and measured FENO (ppb) Difference between F ^ENO and FENO (ppb) Figure 2 Bland–Altman plot with mean of measured F ENO and estimated ^ FENO from 100 ml s 1 in asthmatics (grey dots, n= 40) and our mixed adult population (black dots, n= 30), plotted against the differences in F ENO . In asthmatics: mean differences (grey dotted line), 196 standard deviations (grey dot-slashed line). In mixed adult population: mean differences (black solid line), 196 standard deviation (black slashed line). In asthmatics F ENO measured at 40 ml s 1 .In mixed adult population F ENO measured at 50 ml s 1 after carbonated water mouthwash. ©2019 The Authors. Clinical Physiology and Functional Imaging published by John Wiley & Sons Ltd on behalf of Scandinavian Society of Clinical Physiology and Nuclear Medicine 39, 5, 315–321 FENO conversion model, P. G. Lassmann-Klee et al.318
mixed asthmatic and healthy adult population. In the steroidnaive asthmatic, alveolitis and COPD populations, the average differences in F ENO were moderate with moderate limits of agreement. In the population with COPD, some single individuals showed a considerable deviation. We acknowledge the limitation of this conversion procedure, that is being only an approximation that may result in a considerable deviation between estimated and physiological values especially at extreme F ENO and/or flow levels, as observed in conversions from low flow (30 ml s 1 ) or high expiratory flow (300 ml s 1 ) levels. Nevertheless, this equation is useful when comparing the F ENO medians of large population data measured at different flow levels, being very reliable on the group level, although not on individual level. The conversion model developed suits best F ENO conversions in healthy adults, healthy children and in a mixed adult population, showing the lowest deviation. This novel conversion model mimics physiological expiratory NO values proportional to expiratory flows. Similar F ENO and expiratory flow curves were previously described by other researchers –30 –20 –10 0 10 0 50 100 150 Mean of estimated F ^ENO and measured FENO (ppb) Difference between F ^ENO and FENO (ppb) Figure 4 Bland–Altman plot with mean of measured F ENO and estimated ^ FENO from 100 ml s 1 in COPD patients (grey dots, n= 72) and patients with alveolitis (black dots, n= 17), plotted against the differences in F ENO . In COPD patients: mean differences (grey dotted line), 196 standard deviations (grey dot-slashed line). In patients with alveolitis: mean differences (black solid line), 196 standard deviation (black slashed line). In patients with alveolitis F ENO measured at 40 ml s 1 . In COPD patients F ENO measured at 50 ml s 1 . Table 2 Bland–Altman statistics with bias a , levels of agreement and standard deviation (SD) of the differences between estimated ^ FENO from 100 ml s 1 (Eq. 1) and measured F ENO at 50 or 40 ml s 1 Population Bias a Level of agreement SDLower Upper Mixed healthy and asthmatic adults 028 742 686 364 Healthy adults 044 387 298 174 Asthmatic 168 1136 799 494 Healthy children 027 194 248 113 COPD 116 1146 913 525 Alveolitis 147 828 1122 498 a average of the differences between estimated ^ FENO and measured F ENO§. 0·0 0·1 0·2 0·3 0·4 –30 –20 –10 0 10 Difference between estimated F ^ENO and measured FENO (ppb) Density Dataset Healthy adults Asthmatic adults Healthy children COPD Alveolitis Mixed adults Figure 5 Density plot with mean differences between F ENO measured at 50 or 40 ml s 1 and estimated ^ FENO from 100 ml s 1 , and the density of the individual mean differences in all study groups. [Colour figure can be viewed at wileyonlinelibrary.com] Table 3 Spearman’s correlation between estimated ^ FENO from 100 ml s 1 and measured F ENO at 50 ml s 1 , with 95% CI and Pvalues Population Correlation 95% CI PLower Upper Mixed healthy and asthmatic adults 099 098 099 <0001 Healthy adults 097 095 098 <0001 Asthmatic 099 098 099 <0001 Healthy children 097 095 098 <0001 COPD 098 096 098 <0001 Alveolitis 087 068 095 <0001 ©2019 The Authors. Clinical Physiology and Functional Imaging published by John Wiley & Sons Ltd on behalf of Scandinavian Society of Clinical Physiology and Nuclear Medicine 39, 5, 315–321 FENO conversion model, P. G. Lassmann-Klee et al.319
(Tsoukias & George, 1998; Silkoff et al., 2000), but this model uses a simplified approach in estimating ^ FENO and makes no claim in predicting flow-independent parameters. Since the conversion model developed derives from healthy and asthmatic adults without alveolar diseases, the slope reflects only very low amounts of alveolar nitric oxide concentration (C ANO ). We previously determined C ANO in our mixed healthy and asthmatic group and all results were under 23 ppb (Lassmann-Klee et al., 2018b). Logically, the slope and the estimating equation would change, if switching the participants with subjects with high alveolar NO. The conversion method produces errors in those subjects in whom the relation between alveolar and bronchial NO production is very different from the group mean, as the slope between F ENO and _ Vis very different in these subjects. Therefore, the model may result in erroneous estimates when applied to subjects with known high alveolar nitric oxide concentrations. Emphasis should be made, not to employ the model without discretion in this type of subjects. The elimination of outliers could represent a limitation of our study, although we did not observe drastic changes when comparing the bias between crude and adjusted data. This statistical adjustment merely narrowed the limits of agreement and served the purpose of demonstrating how the model estimates F ENO values stemming from adjusted datasets. Further on, regression estimates were obtained for F ENO values between the mouthwashes, in order to facilitate an interpolation between F ENO values measured at 50 ml s 1 after carbonated, and tap water, and vice versa. Our estimating equation provides different slopes for both mouthwashes. The mean estimated ^ FENO values were ca. 4% lower for the carbonated water mouthwash than the tap water mouthwash. This approximate difference between these mouthwashes was previously confirmed (Lassmann-Klee et al., 2018a,b). The conversion model succeeds also in considering the mouthwashes. In conclusion, we developed an equation for converting F ENO values obtained with different flow levels to F ENO with standard flow (50 ml s 1 ), taking also into account the eventual mouthwash. We proposed a novel model to convert F ENO in healthy populations, as well in subjects with obstructive pulmonary diseases. We conclude that the model is reliable in converting F ENO in large epidemiological data and might be applied in small scale populations with pulmonary diseases, but not on individual level. Acknowledgements We thank the staff members: Sari Fischer, Helena Punkari and Elina Voutilainen for performing the F ENO measuring and also Tommi Pallasaho for data collection. Funding This work was supported by the Nordic Council of Ministers, NordForsk Institution (The Nordic EpilLung Study), the Nummela Sanatorium Foundation (PP 2015, 2017), (AS 2016), Finnish State Funding for University-level Health Research (TYH: 2013354), The Research Foundation of the Pulmonary Diseases (PLK 2017, 2018, 2019), Tampere Tuberculosis Foundation: Eero H€am€al€ainen (PLK 2017, 2018), Ida Montin Foundation (PLK 2017, 2019), V€ain€o and Laina Kivi Foundation (PLK 2017, 2018, 2019), and University of Helsinki (PLK 2019). Disclosures No conflicts of interest are declared by the author(s). References ATS/ERS. ATS/ERS recommendations for standardized procedures for the online and offline measurement of exhaled lower respiratory nitric oxide and nasal nitric oxide, 2005. Am J Respir Crit Care Med (2005); 171: 912–930. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Int J Nurs Stud (2010); 47: 931–936. Ekroos H, Karjalainen J, Sarna S, et al. Shortterm variability of exhaled nitric oxide in young male patients with mild asthma and in healthy subjects. Respir Med (2002); 96: 895–900. H€ogman M, Str€omberg S, Schedin U, et al. Nitric oxide from the human respiratory tract efficiently quantified by standardized single breath measurements. Acta Physiol Scand (1997); 159: 345–346. H€ogman M, Drca N, Ehrstedt C, et al. Exhaled nitric oxide partitioned into alveolar, lower airways and nasal contributions. Respir Med (2000); 94: 985–991. Horvath I, Barnes PJ, Loukides S, et al.A European Respiratory Society technical standard: exhaled biomarkers in lung disease. Eur Respir J (2017); 49: 1600965. Kharitonov SA, Barnes PJ. Exhaled markers of pulmonary disease. Am J Respir Crit Care Med (2001); 163: 1693–1722. Kharitonov S, Alving K, Barnes PJ. Exhaled and nasal nitric oxide measurements: recommendations. The European Respiratory Society Task Force. Eur Respir J (1997); 10: 1683–1693. Lassmann-Klee PG, Lindholm T, Mets€al€aM, et al. Reduction of FENO by tap water and carbonated water mouthwashes: magnitude and time course. Scand J Clin Lab Invest (2018a); 78: 153–156. Lassmann-Klee PG, Lehtim€aki L, Lindholm T, et al. Influence of mouthwashes on extended exhaled nitric oxide (F ENO ) analysis. Scand J Clin Lab Invest (2018b); 78: 450–455. Lehtim€aki L, Kankaanranta H, Saarelainen S, et al. Extended Exhaled NO Measurement Differentiates between Alveolar and Bronchial Inflammation. Am J Respir Crit Care Med (2001); 163: 1557–1561. Lehtim€aki L, Kankaanranta H, Saarelainen S, et al. Bronchial nitric oxide is related to symptom relief during fluticasone treatment in COPD. Eur Respir J (2010a); 35:72–78. ©2019 The Authors. Clinical Physiology and Functional Imaging published by John Wiley & Sons Ltd on behalf of Scandinavian Society of Clinical Physiology and Nuclear Medicine 39, 5, 315–321 FENO conversion model, P. G. Lassmann-Klee et al.320
Lehtim€aki L, Oksa P, J€arvenp€a€aR,et al. Pulmonary inflammation in asbestos-exposed subjects with borderline parenchymal changes on HRCT. Respir Med (2010b); 104: 1042–1049. Leys C, Ley C, Klein O, et al. Detecting outliers: do not use standard deviation around the mean, use absolute deviation around the median. J Exp Soc Psychol (2013); 49: 764–766. NICE. (2017). Asthma: diagnosis, monitoring and chronic asthma management|Guidance and guidelines |NICE. Pellegrino R, Viegi G, Brusasco V, et al. Interpretative strategies for lung function tests. Eur Respir J (2005); 26: 948–968. R Core Team. R: A Language and Environment for Statistical Computing. (2018). R Foundation for Statistical Computing, Vienna, Austria. Rouhos A, Kainu A, Karjalainen J, et al. Atopic sensitization to common allergens without symptoms or signs of airway disorders does not increase exhaled nitric oxide. Clin Respir J (2008); 2: 141–148. Sepponen A, Lehtim€aki L, Huhtala H, et al. Alveolar and bronchial nitric oxide output in healthy children. Pediatr Pulmonol (2008); 43: 1242–1248. Silkoff PE, Sylvester JT, Zamel N, et al. Airway Nitric Oxide Diffusion in Asthma. Am J Respir Crit Care Med (2000); 161: 1218–1228. Tsoukias NM, George SC. A two-compartment model of pulmonary nitric oxide exchange dynamics. J Appl Physiol (1998); 85: 653– 666. World Medical Association Declaration of Helsinki. Ethical Principles for Medical Research Involving Human Subjects. JAMA (2013); 310: 2191. ©2019 The Authors. Clinical Physiology and Functional Imaging published by John Wiley & Sons Ltd on behalf of Scandinavian Society of Clinical Physiology and Nuclear Medicine 39, 5, 315–321 FENO conversion model, P. G. Lassmann-Klee et al.321