- Email: [email protected]
The Diagnostic Accuracy of Fractional Exhaled Nitric Oxide Testing in Asthma: A Systematic Review and Meta-analyses
ORIGINAL ARTICLE
The Diagnostic Accuracy of Fractional Exhaled Nitric Oxide Testing in Asthma: A Systematic Review and Meta-analyses Zhen Wang, PhD; Paolo T. Pianosi, MD; Karina A. Keogh, MD; Feras Zaiem, MD; Mouaz Alsawas, MD, MSc; Fares Alahdab, MD; Jehad Almasri, MD; Khaled Mohammed, MBBCh; Laura Larrea-Mantilla, MD; Wigdan Farah, MBBS; Lubna Daraz, PhD; Patricia Barrionuevo, MD; Allison S. Morrow, BA; Larry J. Prokop, MLS; and Mohammad Hassan Murad, MD, MPH Abstract Objective: To evaluate the diagnostic accuracy of fractional exhaled nitric oxide (FeNO) measurement in individuals with suspected asthma. Methods: We searched MEDLINE, EMBASE, PsycINFO, Cochrane databases, and SciVerse Scopus from the databases’ inception through April 4, 2017, for studies that enrolled patients aged 5 years and older with suspected asthma and evaluated FeNO diagnostic accuracy. Independent reviewers selected studies and extracted data. We used the symmetric hierarchical summary receiver operating characteristic models to estimate test performance. Results: We included 43 studies with a total of 13,747 patients. In adults, using FeNO cutoffs of less than 20, 20 to 29, 30 to 39, and 40 or more parts per billion, FeNO testing had sensitivities of 0.80, 0.69, 0.53, and 0.41, respectively, and specificities of 0.64, 0.78, 0.85, and 0.93, respectively. In children, using FeNO cutoffs of less than 20 and 20 to 29 parts per billion, FeNO testing had sensitivities of 0.78 and 0.61, respectively, and specificities of 0.79 and 0.89, respectively. Depending on the FeNO cutoff, the posttest odds of having asthma with a positive FeNO test result increased by 2.80- to 7.00-fold. Diagnostic accuracy was modestly better in corticosteroid-naive asthmatics, children, and nonsmokers than in the overall population. Conclusion: Fractional exhaled nitric oxide measurement has moderate accuracy to diagnose asthma in individuals aged 5 years and older. Test performance may be modestly better in corticosteroid-naive asthmatics, children, and nonsmokers than in the general population with suspected asthma. Trial Registration: International Prospective Register of Systematic Reviews (PROSPERO) Identifier: CRD42016047887 ª 2017 Mayo Foundation for Medical Education and Research
A
sthma is a chronic inflammatory disorder of the airways characterized by varying degrees of airflow obstruction. Bronchoconstriction, inflammatory cell infiltration, and airway edema reduce airflow intermittently, often in response to specific exposures, resulting in respiratory symptoms.1 It is estimated that 24.6 million Americans had asthma in 2015.2 Diagnosing asthma is challenging. The common symptoms, such as shortness of breath, wheezing, and cough, are relatively nonspecific. Various tests, including bronchodilator response
n
Mayo Clin Proc. 2017;nn(n):1-8
and positive results on bronchial challenge, may be used by clinicians to aid in the diagnosis of asthma in the appropriate clinical context. However, the diagnosis remains clinical, based on compatible symptoms and evidence of reversible airway obstruction; no single criterion standard diagnostic test exists. More recently, fractional exhaled nitric oxide (FeNO) concentration has been added to the list of tests that clinicians may use to diagnose asthma. Thus, the objective of this systematic review was to evaluate the diagnostic accuracy of FeNO concentration in individuals
Mayo Clin Proc. n XXX 2017;nn(n):1-8 n https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org n ª 2017 Mayo Foundation for Medical Education and Research
From the Mayo Clinic Evidence-based Practice Center, Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery (Z.W., F.Z., M.A., F.A., J.A., K.M., L.L.-M., W.F., L.D., P.B., A.S.M., M.H.M.), Division of Health Care Policy and Research, Department of Health Sciences Research Affiliations continued at the end of this article.
1
MAYO CLINIC PROCEEDINGS
aged 5 years and older with suspected asthma. METHODS The reporting of this systematic review complies with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statements.3 We developed the study protocol with input from clinical and research experts and professional organizations. The study protocol is registered in the International Prospective Register of Systematic reviews (PROSPERO Identifier: CRD42016047887). Data Sources and Searches We conducted a comprehensive literature search of 6 databases, including Ovid MEDLINE In-Process & Other Non-Indexed Citations, Ovid MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and SciVerse Scopus from the databases’ inception to April 4, 2017. We searched relevant systematic reviews and conducted reference mining of relevant publications to identify additional literature. We searched gray literature through all of the following: US Food and Drug Administration device registration studies, ClinicalTrials.gov, Health Canada, Medicines and Healthcare Products Regulatory Agency, Agency for Healthcare Research and Quality Horizon Scanning System, conference proceedings, patient advocate group websites, and medical society websites. An experienced medical librarian (L.J.P.), with input from the study investigators, developed and executed the search strategy (Supplemental Appendix, available online at http://www. mayoclinicproceedings.org). An independent librarian reviewed the search strategy. Study Selection We included randomized clinical trials and observational studies that (1) enrolled patients aged 5 years and older with suspected asthma, (2) compared FeNO testing (diagnostic test) to standard diagnostic testing of asthma by health care professionals based on history, clinical course, or other diagnostic tests (clinical diagnosis, bronchodilator response, and positive results on bronchial challenge) (reference test), and (3) reported FeNO diagnostic accuracy. We excluded studies with mixed 2
Mayo Clin Proc.
n
populations (eg, patients with asthma and chronic obstructive lung disease) without reporting separate results for individuals with asthma. We also excluded surveys, narrative reviews, editorials, letters, or errata, qualitative research, in vitro studies, and animal studies. We did not restrict study location, publication time, or language. Independent reviewers working in pairs screened the titles and abstracts of all citations and then the full text of eligible references. Discrepancies between the reviewers were resolved through discussions and consensus. If they did not reach consensus, a third reviewer was added to resolve the difference. Data Extraction and Quality Assessment We developed a pilot-tested standardized data extraction form at the beginning of the study. The following information was extracted: author, study design, inclusion and exclusion criteria, patient characteristics, characteristics of FeNO test and reference tests, diagnostic accuracy measures (reported as true-positives, true-negatives, false-positives, and falsenegatives), and related FeNO cutoff values. We used the QUADAS-2 instrument to evaluate risk of bias of the included studies.4 Data extraction and quality assessment were completed by pairs of independent reviewers. Main Outcome Measures The outcomes of interest were diagnostic accuracy measures, including sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio (DOR). Data Synthesis and Analysis We categorized FeNO cutoff values as less than 20, 20 to 29, 30 to 39, and 40 or more parts per billion (ppb). Analyses were conducted by age group (less than 18 years vs 18 years or older) as well as overall. We extracted true-positives, true-negatives, false-positives, and false-negatives from the included studies for each FeNO cutoff value and reference test. If multiple cutoffs within the same category were reported from the same study, we selected the one with the highest DOR. The DOR is a single indicator of diagnostic performance that facilitates comparison across tests. It was defined as the ratio of the odds of positivity in XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org
FENO TESTING IN ASTHMA
individuals with disease relative to the odds in individuals without disease and is calculated as (true-positives true-negatives)/ (false-positives false-negatives).5 We used the symmetric hierarchical summary receiver operating characteristic (HSROC) models to jointly estimate sensitivity and specificity, positive and negative likelihood ratios, and DOR.6 We also drew the HSROC curves based on the estimates. We were unable to pool estimates when the number of studies was less than 4. We examined potential publication bias by evaluating funnel plots symmetry and Deeks funnel plot asymmetry tests if the number of studies was large (>20). To explore heterogeneity, we conducted the following subgroup analyses based on factors defined a priori: risk of bias (low, moderate, and high risk of bias), health status of reference group (healthy vs symptomatic individuals without asthma), tobacco use (nonsmoker vs smoker), and type of reference tests (clinical diagnosis, positive results on bronchial challenge, bronchodilator response, or the combination). We were unable to conduct preplanned other subgroup analyses, such as those defined by asthma phenotype, adequate testing procedures, body mass index or weight, manufacturer and device model, and exhalation flow rate because studies did not provide sufficient stratified data per subgroup variable. We conducted sensitivity analyses based on use of inhaled corticosteroids (ICS) before FeNO testing and atopy status to evaluate robustness of our findings. All statistical analyses were conducted using Stata statistical software, version 14.1 (StataCorp). Grading the Quality of Evidence (ie, Certainty in Estimates) We found only observational studies in this systematic review. Diagnostic observational studies start as high quality of evidence and can be rated downward because of methodological limitations of the studies, lack of precision, and likelihood of publication bias. We did not rate downward for statistical heterogeneity (which is always high in diagnostic metaanalyses) or consider diagnostic accuracy measures as surrogate outcomes.7,8
Mayo Clin Proc. n XXX 2017;nn(n):1-8 www.mayoclinicproceedings.org
n
RESULTS Description of the Evidence Database searches identified 3884 citations. An additional 61 references were identified through gray literature search and crossreferencing. We included in our analyses 43 observational studies with 13,747 patients (Figure 1). We included one study that enrolled children aged 2 to 7 years with an average age of 4.49 years.9 Most of the studies (33) included only adults 18 years or older; 10 studies included children less than 18 years old. Nineteen studies were longitudinal studies, 23 were cross-sectional studies, and 1 was a case-control study. The studies were conducted in the United States (n¼2), Canada (n¼2), Europe (n¼26), and other countries (n¼13). The FeNO concentration was measured online in 10 studies and offline in 3, and 1 study used both methods. In terms of reference tests used to compare with FeNO, 10 studies used clinical diagnosis, 13 used positive bronchial challenge test results, and 21 combined tests (clinical diagnosis, positive results on bronchial challenge, and/or bronchodilator response). The characteristics of these studies are summarized in Supplemental Table 1 (available online at http://www.mayoclinicproceedings.org). Most of the included studies had low (46.51%) or medium (32.56%) risk of bias. High risk of bias was noted primarily in the areas of cohort selection, including representativeness of the study population (whether patients were consecutive and represented the total number of eligible patients at a particular institution) and whether studies enrolled patients with diagnostic uncertainty (ie, with symptoms suggestive of asthma). The details of risk of bias assessment are presented in Supplemental Table 2 (available online at http://www.mayoclinicproceedings.org) and summarized in Figure 2. We found potential publication bias for cutoffs of less than 20 ppb and no indication of publication bias for cutoffs of 20 to 29 ppb (Supplemental Figures 1 and 2, available online at http://www.mayoclinicproceedings. org). We were not able to evaluate potential publication bias for other cutoffs. Overall there was no strong evidence of publication bias.
https://doi.org/10.1016/j.mayocp.2017.11.012
3
MAYO CLINIC PROCEEDINGS
3884 Citations identified through database searching
61 Citations identified from gray literature and cross-references
2990 Citations excluded
3945 Records screened
912 Full-text articles excluded 69 Did not evaluate FeNO diagnostic accuracy 67 No comparison group 7 Not original study 13 Abstracts, systematic reviews 117 No population of interest 622 No outcome of interest 17 Other
955 Full text screened
43 Studies included in the systematic review
FIGURE 1. Study flowchart. FeNO ¼ fractional exhaled nitric oxide.
Diagnostic Accuracy of FeNO Testing In adults (aged 18 years or older), using FeNO cutoffs of less than 20, 20 to 29, 30 to 39, and 40 or more ppb, FeNO testing had sensitivities of 0.80 (95% CI, 0.72-0.86), 0.69 (95% CI, 0.57-0.79), 0.53 (95% CI, 0.35-0.70), and 0.41 (95% CI, 0.24-0.62), respectively, and specificities of 0.64 (95% CI, 0.46-0.79), 0.78 (95% CI, 0.66-0.86), 0.85 (95% CI, 0.77-0.90), and 0.93 (95% CI, 0.86-0.97), respectively (moderate quality of evidence). In children (aged 18 years), using FeNO cutoffs of less than 20 and 20 to 29 ppb, FeNO testing had sensitivities of 0.78 (95% CI, 0.59-0.90) and 0.61 (95% CI, 0.44-0.76), respectively, and specificities of 0.79 (95% CI, 0.55-0.92) and 0.89 (95% CI, 0.800.94), respectively (moderate quality of evidence). Diagnostic accuracy was overall higher in children than in adults. Overall, the posttest odds of having asthma given a positive FeNO test result increased by 2.80- to 7.00-fold. The Table presents a summary of the diagnostic accuracy of FeNO testing. Supplemental Figures 3 through 6 present HSROC curves for FeNO cutoffs of less than 20, 20 to 29, 30 to 39, and 40 or more ppb, respectively.
Subgroup and Sensitivity Analyses Supplemental Tables 3 through 6 (available online at http://www.mayoclinicproceedings.org) 4
Mayo Clin Proc.
n
list the findings for subgroup analyses. Health status of reference groups impacted the diagnostic accuracy of FeNO. The diagnostic accuracy of FeNO may be overestimated in studies that used healthy controls (DOR, 16.45; 95% CI, 7.85-34.49) compared with symptomatic controls (DOR, 4.42; 95% CI, 2.33-8.35) for FeNO cutoffs of less than 20 ppb. Risk of bias might partially explain heterogeneity in the diagnostic accuracy of FeNO, with greater reported diagnostic accuracy as the risk of bias increased (DORs across cutoffs of 11.29, 10.33, and 8.81 for high, medium, and low risk, respectively, for FeNO cutoffs of less than 20 ppb). With regard to specific reference tests (criterion standard), FeNO testing was found to have higher DORs compared with clinical diagnosis than when compared with positive results on bronchial challenge or the combination of clinical diagnosis, bronchial challenge, or bronchodilator response. We also found that diagnostic accuracy was higher in nonsmokers (sensitivity, 0.70; 95% CI, 0.61-0.78; specificity, 0.80; 95% CI, 0.74-0.85; DOR, 9.49; 95% CI, 5.62-16.01) than in smokers (DORs ranged from 2.53 to 5.4610,11). In a sensitivity analysis, we were only able to analyze studies that evaluated the diagnostic accuracy of FeNO in corticosteroid-naive asthmatics (the remaining studies had a mix of populations, including corticosteroid-naive XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org
FENO TESTING IN ASTHMA
Reference test
Flow and timing
Over-all
Low risk
Medium risk
High risk
Unclear risk
Overall risk of bias
14
20
All were patients included in the analysis
30
2
Index test
0
10
1
Patients received the same reference test
40
01 2
All patients received a reference test
40
02 1
Appropriate interval between index test and reference test
33
01
35
Concern that the target condition did not match the review question Reference test results interpreted without knowledge of the results of index test
38
Reference test correctly classifed the target condition
37
7
1
0 4 1 0
6
The cutoff used was prespecified
02
33
8 101
41
12 2
38
Concern that the included patients did not match the review question Study avoided inappropriate exclusions
31
1 4
Case-control design avoided
31
0
19
0
0 010
42
Index test results interpreted without knowledge of the results of the reference standard
Study enrolled consecutive or random sample of patients
9 0
Concern that the index test differed from the review question
Patient selection
9
15
7 12
0 9
FIGURE 2. Summary of risk of bias assessment using the QUADAS-2 tool (n¼43).
patients and corticosteroid users) (Supplemental Table 7, available online at http://www. mayoclinicproceedings.org). At cutoffs of less than 20 ppb, FeNO had the highest accuracy in this group of patients compared with patients in the main results (sensitivity, 0.79; 95% CI, 0.67-0.87; specificity, 0.77; 95% CI, 0.56-0.90; and DOR, 12.25; 95% CI, 5.73-26.21). Results in other cutoffs were different and inconsistent. In another sensitivity analysis, we analyzed only studies that evaluated the diagnostic accuracy of FeNO in asthmatic patients with atopy. The results, which included a small number of studies (4), revealed accuracy measures that were similar to those from the main analysis (sensitivity, 0.63; 95% CI, 0.43-0.80; specificity, 0.79; 95% CI, 0.65-0.89; and DOR, 6.67; 95% Mayo Clin Proc. n XXX 2017;nn(n):1-8 www.mayoclinicproceedings.org
n
CI, 1.59-27.95) (Supplemental Table 8, available online at http://www.mayoclinicproceedings. org). DISCUSSION We conducted a systematic review with metaanalyses to evaluate the diagnostic accuracy of FeNO testing in children and adults with suspected asthma. This is an important question because of the high prevalence of asthma and the risks associated with undiagnosed asthma as well as the increasingly recognized issue of asthma overdiagnosis and the risks of overtreatment.12,13 We found 43 observational studies addressing this question. Most had low or medium risk of bias. The FeNO concentration has moderate diagnostic
https://doi.org/10.1016/j.mayocp.2017.11.012
5
MAYO CLINIC PROCEEDINGS
TABLE. Summary of Diagnostic Accuracy of FeNO Testing for Suspected Asthma FeNO cutoff <20 ppb
Subgroup
Conclusion
18 Years or older
Sensitivity: 0.80; 95% CI, 0.72-0.86 Specificity: 0.64; 95% CI, 0.46-0.79 DOR: 7.28; 95% CI, 4.04-13.11 LRþ: 2.25; 95% CI, 1.47-3.44 LR: 0.31; 95% CI, 0.23-0.41 Sensitivity: 0.78; 95% CI, 0.59-0.90 Specificity: 0.79; 95% CI, 0.55-0.92 DOR: 13.44; 95% CI, 3.56-50.71 LRþ: 3.76; 95% CI, 1.53-9.26 LR: 0.28; 95% CI, 0.14-0.56 Sensitivity: 0.69; 95% CI, 0.57-0.79 Specificity: 0.78; 95% CI, 0.66-0.86 DOR: 7.70; 95% CI, 4.12-14.40 LRþ: 3.10; 95% CI, 2.02-4.74 LR: 0.40; 95% CI, 0.29-0.56 Sensitivity: 0.61; 95% CI, 0.44-0.76 Specificity: 0.89; 95% CI, 0.80-0.94 DOR: 12.13; 95% CI, 5.98-24.63 LRþ: 5.34; 95% CI, 3.12-9.14 LR: 0.44; 95% CI, 0.30-0.65 Sensitivity: 0.53; 95% CI, 0.35-0.70 Specificity: 0.85; 95% CI, 0.77-0.90 DOR: 6.27; 95% CI, 3.70-10.64 LRþ: 3.46; 95% CI, 2.57-4.66 LR: 0.55; 95% CI, 0.39-0.78 NA Sensitivity: 0.41; 95% CI, 0.24-0.62 Specificity: 0.93; 95% CI, 0.86-0.97 DOR: 9.84; 95% CI, 5.46-17.75 LRþ: 6.18; 95% CI, 3.64-10.47 LR: 0.63; 95% CI, 0.47-0.85 NA
Younger than 18 years
20-29 ppb
18 Years or older
Younger than 18 years
30-39 ppb
18 Years or older
40 ppb
Younger than 18 years 18 Years or older
Younger than 18 years
DOR ¼ diagnostic odds ratio; FeNO ¼ fractional exhaled nitric oxide; LRþ ¼ likelihood ratio for a positive test result; LR ¼ likelihood ratio for a negative test result; ppb ¼ parts per billion; NA ¼ not available.
accuracy for asthma with DORs that range from 5.85 to 16.95 across various cutoff points (in comparison, a test with 0.80 sensitivity and 0.80 specificity would have a DOR of 16). As expected, with increasing cutoff values, FeNO had a gradual decrease in sensitivity and improved specificity (for cutoffs of <20, 20-29, 30-39, and 40 ppb, FeNO testing has sensitivities of 0.78, 0.63, 0.56, and 0.41, respectively, and specificities of 0.71, 0.81, 0.84, and 0.94, respectively). Therefore, knowing the cutoffs used for test interpretation is critical for interpretation by clinicians. The 2011 American Thoracic 6
Mayo Clin Proc.
n
Society guidelines14 have recommended cutoffs of less than 25 ppb (<20 ppb in children) for low FeNO and greater than 50 ppb (>35 ppb in children) for high FeNO as the most useful cutoffs to make the determination that eosinophlic inflammation and response to corticosteroids would be unlikely and likely, respectively. Intermediate results should be interpreted with caution (FeNO of 25-50 ppb [20-35 ppb in children]).14 Inferences from several preplanned subgroup analyses were limited because of the limited number of studies and the heterogeneity of populations, interventions, and outcomes, particularly regarding the impacts of reference test, the presence of atopy, current use of ICS, and smoking status on FeNO diagnostic performance. Asthma can sometimes be difficult to diagnose, and FeNO can be helpfuldmore so in ruling in than ruling out asthmadin making therapeutic decisions more evidence based. For example, in a patient who has compatible symptoms and is clearly atopic (eg, history of eczema, positive skin test results, peripheral blood eosinophilia, or elevated serum IgE level), elevated FeNO implies a diagnosis of asthma and that treatment with ICS is indicated. Conversely, low FeNO (eg, <20 ppb) implies that these compatible symptoms are not likely due to asthma. Thus, clinicians should direct treatment to other causes of symptoms. The caveat is that low FeNO findings do not exclude asthma, but clinical judgment and vigilance are warranted. Such a scenario is common in pediatric practice, in which this review already concluded that evidence was stronger in children and adolescents, likely because most asthma in this population is atopic. It is precisely in this population that overdiagnosis of asthma is a bigger clinical problem than underdiagnosis. In this case, patients with suspected asthma should have further testing to confirm the diagnosis. The first test should be spirometry with an assessment of bronchodilator response. If this test does not confirm the diagnosis but the index of suspicion for asthma is still high, further testing should be performed. Measurement of FeNO should be considered at this stage as a second-line test (other second-line tests would include the various bronchoprovocation studies). XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org
FENO TESTING IN ASTHMA
One challenge relates to the fact that there is no true criterion standard for diagnosing asthma. Although we did not label studies as having increased risk of bias because of this issue, we recognize that it can impact diagnostic accuracy. In addition, a wide range of reference tests were reported. We categorized these reference tests as clinical diagnosis, positive bronchial challenge test results, or a combination of clinical diagnosis, positive results on bronchial challenge, and/or bronchodilator response. However, considerable heterogeneity still exists, such as to how and when these tests were administered. Although categorizations of less than 20, 20 to 29, 30 to 39, and 40 or more ppb helped reduce heterogeneity and facilitated meta-analyses, we were not able to definitively present a best cutoff overall or within each category. We were also not able to conduct some planned subgroup analyses because of lack of data, including asthma phenotype, body mass index, and exhalation flow rate. CONCLUSION The FeNO concentration has moderate accuracy to diagnose asthma in individuals aged 5 years and older. Test performance is modestly better in corticosteroid-naive asthmatics, children, and nonsmokers than in the general population with suspected asthma. ACKNOWLEDGMENTS We gratefully acknowledge the following individuals for their contributions to this study: the task order officers from Agency for Healthcare Research and Quality; Anas Daghestani, MD, Austin Regional Clinic; Serpil C. Erzurum, MD, Cleveland Clinic Foundation; Elliot Israel, MD, Brigham and Women’s Hospital; Rohit Katial, MD, University of Colorado; Nicholas Kenyon, MD, University of California, Davis; Todd A. Mahr, MD, Gundersen Lutheran Medical Center; John G. Mastronarde, MD, MSc, Providence Portland Medical Center; Wayne J. Morgan, MD, University of Arizona College of Medicine; Shahzeb Munir, MD, University of Wisconsin School of Medicine and Public Health; Augusto Litonjua, MD, MPH, Harvard Medical School; John Sundy, MD, PhD, Duke University School of Medicine; and Sally Wenzel, MD, University of Pittsburgh Medical Center. Mayo Clin Proc. n XXX 2017;nn(n):1-8 www.mayoclinicproceedings.org
n
The funding source had no role in conduct of the review; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the submitted manuscript; and decision to submit the manuscript for publication. The findings and conclusions in this article are those of the authors, who are responsible for its contents, and do not necessarily represent the views of the Agency for Healthcare Research and Quality or the US government. Therefore, no statement in this article should be construed as the official position of the Agency for Healthcare Research and Quality. Author contributions: All authors contributed to the acquisition, analysis, or interpretation of data for the study and critical revision of the submitted manuscript for important intellectual content. Concept and design: Drs Wang, Pianosi, Keogh, and Murad; drafting of the manuscript: Drs Wang, Pianosi, Keogh, and Murad; statistical analysis: Drs Wang and Murad; administrative, technical, or material support: Dr Alsawas. As the guarantors of the study, Drs Wang and Murad take full responsibility for the work as a whole, including the study design, access to data, and the decision to submit and publish the manuscript. SUPPLEMENTAL ONLINE MATERIAL Supplemental material can be found online at http://www.mayoclinicproceedings.org. Supplemental material attached to journal articles has not been edited, and the authors take responsibility for the accuracy of all data. Abbreviations and Acronyms: DOR = diagnostic odds ratio; FeNO = fractional exhaled nitric oxide; HSROC = hierarchical summary receiver operating characteristic; ICS = inhaled corticosteroids; ppb = parts per billion Affiliations (Continued from the first page of this article.): (Z.W.), Division of Pediatric Pulmonology, Department of Pediatric and Adolescent Medicine (P.T.P.), Division of Pulmonary and Critical Care Medicine (K.A.K.), and Mayo Clinic Libraries (L.J.P.), Mayo Clinic, Rochester, MN; and Paediatric Respirology, King’s College Hospital NHS Foundation Trust, London, UK (P.T.P.). Grant Support: This work was funded by the Agency for Healthcare Research and Quality under contract HHSA 290201500013I to support evidence synthesis as part of the Evidence-based Practice Center Program.
https://doi.org/10.1016/j.mayocp.2017.11.012
7
MAYO CLINIC PROCEEDINGS
Correspondence: Address to Mohammad Hassan Murad, MD, MPH, Mayo Clinic Evidence-based Practice Center, Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, 200 First St SW, Rochester, MN 55905 ([email protected]).
REFERENCES 1. National Asthma Education and Prevention Program (National Heart, Lung, and Blood Institute) Third Expert Panel on the Management of Asthma. Expert Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma. Bethesda, MD: US Dept of Health and Human Services, National Institutes of Health; 2007. NIH publication 07-4051. 2. Centers for Disease Control and Prevention. Most recent asthma data. Centers for Disease Control and Prevention website. https://www.cdc.gov/asthma/most_recent_data.htm. Updated June 2017. Accessed December 4, 2017. 3. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and metaanalyses: the PRISMA statement. BMJ. 2009;339:b2535. 4. Whiting PF, Rutjes AW, Westwood ME, et al; QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536. 5. Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol. 2003;56(11):1129-1135. 6. Rutter CM, Gatsonis CA. A hierarchical regression approach to meta-analysis of diagnostic test accuracy evaluations. Stat Med. 2001;20(19):2865-2884.
8
Mayo Clin Proc.
n
7. Schünemann HJ, Mustafa R, Brozek J, et al; GRADE Working Group. GRADE Guidelines: 16. GRADE evidence to decision frameworks for tests in clinical practice and public health. J Clin Epidemiol. 2016;76:89-98. 8. Mustafa RA, Santesso N, Khatib R, et al. Systematic reviews and meta-analyses of the accuracy of HPV tests, visual inspection with acetic acid, cytology, and colposcopy. Int J Gynaecol Obstet. 2016;132(3):259-265. 9. Avital A, Uwyyed K, Berkman N, Godfrey S, Bar-Yishay E, Springer C. Exhaled nitric oxide and asthma in young children. Pediatr Pulmonol. 2001;32(4):308-313. 10. Matsunaga K, Hirano T, Akamatsu K, et al. Exhaled nitric oxide cutoff values for asthma diagnosis according to rhinitis and smoking status in Japanese subjects. Allergol Int. 2011;60(3): 331-337. 11. Malinovschi A, Backer V, Harving H, Porsbjerg C. The value of exhaled nitric oxide to identify asthma in smoking patients with asthma-like symptoms. Respir Med. 2012; 106(6):794-801. 12. Looijmans-van den Akker I, van Luijn K, Verheij T. Overdiagnosis of asthma in children in primary care: a retrospective analysis. Br J Gen Pract. 2016;66(644):e152-e157. 13. Aaron SD, Vandemheen KL, FitzGerald JM, et al; Canadian Respiratory Research Network. Reevaluation of diagnosis in adults with physician-diagnosed asthma. JAMA. 2017;317(3): 269-279. 14. Dweik RA, Boggs PB, Erzurum SC, et al; American Thoracic Society Committee on Interpretation of Exhaled Nitric Oxide Levels (FENO) for Clinical Applications. An official ATS clinical practice guideline: interpretation of exhaled nitric oxide levels (FENO) for clinical applications. Am J Respir Crit Care Med. 2011;184(5):602-615.
XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org
The Diagnostic Accuracy of Fractional Exhaled Nitric Oxide Testing in Asthma: A Systematic Review and Meta-analyses Zhen Wang, PhD; Paolo T. Pianosi, MD; Karina A. Keogh, MD; Feras Zaiem, MD; Mouaz Alsawas, MD, MSc; Fares Alahdab, MD; Jehad Almasri, MD; Khaled Mohammed, MBBCh; Laura Larrea-Mantilla, MD; Wigdan Farah, MBBS; Lubna Daraz, PhD; Patricia Barrionuevo, MD; Allison S. Morrow, BA; Larry J. Prokop, MLS; and Mohammad Hassan Murad, MD, MPH Abstract Objective: To evaluate the diagnostic accuracy of fractional exhaled nitric oxide (FeNO) measurement in individuals with suspected asthma. Methods: We searched MEDLINE, EMBASE, PsycINFO, Cochrane databases, and SciVerse Scopus from the databases’ inception through April 4, 2017, for studies that enrolled patients aged 5 years and older with suspected asthma and evaluated FeNO diagnostic accuracy. Independent reviewers selected studies and extracted data. We used the symmetric hierarchical summary receiver operating characteristic models to estimate test performance. Results: We included 43 studies with a total of 13,747 patients. In adults, using FeNO cutoffs of less than 20, 20 to 29, 30 to 39, and 40 or more parts per billion, FeNO testing had sensitivities of 0.80, 0.69, 0.53, and 0.41, respectively, and specificities of 0.64, 0.78, 0.85, and 0.93, respectively. In children, using FeNO cutoffs of less than 20 and 20 to 29 parts per billion, FeNO testing had sensitivities of 0.78 and 0.61, respectively, and specificities of 0.79 and 0.89, respectively. Depending on the FeNO cutoff, the posttest odds of having asthma with a positive FeNO test result increased by 2.80- to 7.00-fold. Diagnostic accuracy was modestly better in corticosteroid-naive asthmatics, children, and nonsmokers than in the overall population. Conclusion: Fractional exhaled nitric oxide measurement has moderate accuracy to diagnose asthma in individuals aged 5 years and older. Test performance may be modestly better in corticosteroid-naive asthmatics, children, and nonsmokers than in the general population with suspected asthma. Trial Registration: International Prospective Register of Systematic Reviews (PROSPERO) Identifier: CRD42016047887 ª 2017 Mayo Foundation for Medical Education and Research
A
sthma is a chronic inflammatory disorder of the airways characterized by varying degrees of airflow obstruction. Bronchoconstriction, inflammatory cell infiltration, and airway edema reduce airflow intermittently, often in response to specific exposures, resulting in respiratory symptoms.1 It is estimated that 24.6 million Americans had asthma in 2015.2 Diagnosing asthma is challenging. The common symptoms, such as shortness of breath, wheezing, and cough, are relatively nonspecific. Various tests, including bronchodilator response
n
Mayo Clin Proc. 2017;nn(n):1-8
and positive results on bronchial challenge, may be used by clinicians to aid in the diagnosis of asthma in the appropriate clinical context. However, the diagnosis remains clinical, based on compatible symptoms and evidence of reversible airway obstruction; no single criterion standard diagnostic test exists. More recently, fractional exhaled nitric oxide (FeNO) concentration has been added to the list of tests that clinicians may use to diagnose asthma. Thus, the objective of this systematic review was to evaluate the diagnostic accuracy of FeNO concentration in individuals
Mayo Clin Proc. n XXX 2017;nn(n):1-8 n https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org n ª 2017 Mayo Foundation for Medical Education and Research
From the Mayo Clinic Evidence-based Practice Center, Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery (Z.W., F.Z., M.A., F.A., J.A., K.M., L.L.-M., W.F., L.D., P.B., A.S.M., M.H.M.), Division of Health Care Policy and Research, Department of Health Sciences Research Affiliations continued at the end of this article.
1
MAYO CLINIC PROCEEDINGS
aged 5 years and older with suspected asthma. METHODS The reporting of this systematic review complies with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statements.3 We developed the study protocol with input from clinical and research experts and professional organizations. The study protocol is registered in the International Prospective Register of Systematic reviews (PROSPERO Identifier: CRD42016047887). Data Sources and Searches We conducted a comprehensive literature search of 6 databases, including Ovid MEDLINE In-Process & Other Non-Indexed Citations, Ovid MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and SciVerse Scopus from the databases’ inception to April 4, 2017. We searched relevant systematic reviews and conducted reference mining of relevant publications to identify additional literature. We searched gray literature through all of the following: US Food and Drug Administration device registration studies, ClinicalTrials.gov, Health Canada, Medicines and Healthcare Products Regulatory Agency, Agency for Healthcare Research and Quality Horizon Scanning System, conference proceedings, patient advocate group websites, and medical society websites. An experienced medical librarian (L.J.P.), with input from the study investigators, developed and executed the search strategy (Supplemental Appendix, available online at http://www. mayoclinicproceedings.org). An independent librarian reviewed the search strategy. Study Selection We included randomized clinical trials and observational studies that (1) enrolled patients aged 5 years and older with suspected asthma, (2) compared FeNO testing (diagnostic test) to standard diagnostic testing of asthma by health care professionals based on history, clinical course, or other diagnostic tests (clinical diagnosis, bronchodilator response, and positive results on bronchial challenge) (reference test), and (3) reported FeNO diagnostic accuracy. We excluded studies with mixed 2
Mayo Clin Proc.
n
populations (eg, patients with asthma and chronic obstructive lung disease) without reporting separate results for individuals with asthma. We also excluded surveys, narrative reviews, editorials, letters, or errata, qualitative research, in vitro studies, and animal studies. We did not restrict study location, publication time, or language. Independent reviewers working in pairs screened the titles and abstracts of all citations and then the full text of eligible references. Discrepancies between the reviewers were resolved through discussions and consensus. If they did not reach consensus, a third reviewer was added to resolve the difference. Data Extraction and Quality Assessment We developed a pilot-tested standardized data extraction form at the beginning of the study. The following information was extracted: author, study design, inclusion and exclusion criteria, patient characteristics, characteristics of FeNO test and reference tests, diagnostic accuracy measures (reported as true-positives, true-negatives, false-positives, and falsenegatives), and related FeNO cutoff values. We used the QUADAS-2 instrument to evaluate risk of bias of the included studies.4 Data extraction and quality assessment were completed by pairs of independent reviewers. Main Outcome Measures The outcomes of interest were diagnostic accuracy measures, including sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio (DOR). Data Synthesis and Analysis We categorized FeNO cutoff values as less than 20, 20 to 29, 30 to 39, and 40 or more parts per billion (ppb). Analyses were conducted by age group (less than 18 years vs 18 years or older) as well as overall. We extracted true-positives, true-negatives, false-positives, and false-negatives from the included studies for each FeNO cutoff value and reference test. If multiple cutoffs within the same category were reported from the same study, we selected the one with the highest DOR. The DOR is a single indicator of diagnostic performance that facilitates comparison across tests. It was defined as the ratio of the odds of positivity in XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org
FENO TESTING IN ASTHMA
individuals with disease relative to the odds in individuals without disease and is calculated as (true-positives true-negatives)/ (false-positives false-negatives).5 We used the symmetric hierarchical summary receiver operating characteristic (HSROC) models to jointly estimate sensitivity and specificity, positive and negative likelihood ratios, and DOR.6 We also drew the HSROC curves based on the estimates. We were unable to pool estimates when the number of studies was less than 4. We examined potential publication bias by evaluating funnel plots symmetry and Deeks funnel plot asymmetry tests if the number of studies was large (>20). To explore heterogeneity, we conducted the following subgroup analyses based on factors defined a priori: risk of bias (low, moderate, and high risk of bias), health status of reference group (healthy vs symptomatic individuals without asthma), tobacco use (nonsmoker vs smoker), and type of reference tests (clinical diagnosis, positive results on bronchial challenge, bronchodilator response, or the combination). We were unable to conduct preplanned other subgroup analyses, such as those defined by asthma phenotype, adequate testing procedures, body mass index or weight, manufacturer and device model, and exhalation flow rate because studies did not provide sufficient stratified data per subgroup variable. We conducted sensitivity analyses based on use of inhaled corticosteroids (ICS) before FeNO testing and atopy status to evaluate robustness of our findings. All statistical analyses were conducted using Stata statistical software, version 14.1 (StataCorp). Grading the Quality of Evidence (ie, Certainty in Estimates) We found only observational studies in this systematic review. Diagnostic observational studies start as high quality of evidence and can be rated downward because of methodological limitations of the studies, lack of precision, and likelihood of publication bias. We did not rate downward for statistical heterogeneity (which is always high in diagnostic metaanalyses) or consider diagnostic accuracy measures as surrogate outcomes.7,8
Mayo Clin Proc. n XXX 2017;nn(n):1-8 www.mayoclinicproceedings.org
n
RESULTS Description of the Evidence Database searches identified 3884 citations. An additional 61 references were identified through gray literature search and crossreferencing. We included in our analyses 43 observational studies with 13,747 patients (Figure 1). We included one study that enrolled children aged 2 to 7 years with an average age of 4.49 years.9 Most of the studies (33) included only adults 18 years or older; 10 studies included children less than 18 years old. Nineteen studies were longitudinal studies, 23 were cross-sectional studies, and 1 was a case-control study. The studies were conducted in the United States (n¼2), Canada (n¼2), Europe (n¼26), and other countries (n¼13). The FeNO concentration was measured online in 10 studies and offline in 3, and 1 study used both methods. In terms of reference tests used to compare with FeNO, 10 studies used clinical diagnosis, 13 used positive bronchial challenge test results, and 21 combined tests (clinical diagnosis, positive results on bronchial challenge, and/or bronchodilator response). The characteristics of these studies are summarized in Supplemental Table 1 (available online at http://www.mayoclinicproceedings.org). Most of the included studies had low (46.51%) or medium (32.56%) risk of bias. High risk of bias was noted primarily in the areas of cohort selection, including representativeness of the study population (whether patients were consecutive and represented the total number of eligible patients at a particular institution) and whether studies enrolled patients with diagnostic uncertainty (ie, with symptoms suggestive of asthma). The details of risk of bias assessment are presented in Supplemental Table 2 (available online at http://www.mayoclinicproceedings.org) and summarized in Figure 2. We found potential publication bias for cutoffs of less than 20 ppb and no indication of publication bias for cutoffs of 20 to 29 ppb (Supplemental Figures 1 and 2, available online at http://www.mayoclinicproceedings. org). We were not able to evaluate potential publication bias for other cutoffs. Overall there was no strong evidence of publication bias.
https://doi.org/10.1016/j.mayocp.2017.11.012
3
MAYO CLINIC PROCEEDINGS
3884 Citations identified through database searching
61 Citations identified from gray literature and cross-references
2990 Citations excluded
3945 Records screened
912 Full-text articles excluded 69 Did not evaluate FeNO diagnostic accuracy 67 No comparison group 7 Not original study 13 Abstracts, systematic reviews 117 No population of interest 622 No outcome of interest 17 Other
955 Full text screened
43 Studies included in the systematic review
FIGURE 1. Study flowchart. FeNO ¼ fractional exhaled nitric oxide.
Diagnostic Accuracy of FeNO Testing In adults (aged 18 years or older), using FeNO cutoffs of less than 20, 20 to 29, 30 to 39, and 40 or more ppb, FeNO testing had sensitivities of 0.80 (95% CI, 0.72-0.86), 0.69 (95% CI, 0.57-0.79), 0.53 (95% CI, 0.35-0.70), and 0.41 (95% CI, 0.24-0.62), respectively, and specificities of 0.64 (95% CI, 0.46-0.79), 0.78 (95% CI, 0.66-0.86), 0.85 (95% CI, 0.77-0.90), and 0.93 (95% CI, 0.86-0.97), respectively (moderate quality of evidence). In children (aged 18 years), using FeNO cutoffs of less than 20 and 20 to 29 ppb, FeNO testing had sensitivities of 0.78 (95% CI, 0.59-0.90) and 0.61 (95% CI, 0.44-0.76), respectively, and specificities of 0.79 (95% CI, 0.55-0.92) and 0.89 (95% CI, 0.800.94), respectively (moderate quality of evidence). Diagnostic accuracy was overall higher in children than in adults. Overall, the posttest odds of having asthma given a positive FeNO test result increased by 2.80- to 7.00-fold. The Table presents a summary of the diagnostic accuracy of FeNO testing. Supplemental Figures 3 through 6 present HSROC curves for FeNO cutoffs of less than 20, 20 to 29, 30 to 39, and 40 or more ppb, respectively.
Subgroup and Sensitivity Analyses Supplemental Tables 3 through 6 (available online at http://www.mayoclinicproceedings.org) 4
Mayo Clin Proc.
n
list the findings for subgroup analyses. Health status of reference groups impacted the diagnostic accuracy of FeNO. The diagnostic accuracy of FeNO may be overestimated in studies that used healthy controls (DOR, 16.45; 95% CI, 7.85-34.49) compared with symptomatic controls (DOR, 4.42; 95% CI, 2.33-8.35) for FeNO cutoffs of less than 20 ppb. Risk of bias might partially explain heterogeneity in the diagnostic accuracy of FeNO, with greater reported diagnostic accuracy as the risk of bias increased (DORs across cutoffs of 11.29, 10.33, and 8.81 for high, medium, and low risk, respectively, for FeNO cutoffs of less than 20 ppb). With regard to specific reference tests (criterion standard), FeNO testing was found to have higher DORs compared with clinical diagnosis than when compared with positive results on bronchial challenge or the combination of clinical diagnosis, bronchial challenge, or bronchodilator response. We also found that diagnostic accuracy was higher in nonsmokers (sensitivity, 0.70; 95% CI, 0.61-0.78; specificity, 0.80; 95% CI, 0.74-0.85; DOR, 9.49; 95% CI, 5.62-16.01) than in smokers (DORs ranged from 2.53 to 5.4610,11). In a sensitivity analysis, we were only able to analyze studies that evaluated the diagnostic accuracy of FeNO in corticosteroid-naive asthmatics (the remaining studies had a mix of populations, including corticosteroid-naive XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org
FENO TESTING IN ASTHMA
Reference test
Flow and timing
Over-all
Low risk
Medium risk
High risk
Unclear risk
Overall risk of bias
14
20
All were patients included in the analysis
30
2
Index test
0
10
1
Patients received the same reference test
40
01 2
All patients received a reference test
40
02 1
Appropriate interval between index test and reference test
33
01
35
Concern that the target condition did not match the review question Reference test results interpreted without knowledge of the results of index test
38
Reference test correctly classifed the target condition
37
7
1
0 4 1 0
6
The cutoff used was prespecified
02
33
8 101
41
12 2
38
Concern that the included patients did not match the review question Study avoided inappropriate exclusions
31
1 4
Case-control design avoided
31
0
19
0
0 010
42
Index test results interpreted without knowledge of the results of the reference standard
Study enrolled consecutive or random sample of patients
9 0
Concern that the index test differed from the review question
Patient selection
9
15
7 12
0 9
FIGURE 2. Summary of risk of bias assessment using the QUADAS-2 tool (n¼43).
patients and corticosteroid users) (Supplemental Table 7, available online at http://www. mayoclinicproceedings.org). At cutoffs of less than 20 ppb, FeNO had the highest accuracy in this group of patients compared with patients in the main results (sensitivity, 0.79; 95% CI, 0.67-0.87; specificity, 0.77; 95% CI, 0.56-0.90; and DOR, 12.25; 95% CI, 5.73-26.21). Results in other cutoffs were different and inconsistent. In another sensitivity analysis, we analyzed only studies that evaluated the diagnostic accuracy of FeNO in asthmatic patients with atopy. The results, which included a small number of studies (4), revealed accuracy measures that were similar to those from the main analysis (sensitivity, 0.63; 95% CI, 0.43-0.80; specificity, 0.79; 95% CI, 0.65-0.89; and DOR, 6.67; 95% Mayo Clin Proc. n XXX 2017;nn(n):1-8 www.mayoclinicproceedings.org
n
CI, 1.59-27.95) (Supplemental Table 8, available online at http://www.mayoclinicproceedings. org). DISCUSSION We conducted a systematic review with metaanalyses to evaluate the diagnostic accuracy of FeNO testing in children and adults with suspected asthma. This is an important question because of the high prevalence of asthma and the risks associated with undiagnosed asthma as well as the increasingly recognized issue of asthma overdiagnosis and the risks of overtreatment.12,13 We found 43 observational studies addressing this question. Most had low or medium risk of bias. The FeNO concentration has moderate diagnostic
https://doi.org/10.1016/j.mayocp.2017.11.012
5
MAYO CLINIC PROCEEDINGS
TABLE. Summary of Diagnostic Accuracy of FeNO Testing for Suspected Asthma FeNO cutoff <20 ppb
Subgroup
Conclusion
18 Years or older
Sensitivity: 0.80; 95% CI, 0.72-0.86 Specificity: 0.64; 95% CI, 0.46-0.79 DOR: 7.28; 95% CI, 4.04-13.11 LRþ: 2.25; 95% CI, 1.47-3.44 LR: 0.31; 95% CI, 0.23-0.41 Sensitivity: 0.78; 95% CI, 0.59-0.90 Specificity: 0.79; 95% CI, 0.55-0.92 DOR: 13.44; 95% CI, 3.56-50.71 LRþ: 3.76; 95% CI, 1.53-9.26 LR: 0.28; 95% CI, 0.14-0.56 Sensitivity: 0.69; 95% CI, 0.57-0.79 Specificity: 0.78; 95% CI, 0.66-0.86 DOR: 7.70; 95% CI, 4.12-14.40 LRþ: 3.10; 95% CI, 2.02-4.74 LR: 0.40; 95% CI, 0.29-0.56 Sensitivity: 0.61; 95% CI, 0.44-0.76 Specificity: 0.89; 95% CI, 0.80-0.94 DOR: 12.13; 95% CI, 5.98-24.63 LRþ: 5.34; 95% CI, 3.12-9.14 LR: 0.44; 95% CI, 0.30-0.65 Sensitivity: 0.53; 95% CI, 0.35-0.70 Specificity: 0.85; 95% CI, 0.77-0.90 DOR: 6.27; 95% CI, 3.70-10.64 LRþ: 3.46; 95% CI, 2.57-4.66 LR: 0.55; 95% CI, 0.39-0.78 NA Sensitivity: 0.41; 95% CI, 0.24-0.62 Specificity: 0.93; 95% CI, 0.86-0.97 DOR: 9.84; 95% CI, 5.46-17.75 LRþ: 6.18; 95% CI, 3.64-10.47 LR: 0.63; 95% CI, 0.47-0.85 NA
Younger than 18 years
20-29 ppb
18 Years or older
Younger than 18 years
30-39 ppb
18 Years or older
40 ppb
Younger than 18 years 18 Years or older
Younger than 18 years
DOR ¼ diagnostic odds ratio; FeNO ¼ fractional exhaled nitric oxide; LRþ ¼ likelihood ratio for a positive test result; LR ¼ likelihood ratio for a negative test result; ppb ¼ parts per billion; NA ¼ not available.
accuracy for asthma with DORs that range from 5.85 to 16.95 across various cutoff points (in comparison, a test with 0.80 sensitivity and 0.80 specificity would have a DOR of 16). As expected, with increasing cutoff values, FeNO had a gradual decrease in sensitivity and improved specificity (for cutoffs of <20, 20-29, 30-39, and 40 ppb, FeNO testing has sensitivities of 0.78, 0.63, 0.56, and 0.41, respectively, and specificities of 0.71, 0.81, 0.84, and 0.94, respectively). Therefore, knowing the cutoffs used for test interpretation is critical for interpretation by clinicians. The 2011 American Thoracic 6
Mayo Clin Proc.
n
Society guidelines14 have recommended cutoffs of less than 25 ppb (<20 ppb in children) for low FeNO and greater than 50 ppb (>35 ppb in children) for high FeNO as the most useful cutoffs to make the determination that eosinophlic inflammation and response to corticosteroids would be unlikely and likely, respectively. Intermediate results should be interpreted with caution (FeNO of 25-50 ppb [20-35 ppb in children]).14 Inferences from several preplanned subgroup analyses were limited because of the limited number of studies and the heterogeneity of populations, interventions, and outcomes, particularly regarding the impacts of reference test, the presence of atopy, current use of ICS, and smoking status on FeNO diagnostic performance. Asthma can sometimes be difficult to diagnose, and FeNO can be helpfuldmore so in ruling in than ruling out asthmadin making therapeutic decisions more evidence based. For example, in a patient who has compatible symptoms and is clearly atopic (eg, history of eczema, positive skin test results, peripheral blood eosinophilia, or elevated serum IgE level), elevated FeNO implies a diagnosis of asthma and that treatment with ICS is indicated. Conversely, low FeNO (eg, <20 ppb) implies that these compatible symptoms are not likely due to asthma. Thus, clinicians should direct treatment to other causes of symptoms. The caveat is that low FeNO findings do not exclude asthma, but clinical judgment and vigilance are warranted. Such a scenario is common in pediatric practice, in which this review already concluded that evidence was stronger in children and adolescents, likely because most asthma in this population is atopic. It is precisely in this population that overdiagnosis of asthma is a bigger clinical problem than underdiagnosis. In this case, patients with suspected asthma should have further testing to confirm the diagnosis. The first test should be spirometry with an assessment of bronchodilator response. If this test does not confirm the diagnosis but the index of suspicion for asthma is still high, further testing should be performed. Measurement of FeNO should be considered at this stage as a second-line test (other second-line tests would include the various bronchoprovocation studies). XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org
FENO TESTING IN ASTHMA
One challenge relates to the fact that there is no true criterion standard for diagnosing asthma. Although we did not label studies as having increased risk of bias because of this issue, we recognize that it can impact diagnostic accuracy. In addition, a wide range of reference tests were reported. We categorized these reference tests as clinical diagnosis, positive bronchial challenge test results, or a combination of clinical diagnosis, positive results on bronchial challenge, and/or bronchodilator response. However, considerable heterogeneity still exists, such as to how and when these tests were administered. Although categorizations of less than 20, 20 to 29, 30 to 39, and 40 or more ppb helped reduce heterogeneity and facilitated meta-analyses, we were not able to definitively present a best cutoff overall or within each category. We were also not able to conduct some planned subgroup analyses because of lack of data, including asthma phenotype, body mass index, and exhalation flow rate. CONCLUSION The FeNO concentration has moderate accuracy to diagnose asthma in individuals aged 5 years and older. Test performance is modestly better in corticosteroid-naive asthmatics, children, and nonsmokers than in the general population with suspected asthma. ACKNOWLEDGMENTS We gratefully acknowledge the following individuals for their contributions to this study: the task order officers from Agency for Healthcare Research and Quality; Anas Daghestani, MD, Austin Regional Clinic; Serpil C. Erzurum, MD, Cleveland Clinic Foundation; Elliot Israel, MD, Brigham and Women’s Hospital; Rohit Katial, MD, University of Colorado; Nicholas Kenyon, MD, University of California, Davis; Todd A. Mahr, MD, Gundersen Lutheran Medical Center; John G. Mastronarde, MD, MSc, Providence Portland Medical Center; Wayne J. Morgan, MD, University of Arizona College of Medicine; Shahzeb Munir, MD, University of Wisconsin School of Medicine and Public Health; Augusto Litonjua, MD, MPH, Harvard Medical School; John Sundy, MD, PhD, Duke University School of Medicine; and Sally Wenzel, MD, University of Pittsburgh Medical Center. Mayo Clin Proc. n XXX 2017;nn(n):1-8 www.mayoclinicproceedings.org
n
The funding source had no role in conduct of the review; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the submitted manuscript; and decision to submit the manuscript for publication. The findings and conclusions in this article are those of the authors, who are responsible for its contents, and do not necessarily represent the views of the Agency for Healthcare Research and Quality or the US government. Therefore, no statement in this article should be construed as the official position of the Agency for Healthcare Research and Quality. Author contributions: All authors contributed to the acquisition, analysis, or interpretation of data for the study and critical revision of the submitted manuscript for important intellectual content. Concept and design: Drs Wang, Pianosi, Keogh, and Murad; drafting of the manuscript: Drs Wang, Pianosi, Keogh, and Murad; statistical analysis: Drs Wang and Murad; administrative, technical, or material support: Dr Alsawas. As the guarantors of the study, Drs Wang and Murad take full responsibility for the work as a whole, including the study design, access to data, and the decision to submit and publish the manuscript. SUPPLEMENTAL ONLINE MATERIAL Supplemental material can be found online at http://www.mayoclinicproceedings.org. Supplemental material attached to journal articles has not been edited, and the authors take responsibility for the accuracy of all data. Abbreviations and Acronyms: DOR = diagnostic odds ratio; FeNO = fractional exhaled nitric oxide; HSROC = hierarchical summary receiver operating characteristic; ICS = inhaled corticosteroids; ppb = parts per billion Affiliations (Continued from the first page of this article.): (Z.W.), Division of Pediatric Pulmonology, Department of Pediatric and Adolescent Medicine (P.T.P.), Division of Pulmonary and Critical Care Medicine (K.A.K.), and Mayo Clinic Libraries (L.J.P.), Mayo Clinic, Rochester, MN; and Paediatric Respirology, King’s College Hospital NHS Foundation Trust, London, UK (P.T.P.). Grant Support: This work was funded by the Agency for Healthcare Research and Quality under contract HHSA 290201500013I to support evidence synthesis as part of the Evidence-based Practice Center Program.
https://doi.org/10.1016/j.mayocp.2017.11.012
7
MAYO CLINIC PROCEEDINGS
Correspondence: Address to Mohammad Hassan Murad, MD, MPH, Mayo Clinic Evidence-based Practice Center, Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, 200 First St SW, Rochester, MN 55905 ([email protected]).
REFERENCES 1. National Asthma Education and Prevention Program (National Heart, Lung, and Blood Institute) Third Expert Panel on the Management of Asthma. Expert Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of Asthma. Bethesda, MD: US Dept of Health and Human Services, National Institutes of Health; 2007. NIH publication 07-4051. 2. Centers for Disease Control and Prevention. Most recent asthma data. Centers for Disease Control and Prevention website. https://www.cdc.gov/asthma/most_recent_data.htm. Updated June 2017. Accessed December 4, 2017. 3. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and metaanalyses: the PRISMA statement. BMJ. 2009;339:b2535. 4. Whiting PF, Rutjes AW, Westwood ME, et al; QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536. 5. Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol. 2003;56(11):1129-1135. 6. Rutter CM, Gatsonis CA. A hierarchical regression approach to meta-analysis of diagnostic test accuracy evaluations. Stat Med. 2001;20(19):2865-2884.
8
Mayo Clin Proc.
n
7. Schünemann HJ, Mustafa R, Brozek J, et al; GRADE Working Group. GRADE Guidelines: 16. GRADE evidence to decision frameworks for tests in clinical practice and public health. J Clin Epidemiol. 2016;76:89-98. 8. Mustafa RA, Santesso N, Khatib R, et al. Systematic reviews and meta-analyses of the accuracy of HPV tests, visual inspection with acetic acid, cytology, and colposcopy. Int J Gynaecol Obstet. 2016;132(3):259-265. 9. Avital A, Uwyyed K, Berkman N, Godfrey S, Bar-Yishay E, Springer C. Exhaled nitric oxide and asthma in young children. Pediatr Pulmonol. 2001;32(4):308-313. 10. Matsunaga K, Hirano T, Akamatsu K, et al. Exhaled nitric oxide cutoff values for asthma diagnosis according to rhinitis and smoking status in Japanese subjects. Allergol Int. 2011;60(3): 331-337. 11. Malinovschi A, Backer V, Harving H, Porsbjerg C. The value of exhaled nitric oxide to identify asthma in smoking patients with asthma-like symptoms. Respir Med. 2012; 106(6):794-801. 12. Looijmans-van den Akker I, van Luijn K, Verheij T. Overdiagnosis of asthma in children in primary care: a retrospective analysis. Br J Gen Pract. 2016;66(644):e152-e157. 13. Aaron SD, Vandemheen KL, FitzGerald JM, et al; Canadian Respiratory Research Network. Reevaluation of diagnosis in adults with physician-diagnosed asthma. JAMA. 2017;317(3): 269-279. 14. Dweik RA, Boggs PB, Erzurum SC, et al; American Thoracic Society Committee on Interpretation of Exhaled Nitric Oxide Levels (FENO) for Clinical Applications. An official ATS clinical practice guideline: interpretation of exhaled nitric oxide levels (FENO) for clinical applications. Am J Respir Crit Care Med. 2011;184(5):602-615.
XXX 2017;nn(n):1-8
n
https://doi.org/10.1016/j.mayocp.2017.11.012 www.mayoclinicproceedings.org