Accuracy of objective tests for diagnosing adult asthma in symptomatic patients: A systematic literature review and hierarchical Bayesian latent-class meta-analysis

Background: We obtain summary estimates of the accuracy of additional objective tests for the diagnosis of adult asthma using systematic review and meta-analysis of diagnostic test accuracy studies. Methods: Medline, Embase, and other relevant electronic databases were searched for papers published between January 1989 and December 2016. Studies were included if they evaluated the diagnostic accuracy of objective tests, including airway reversibility (AR), airway hyperresponsiveness (AHR), and fractionated exhaled nitric oxide (FeNO) for the diagnosis of adult asthma in patients with symptoms suggestive of asthma. If papers were assessed appropriate using the adapted QUADAS-2 tool, meta-analysis was conducted using the hierarchical bivariate model. This hierarchical model accounts for both within and between study variability. Results: Sixteen studies reported the performance of the evaluated objective tests at presentation. For diagnosis of adult asthma, overall sensitivity and specificity for AR were 0.39 (95% confidence interval [CI] 0.18 to 0.66) and 0.95 (95% CI 0.86 to 1.00); for AHR, 0.86 (95% CI 0.61 to 1.00) and 0.95 (95% CI 0.77 to 1.00); for FeNO, 0.65 (95% CI 0.53 to 0.77) and 0.83 (95% CI 0.75 to 0.90). Comprehensive comparison of three diagnostic tools for adult asthma using the back-calculated likelihood rate (LR) showed that AR and AHR corresponded to a higher LR+, and AHR gave a lower LR-. Conclusions: In the current situation of no gold standard for diagnosis of adult asthma, AR and AHR are appropriate for ruling-in the true diagnosis, and AHR is superior for ruling-out a diagnosis. Since each objective test had a specific characteristic, it should be chosen depending on the situation, such as the capacity of the institution and the conditions of patients. Copyright (C) 2018, Japanese Society of Allergology. Production and hosting by Elsevier B.V.