Abstract
BACKGROUND: Allergic bronchopulmonary aspergillosis (ABPA) has been reported in various degrees among patients with persistent allergic asthma (PAA). Currently, there is no gold standard approach for diagnosis of ABPA. OBJECTIVES: In the current study, we aimed the evaluation of three different mainly used algorithms as Rosenberg & Patterson (A), ISHAM Working Group (B) and Greenberger (C) for diagnosis of ABPA in 200 patients with underlying PAA. METHODS: All patients were evaluated using Aspergillus skin prick test (SPTAf), Aspergillus-specific IgE (sIgEAf) and IgG (sIgGAf), total IgE (tIgE), pulmonary function tests, radiological findings and peripheral blood eosinophil count. The prevalence rate of ABPA in PAA patients was estimated by three diagnostic criteria. We used Latent Class Analysis for the evaluation of different diagnostic parameters in different applied ABPA diagnostic algorithms. RESULTS: Aspergillus sensitisation was observed in 30 (15.0%) patients. According to algorithms A, B and C, nine (4.5%), six (3.0%) and 11 (5.5%) of patients were diagnosed with ABPA, respectively. The sensitivity and specificity of criteria B and C were (55.6% and 99.5%) and (100.0% and 98.9%) respectively. sIgEAf and sIgGAf showed the high significant sensitivity. The performance of algorithm A, in terms of sensitivity and specificity, was somewhat better than algorithm B. CONCLUSION: Our study demonstrated that the sensitivity of different diagnostic algorithms could change the prevalence rate of ABPA. We also found that all of three criteria resulted an adequate specificity for ABPA diagnosis. A consensus patterns combining elements of all three criteria may warrant a better diagnostic algorithm.