A novel computerized phenotype-oriented algorithm for asthma diagnostics
By Dennis Lo
Physicians in the US currently rely on two guidelines for asthma diagnostics: the EPR-3 and the GINA report. Due to the ten year difference in publication time, a comparison between both guidelines is necessary. Additionally, while the guidelines include defined parameters for impairment factors, patient-specific risk factors remain unparameterized. By parameterizing the risk factors and following a phenotype-oriented diagnostic approach, clinicians may be may be able to improve asthma diagnostics with respect to speed and accuracy consistency. Computerized algorithms for the EPR-3 and GINA report were coded in Java and compared with respect to therapy recommendations for mild, moderate, and severe asthmatics. A systematic literature review and meta-synthesis was performed in the PubMed database to determine parameters for asthma phenotype categorization. A computerized algorithm was coded based on determined parameters. and compared to the EPR-3- and GINA report-based algorithms with respect to therapy recommendation for asthmatics of five established phenotypes. The GINA reports recommendation of specialist assessment for severe asthma supported the emphasis on asthma as a heterogeneous disease since publication of the EPR-3. The systematic-review-based algorithm recommended targeted therapy such as anti-IL-5 for the late-onset eosinophilic phenotype when compared to the EPR-3 and GINA algorithms recommended inhaled corticosteroids with long-acting agonist, suggesting the potential of the phenotype-oriented approach for personalizing clinical decisions.