Severe bronchial asthma: probability of frequent exacerbations considering occupational exposure

Bronchial asthma remains one of the most common chronic inflammatory diseases of the airways and represents a significant medical and social problem due to highly frequent exacerbations, hospitalizations, and declining quality of life. Severe asthma is a particularly challenging clinical condition, in which frequent exacerbations (three or more episodes per year) place a major burden on the healthcare system and are associated with adverse outcomes. Therefore, development of scales for assessing the probability of frequent exacerbations based on accessible clinical, functional, and occupational parameters is a relevant task foridentifying patient groups requiring more detailed examination and follow-up by specialized physicians. The aim of this study was to develop a scale for assessing the probability of frequent exacerbations in patients with severe bronchial asthma. We analyzed the examination results of 174 patients with severe bronchial asthma aged 18 to 71 years. The modeling outcome was defined as “frequent exacerbations” as three or more episodes during the previous year. A hybrid machine-learning algorithm was used to construct the model: gradient boosting was applied to select the most significant predictors. The list of predictors included sex, age, duration of bronchial asthma, occupational exposure, body mass index, gastroesophageal reflux disease, eosinophil level, and FEV1. Model validation was performed using 5-fold stratified cross-validation. The key factors associated with the probability of frequent exacerbations were occupational exposure to allergens orirritants in the workplace, reduced pulmonary function, and presence of comorbid pathology. Stratification by tertiles of predicted probability made it possible to identify groups with low, moderate, and high probability of frequent exacerbations. According to the developed scale, the combination of FEV1 < 40 %, gastroesophageal reflux disease, occupational exposure, and a higheosinophil level increased the probability of frequent exacerbations up to 97 %. The results were visualized as a heat map. We proposed a scale for assessing the probability of frequent exacerbations in patients with severe bronchial asthma, intended primarily for primary care physicians, including therapists and general practitioners. Its application is aimed at timely identification of patients with moderate and high probability of severe bronchial asthma exacerbations for subsequent referral to specialized physicians, including allergists-immunologists and pulmonologists. This approach may contribute to reducing hospitalization rates and may have significant pharmacoeconomic relevance.