Risk Factors associated with frequent exacerbations in asthma

Background: Asthma is a chronic airway inflammatory disease with various degrees of severity. Exacerbations are
commonly seen in uncontrolled asthma and their treatment involves oral corticosteroids use with a lot of side
effects.
Objective: The aim of the study was to identify easily available predictors for future exacerbations in patients with
asthma.
Methods: This is a prospective study on 250 consecutive patients with asthma with a successful sputum induction.
Exacerbation rate in the following year was assessed by telephone interview. Logistic regression was used to test
the relationship between the binary outcomes (<1 or ≥1 exacerbation, <2 or ≥2 exacerbations) and a set of
covariates including demographic, clinical, functional and inflammatory characteristics such as FeNO, sputum
and blood cell counts. The results were then applied and validated in a new cohort of 1450 patients.
Results: Sputum and blood eosinophils were able to identify patients presenting ≥1 or ≥2 exacerbations with the
same discriminative power (AUC:0.65 and 0.64 respectively). The multiple regression analysis identified that
exacerbations in the previous year (OR = 9.3), treatment with high doses ICS (OR = 27.1), blood eosinophils
(cells/mm3, OR = 1.8) and FEV1/FVC (OR = 0.93) were independent predictors of exacerbations in the year
following the visit with an AUC of 0.93 for this model. Frequent exacerbations (≥2) were also predicted by
exacerbations in the previous year (OR = 10.5), treatment with high doses ICS (OR = 39.2) and blood eosinophils
(OR = 3.5) with an AUC of 0.95 for the model.
Conclusion: Blood and sputum eosinophils have similar predictive value for future exacerbations. Prediction
could be improved by combining this information with lung function, ICS dose and history of previous
exacerbations.