B45-02 Identifiable Clinical Factors to Predict Bronchiectasis Exacerbations

Abstract Introduction Bronchiectasis, a once orphan disease, has been growing in prevalence due to the expansion of available cross-sectional chest imaging as well as a greater understanding among clinicians. To that end, it becomes important that means to prevent bronchiectasis exacerbations, which portend for morbidity and mortality, are implemented more broadly as well. Modern medicine provides an ample number of readily available, standard-of-care clinical biomarkers that may grant insight into future exacerbation risk. In this small study, we attempt to discover potentially predictive aspects of a bronchiectasis patient encounter that may predict future exacerbation. Methods Participants were drawn from the bronchiectasis clinic at the University of Cincinnati Bronchiectasis Center for either the most recent clinic visit or the visit immediately preceding a severe (inpatient) exacerbation. Patient charts were investigated for demographic, spirometric, microbiological, and laboratory data. We compared exacerbators and non-exacerbators using bivariable approaches, chi2 for categorical variables and two-sided Student’s T-testing for continuous variables with standard deviations presented. Variables found to be statistically significant between groups were then fit into a logistic regression model. Variables with attenuated significance were removed from the model to generate a parsimonious, predictive model. Results The cohort was 80% female, 90% white, and had an average age of 68 years. There was no difference in age, race, sex, spirometry values, smoking status, mucous plugging on CT scan, pectoralis major thickness, or medication use between exacerbators and non-exacerbators. Exacerbators (88% vs 45%) reported subjective mucous hypersecretion/chest congestion more frequently (p = 0.043), and had a trend toward being more likely to be colonized with Pseudomonas (33% vs 19%; p = 0.07). In logistic regression containing sex, race, age, subjective mucous hyper secretion, and Pseudomonas colonization, all variables were attenuated with subjective mucous hypersecretion trending toward significance (OR 11.68; 95% CI 0.80-168.85; p = 0.07). Conclusion This very small study suggests some trends in readily available clinical variables that may predict future bronchiectasis exacerbations. Data collection is ongoing and the next steps will enlarge the cohort, limiting variability in statistical calculation with a goal of generating a predictive scoring system that may be tested for clinical implementation. This abstract is funded by: None