Background: Exacerbations are a hallmark of chronic obstructive pulmonary disease (COPD). Evidence suggests the presence of substantial between-individual variability (heterogeneity) in exacerbation rate. Whether individuals vary in tendency towards experiencing severe (versus mild) exacerbations, or whether there is an association between exacerbation rate and severity, is not yet studied.

Methods: We used data from a one-year randomized trial of azithromycin for prevention of COPD exacerbations (the MACRO study, 2006–2010, N=1,107, mean age 65.2, 59.1% male) to estimate the joint heterogeneity in exacerbation rate and severity and to examine whether frequent exacerbators experience a higher proportion of severe exacerbations. A parametric frailty model was combined with a logistic regression model, with bivariate random-effects capturing heterogeneity in rate and severity.

Results: The average rate of exacerbation was 1.53/year, with 95% of subjects having model-estimated rate of 0.47 to 4.22. The overall ratio of severe to total exacerbations was 0.22, with 95% of subjects having model-estimated ratio of 0.04 to 0.60. We did not confirm an association between exacerbations rate and severity (P=0.099).

Conclusion: A unified model, implemented in standard software, could estimate joint heterogeneity in exacerbation rate and severity and can have applications in similar contexts where inference on event time and intensity are considered.