Infectious Disease
Effect of leukotriene-modifying agents on severe influenza illness: Comparative analysis of three analytical approaches Brittney M. Snyder* Brittney Snyder Snyder Snyder Snyder Snyder Snyder Snyder Snyder Snyder Snyder Snyder brittney.m.snyder@vumc.org
Animal and in vitro studies suggest that leukotriene modifying agents (LMAs) may decrease susceptibility to influenza illness. We estimated the effect of LMAs on severe human influenza illness in a retrospective cohort using three analytic designs and corresponding methods to mitigate potential biases: 1) a marginal structural model, 2) a proportional hazards model, and 3) a case-time-control design. Study populations included individuals with asthma and/or allergic rhinitis for whom LMAs are FDA approved and who were enrolled in Tennessee Medicaid (TennCare) or Department of Defense Military Health System (DoD MHS) from 1995–2019. Exposed periods were defined by LMA prescription start dates plus days’ supply. Severe influenza illness was defined using previously validated International Classification of Diseases criteria. Montelukast accounted for 99% of LMA prescription fills. Adjusted incidence rate ratios from the marginal structural model and adjusted hazard ratios from the proportional hazards model were 1.26 (95% CI 0.99, 1.59) and 1.14 (95% CI 0.92, 1.41) for TennCare and 1.01 (95% CI 0.84, 1.21) and 0.97 (95% CI 0.82, 1.16) for DoD MHS, respectively. Adjusted odds ratios from the case-time-control were 1.11 (95% CI 0.74, 1.65) for TennCare and 1.91 (95% CI 1.29, 2.84) for DoD MHS. Findings in different populations and designs do not support the use of LMAs for chemoprophylaxis of severe influenza illness. Conclusions regarding LMAs other than montelukast were not possible.
