Controller Medications and Serious Early Childhood Lower Respiratory Tract Illnesses

JOHN WATSON, MD
University of Colorado Anschutz Medical Campus, Children’s Hospital Colorado

While asthma controllers (inhaled corticosteroids and leukotriene inhibitors) have been shown to reduce exacerbations in children 2-5 years old with asthma or recurrent wheeze, deciding when to prescribe controllers in children <2 years old remains challenging given the substantial clinical overlap between asthma and lower respiratory tract infections (LRTIs). Our objective was to assess the association between time on controller medications and emergency department (ED) and inpatient (IP) visits for LRTI or asthma in children <2 years old.

Children <2 years old with at least one prior LRTI (bronchiolitis and pneumonia) were identified in the 2009-2017 Colorado All Payer Claims Database. Those with complex chronic conditions, a diagnosis of asthma prior to first LRTI, or controller use prior to first LRTI were excluded.

Retrospective cohort study using administrative claims data. The primary exposure variable was a time-dependent indicator for presence of a prescription for controller medication. The primary outcome was count of ED/IP visits for any diagnosis of LRTI or ED/IP visit for a primary diagnosis of asthma after the first LRTI (using ICD9/10 codes). A Poisson regression model accounting for correlation within subjects was used. Adjusted models included baseline covariates for gender, insurance type, prematurity, family history of asthma claim, and time-dependent covariates of prior wheeze claim, atopy claim, number of LRTI visits, LRTI hospitalizations, prior subspecialty claim (Allergy/Immunology or Pulmonology), and outpatient beta agonist prescriptions.

We identified 40,473 children meeting inclusion criteria, ultimately constituting 547,082 person-months. A larger percentage of person-time on controller compared to off controller was seen with older age, male gender, Medicaid insurance, family history of asthma claim, prior atopy claim, prior wheeze claim, more prior LRTI visits, and prior outpatient beta agonist prescription. Controller medication use was not significantly associated with a reduction in ED/IP visits for LRTI or asthma in the adjusted model (RR 0.77; 95% CI: 0.57, 1.05).

In children under 2 with LRTI, controllers are more often prescribed in those who have more risk factors for future asthma. However, we found that time on controllers did not statistically reduce ED/IP visits for related respiratory diagnoses in this age group, potentially indicating an area for increased prescription stewardship.