When a regulatory agency attaches its strongest safety label to a widely used medication, the assumption is that prescribing behavior changes meaningfully. For montelukast — a leukotriene receptor antagonist taken by hundreds of thousands of asthma patients — that assumption deserves scrutiny, particularly given the drug's neuropsychiatric risk profile and its heavy use in children.
This interrupted time series analysis drew on a national commercial insurance claims database tracking monthly cohorts of over 590,000 asthma patients aged six and older from October 2017 through December 2022. The exposure of interest was the FDA's March 2020 boxed warning — its most serious safety designation — flagging neuropsychiatric adverse events including suicidal ideation, depression, and behavioral changes. Using segmented ordinary least-squares regression with Newey-West standard errors, researchers evaluated changes in both the monthly incidence and prevalence of montelukast use before and after the warning. Results were stratified by age, inhaled corticosteroid (ICS) use, and prior asthma-related emergency department visits or hospitalizations — allowing a more granular view of which patient subgroups drove or resisted any prescribing shifts.
The findings fit a familiar pattern in drug-safety communication research: regulatory warnings produce statistically detectable but clinically modest reductions in prescribing. Montelukast is particularly sticky as a therapeutic option because it offers oral administration and a well-established efficacy profile in both allergic asthma and exercise-induced bronchospasm — attributes that make physicians and patients reluctant to switch. Prior work on boxed warnings for antidepressants in adolescents and anticoagulants similarly showed partial, rather than dramatic, prescribing adjustments. The pediatric subgroup result here is especially consequential, as children represent the population at greatest theoretical neuropsychiatric risk and the one where guideline bodies have increasingly pushed for alternative or add-on approaches. Key limitations include the commercial insurance cohort, which excludes Medicaid and uninsured populations where asthma burden is often highest, and the observational design, which cannot confirm whether prescribers explicitly cited the warning as their rationale. This is a solid real-world pharmacovigilance study — confirmatory rather than paradigm-shifting — with direct relevance to how regulators and clinicians assess the downstream impact of safety labeling.