Antibiotics are frequently prescribed to children with asthma for non-specific upper respiratory tract infections (NS-URTIs), but it is unclear whether such prescriptions affect the risk of subsequent asthma exacerbation. The authors aimed to determine whether antibiotic use (narrow- or broad-spectrum) initiated within 3 days of an outpatient visit for NS-URTI is associated with exacerbation risk during the 6 months following an initial induction period.
The investigation used a sequential target trial emulation approach applied to Quebec administrative health and drug claims data. The design emulated a series of hypothetical randomized trials at each eligible outpatient visit for an NS-URTI among children with pre-existing asthma.
Children aged 0–17 years with asthma who attended an outpatient visit for a NS-URTI were assessed for eligibility at each potential index visit. Exclusion criteria applied relative to the index visit included chronic lung disease, chronic lung infection, acute infection or asthma exacerbation in the 30 days prior to the visit, lack of public drug insurance, and prior antibiotic use within the preceding 6 months. Eligible visits could re-enter subsequent target trials if criteria were again met.
Treatment strategies were classified according to drug claims within a 3-day treatment ascertainment window after the index visit: no antibiotic (reference), narrow-spectrum antibiotic, or broad-spectrum antibiotic. The analysis compared these spectrum-specific treatment strategies.
The primary outcome was an asthma exacerbation occurring during the 6 months after a 10-day induction period. Exacerbation was defined as oral corticosteroid use, an emergency department visit for asthma, or hospitalization for asthma. To reduce reverse causation and allow for biological latency, exacerbation events occurring within the first 10 days of follow-up were excluded from the primary analysis. The outcome assessment period therefore began 14 days after the index visit (T0) and continued for 182 days.
A secondary outcome window was also specified for events occurring shortly after the index visit; for the secondary outcome, follow-up began at T0+4 days until 13 days after the index visit (inclusive).
Risk ratios (RRs) and risk differences (RDs) were estimated using inverse probability of treatment weighted log-binomial models. Eligible visits were paired by season and calendar year through propensity score model stratification. Subgroup analyses were planned by age (<5 years vs ≥5 years), sex, use of asthma controller therapy, short-acting beta-agonist (SABA) use (≤2 vs >2 doses), and type of provider (specialist vs general practitioner).
The analytic sample included 13,646 children contributing 24,139 person-trials. Among these person-trials, 8.8% received narrow-spectrum antibiotics and 11.4% received broad-spectrum antibiotics within the 3-day treatment window after the NS-URTI visit.
Compared with no antibiotics, the estimated RR for asthma exacerbation associated with narrow-spectrum antibiotics was 1.26 (95% CI, 0.99–1.53) with an RD of 0.07%. For broad-spectrum antibiotics versus no antibiotics, the RR was 1.28 (95% CI, 1.03–1.54) with an RD of 0.08%.
The authors reported that broad-spectrum antibiotics were associated with a modestly higher risk of exacerbation compared with no antibiotics. However, the absolute risk differences were very small (less than 0.1%).
The study protocol described subgroup analyses by age, sex, controller therapy use, SABA use, and provider type. The abstract and main summary report that subgroup analyses were conducted; specific subgroup effect estimates and statistical details beyond what is summarized above are reported in the full text and figures. If additional numerical subgroup results are required, they were not fully reported in the abstract-level summary provided here.
In this target trial emulation of children with asthma presenting with an NS-URTI, initiation of broad-spectrum antibiotics within 3 days of the visit was associated with a modestly increased relative risk of asthma exacerbation over the subsequent 6 months, whereas narrow-spectrum antibiotics showed an elevated point estimate but a confidence interval that included the null. Importantly, the reported absolute risk differences were very small (<0.1%), indicating that although relative risks were modestly increased, the excess absolute risk attributable to antibiotic treatment in this population was minimal.
Clinicians should weigh the small absolute change in exacerbation risk against other considerations when deciding on antibiotic treatment for NS-URTIs in children with asthma, including the likelihood of bacterial infection, antimicrobial stewardship, and individual patient factors.
Author disclosures and funding are reported. Two authors (Carmen Cañete Ramirez and Albina Tskhay) received graduate scholarships from Fonds de Recherche Québec-Santé and CHU Sainte-Justine and declared no conflicts. One author (Cristina Longo) discloses salary support from multiple research agencies and foundations and declared no conflicts. Lucie Blais discloses grants, personal fees, and research contracts from AstraZeneca and grants from Teva outside the submitted work. Other authors reported no conflicts. These disclosures were stated in the source material.
Note: This summary reflects the abstract and front-matter information available in the source. Detailed methods, complete subgroup estimates, and additional figures are available in the full article; where specific numerical details were not reported in the abstract, the summary indicates that those details are reported in the full text.