ABSTRACT
Background Concerns that antimicrobial treatment may foster selection and transmission of resistant bacterial lineages have led to conflicting guidelines for clinical management of common non-severe infections. However, the impact of antimicrobial treatment on colonization dynamics is poorly understood. We used data from a previously-conducted trial of amoxicillin-clavulanate therapy for acute otitis media (AOM) to understand how antimicrobial treatment impacts the acquisition and clearance of Streptococcus pneumoniae lineages with varying susceptibility to penicillin.
Methods and findings We measured impacts of antimicrobial treatment on nasopharyngeal carriage of penicillin-susceptible S. pneumoniae (PSSP) and penicillin–non-susceptible S. pneumoniae (PNSP) lineages at end-of-treatment and 15d, 30d, and 60d after treatment in a previously-conducted randomized, double-blind, placebo-controlled trial. Analyses were not specified in the original protocol. Among children 6-35 months of age with stringently-defined AOM, 162 were assigned amoxicillin-clavulanate, and 160 were assigned placebo. Children who did not show clinical improvement received open-label antimicrobial rescue treatment with amoxicillin-clavulanate irrespective of the randomized treatment assignment, to which both patients and physicians were blinded. The intention-to-treat populations of the intervention and placebo arms thus received care resembling immediate antimicrobial therapy and watchful waiting, respectively. Immediate amoxicillin-clavulanate reduced PSSP carriage prevalence by 88% (95%CI: 76-96%) at end-of-treatment and by 27% (–3-49%) after 60d, but did not measurably alter PNSP carriage prevalence throughout follow-up. By end-of-treatment, 7% of children who carried PSSP at enrollment remained colonized in the amoxicillin-clavulanate arm, compared to 61% of PSSP carriers who received placebo; differences in carriage prevalence persisted at least 60d after treatment among children who carried PSSP at enrollment. Among children not carrying pneumococci at enrollment, amoxicillin-clavulanate reduced PSSP acquisition by >80% over 15d. Among children who carried PNSP at enrollment, no differences in carriage prevalence of S. pneumoniae, PSSP, or PNSP were detected at follow-up visits.
Conclusions In a setting with low PNSP prevalence, antimicrobial therapy for AOM conferred a selective impact on colonizing S. pneumoniae by accelerating clearance, and delaying acquisition, of penicillin-susceptible lineages. Absolute risk of carrying PNSP was unaffected by treatment (ClinicalTrials.gov: NCT00299455; Funding: NIH/NIGMS).