Increasing non-susceptibility to antibiotics within carried pneumococcal serotypes - Alaska, 2008-2015

In Alaska, while introduction of 13-valent pneumococcal conjugate vaccine led to declines in invasive pneumococcal disease, carriage prevalence remained stable because of replacement with non-vaccine serotypes. We assessed antibiotic non-susceptibility of carried pneumococci during serotype redistribution, determined the contributions of within-serotype shifts, and assessed factors that could explain changes in non-susceptibility.

Each year from 2008 to 2015, at multiple sites in Alaska, we collected nasopharyngeal swabs and completed surveys for a convenience sample of participants. Pneumococcal serotyping and antimicrobial susceptibility testing for penicillin and erythromycin were performed. We described changes in non-susceptibility of isolates from 2008-2011 to 2012-2015, and assessed the contributions of serotype redistribution and within-serotype changes in non-susceptibility by comparing observed data to modeled data removing either factor. We used weighted logistic regression to assess whether reported risk factors could explain changes over time in non-susceptibility within serotypes.

From 2008-2011 to 2012-2015, the overall proportion of isolates non-susceptible to penicillin or erythromycin increased by 3%, from 23% (n = 1,183) to 26% (n = 1,589; P < 0.05). However, a decrease of 3% would be expected if serotype redistribution occurred without within-serotype changes in non-susceptibility. Standardization by either factor produced hypothetical data significantly different to observed data. Within serotypes, the average annual increase in odds of non-susceptibility to penicillin or erythromycin was 1.08 (95% CI 1.05-1.11). Recent antibiotic exposure, urban residence and increased household size of participants predicted isolate non-susceptibility but did not explain the increase over time.

An overall increase in non-susceptibility of carried pneumococcal isolates to penicillin or erythromycin resulted from increases in non-susceptibility within serotypes, which outweighed a protective effect of serotype redistribution. Characterization of emerging resistant clones within carried non-vaccine serotypes, including risk factors for colonization and disease, would support disease prevention efforts and inform vaccine strategies.