Long-term effects of the pneumococcal conjugate vaccine on antimicrobial resistance in children

Abstract number: P1067

Mera R.M., Miller L.A., Amrine-Madsen H., Sahm D.F.

Objectives: To model the changes in S. pneumoniae resistance to several antimicrobial classes five years before and six years after the introduction of the conjugate pneumococcal vaccine in the US.

Methods: A total of 129,562 isolates from the TSN Network^® surveillance database (Eurofins Medinet) during the period 1996 to 2007, as well as age, specimen source, inpatient or outpatient status and US census region were available for analysis. To appropriately model the rise, drop and subsequent rebound in antimicrobial resistance, cubic splines were utilised in a logistic regression model. Multi-drug resistance was considered as full resistance to two or more antimicrobial classes.

Results: In children less than five years old, antimicrobial resistance continuously increased from 1996 to reach its peak between 2000 and 2001. 2001 rates were 48.8% for TMP/SMX, 44.9% for erythromycin, 32% for penicillin, 24.9% for tetracycline, 7.0% for ceftriaxone, and 29.5% for multiple resistance. All antimicrobial classes subsequently experienced a steep drop in resistance before leveling off between 2003 and 2004. Levels in 2004 were 35.3% for TMP/SMX, 38.1% for erythromycin, 21.1% for penicillin, 19% for tetracycline, 2.6% for ceftriaxone and 22.1% for multiple resistance. A rebound was also experienced in every class, with 2007 levels near or above the 2001 peaks. Resistance rates for 2007 were 41.1% for TMP/SMX, 45.7% for erythromycin, 27.6% for penicillin, 30.4% for tetracycline, 3.1% for ceftriaxone, and 27.9% for multiple resistance. Similar changes were observed by source of the isolate, with blood/CSF isolates having lower peaks and rebounds, and otitis media isolates showing very steep changes over time. Similar changes were also observed by inpatient/outpatient status and in different regions of the country, with larger variations among the regions that started with higher levels.

Conclusion: Most antimicrobial classes experienced a significant drop in resistance after the introduction of the pneumococcal vaccine in 2001. This effect reached its maximum in 2004 with a subsequent and significant rebound by 2007. The same pattern is seen regardless of specimen source, US census region or inpatient/outpatient status, and may be due to the fact that the vaccine serotypes were the most frequent and most resistant in 2001. As they were replaced by non vaccine serotypes, resistance declined initially but later increased as the non-vaccine serotypes acquired resistance.

Resistance prevalence (%) in <5 year olds.