The use of Streptococcus pneumoniae nasopharyngeal isolates from healthy children to predict features of invasive disease

Abstract

Background: The role of sampling nasopharyngeal carriage isolates of Streptococcus pneumoniae to determine characteristics of isolates causing invasive disease has not been established.

Methods: Data were compared from two 1995 studies of S. pneumoniae in Metropolitan Toronto and Peel Region (population, 3.1 million). The first was a prospective survey of nasopharyngeal (NP) carriage in child care centers. The second was a prospective surveillance for all cases of invasive disease.

Results: There were 545 NP S. pneumoniae isolates obtained from 532 children and 96 cases of invasive S. pneumoniae disease in children. The prevalences of reduced antibiotic susceptibility in the NP carriage and invasive studies, respectively, were: penicillin (16% vs. 11%, P=0.29); erythromycin (12% vs. 7%, P=0.25); and multiresistant (16% vs. 12%, P=0.34). The power to rule out a difference between the groups was <30% for each comparison. Trimethoprim/sulfamethoxazole resistance was more common in NP carriage isolates than invasive isolates (38% vs. 23%, P=0.02). Serotype 14 was more common in invasive isolates, whereas serogroup 6 was more common in NP carriage isolates. Antibiotic-resistant isolates were predominantly serogroups 6, 19 and 23 in both studies.

Conclusions: Nasopharyngeal carriage isolates of S. pneumoniae reflect the antibiotic susceptibility rates of invasive isolates found in the same period for most antibiotics. However, even a large study like this may have limited power to detect a difference. The most common NP carriage serotypes are the same as the invasive isolates, although the rank order of specific serotypes is different. Routine surveys of S. pneumoniae NP carriage are not feasible because of the cost of serotyping and limited power of the observations, unless sample sizes are extremely large.