Abstract 1221 Poster Session II, Sunday, 5/2 (poster 4)

We describe here the relation of infant colonization and early-onset disease (EOD) due to GBS in an aggregate of 97,919 births from an ethnically and geographically diverse population from 14 hospitals in 6 urban areas in the U.S. between 7/19/95 and 12/31/97. EOD was defined as isolation of GBS from blood and/or CSF in an infant whose onset of symptoms/signs occurred, and/or in whom the first positive blood/CSF culture was taken within 7 days of birth. EOD cases were identified by active surveillance of microbiology laboratories and intensive care nurseries, while colonized infants were identified by systematic culturing of a sample of infants each month in each of the 14 study hospitals. Cultures were taken from 4 surface sites (throat, anus, umbilicus, ear) shortly after birth and before the first bath. During the study period, among 97,919 births, 106 infants with EOD were identified (overall incidence of 1.1/1000 births; 1.8 in 1995, 0.9 in 1996 and 1.0 in 1997; p = .008) and 13,155 infants had surface cultures. 1370 of 13155 surface cultured infants had GBS recovered from one or more sites for an overall colonization rate of 10.4%(14.0% in 1995, 10.7% in 1996, and 8.7% in 1997, P <.0001). GBS was most frequently recovered from ear swabs (7.6%) and least frequently from throat (4.0%). Of the colonized infants, 62% were positive at 1 or 2 sites, while 38% were positive at 3 or 4 sites. 18 (1.3%) of the 1370 colonized infants developed EOD as opposed to none of the 11,785 non-colonized infants. The attack rate for EOD increased with the degree of surface colonization (see Table). The serotype-specific attack rates of EOD for colonized infants were similar for types Ia (2.2%), Ib (2.9%), III (3.1%), and V (2.8%), while no infants colonized with type II and non-typable strains developed EOD. The ethnic group-specific colonization rate varied from 7.9% for white, to 13.0% for black (P<.001). In summary, we observed a decline in the incidence of EOD in six academic centers and noted a corresponding decline in infant colonization rates from 1995 to 1997. Our data support claims that EOD is decreasing in the U.S., confirm prior studies on the relation between the occurrence of EOD and the degree of surface colonization, and suggest that therapies aimed at reducing the degree of infant colonization with GBS at birth could be useful in reducing the occurrence of EOD.

Table 1 No caption
Full size table