Background: In May 1996, the CDC recommended obtaining a complete blood count and blood culture (BC) from all asymptomatic "at risk" newborns; those > or =35 weeks gestation born to mothers with group B streptococcal vaginal colonization or those with maternal fever, premature rupture of membranes or previous infant with group B streptococcal disease; who did not receive adequate intrapartum antibiotic prophylaxis.
Design/methods: During the study period (May 1996 to July 1999), a complete blood count and BC were obtained within 4 h from all asymptomatic at risk newborns of > or =35 weeks gestation. White blood cell count (WBC) and BC results and prevalence of clinical or culture-proven sepsis were obtained by chart review. We determined the sensitivity/specificity and likelihood ratios of an abnormal WBC (total >30 000 or <5000/mm3; absolute neutrophil count <1500/mm3, or a band form-polymorphonuclear cell ratio of >0.2) to distinguish between clinically septic vs. nonseptic term at risk newborns.
Results: Of 20 554 deliveries 1665 were initially asymptomatic at risk newborns; 17 (1.0%) developed early onset sepsis, all within 48 h. WBC was abnormal in 7 of 17 (41%) and in 447 of 1648 (27%) who remained nonseptic. None of the 1665 term at risk newborns had a positive BC. The sensitivity and specificity of an abnormal WBC in predicting which at risk newborns would develop sepsis were 41 and 73%, respectively. The positive likelihood ratio was 1.52, whereas the negative likelihood ratio was 0.81, with an odds ratio of 1.88.
Conclusions: Since the implementation of the CDC guidelines for maternal intrapartum antibiotic prophylaxis, culture-proved sepsis has become rare at our institution. Although BC did not aid in the diagnosis of sepsis among asymptomatic at risk newborns, close observation in the first 24 h remained critically important.