Use of the Serum Bacterial Antigen Test for the Detection of Group B Streptococcal Neonatal Sepsis

The Revised Guidelines from the Centers for Disease Control for the Prevention of Perinatal Group B Streptococcal Disease were presented in the Morbidity and Mortality Weekly Report. An algorithm for evaluation of Group B Streptococcal sepsis (GBS) included CBC and differential, blood culture, possible chest x-ray, and lumbar puncture when signs of sepsis were present. The serum bacterial antigen was not mentioned in the recommendations although its clinical use for GBS evaluation has continued. We sought to determine if routine use of Group B Streptococcal Serum Antigen screening was indicated. According to hospital practice, serum bacterial antigens were drawn along with blood cultures in evaluation of sepsis for patients up to 2 months of age. Use of serum bacterial antigen testing using the BD Directigen Combo test was analyzed retrospectively over a five year period. Predictive value, sensitivity, and specificity of the analysis were studied relative to presence of a positive blood culture result. Over a five year period (2001-2005), 3336 serum bacterial antigens were performed. During that time, there were 23 positives (0.69%) for GBS. There were 3313 negatives (99%). There were 11 cases where the bacterial antigen predicted the Group B neonatal sepsis. There were 12 cases where the bacterial antigen test was positive for GBS; however, the blood cultures had no growth. In two cases, the antigen test was positive for GBS with a positive blood culture for a different bacterium: one blood culture grew Bacillus species; and the other, coagulase negative Staphylococcus. During 2005, the NICU had no positive blood cultures for GBS, and there were only two other cases hospital-wide where the bacterial antigen test predicted a positive blood culture for GBS. Although sensitivity, specificity, and negative predictive value were 99%-100%, the positive predictive value was 48% with a disease prevalence of 0.33%. Continued evaluation of Group B Streptococcal Antigen would not result in an identifiable risk reduction and is not justifiable for routine screening.