Group B streptoccal infection
Prompt and effective screening can prevent potentially deadly group B streptococcus infections in newborns. An expert examines both conventional and new methods.
Over the past two decades, considerable information has emerged on the epidemiology and pathophysiology of neonatal sepsis caused by group B streptococci (GBS). In 1996, the Centers for Disease Control and Prevention, with input from the American College of Obstetricians and Gynecologists and the American Academy of Pediatrics (AAP), issued recommendations for prophylactic strategies against perinatal GBS transmission. The CDC recommends using clinical risk factors or prenatal culture at 35 to 37 weeks' gestation to identify candidates for antibiotic prophylaxis.
There is now agreement on preventive measures, but we still need prompt and effective techniques to screen for GBS. While culture remains the gold standard, more rapid tests for detection are desirable and may be helpful in some clinical conditions. I will review screening methods and examine their sensitivity, specificity, predictive values, and relevance to clinical practice.
Streptococcus agalactiae is a facultative gram-positive Diplococcus that grows readily on blood agar. Colonies are small, gray-white, flat, and mucoid, and the characteristic finding of ß-hemolysis occurs in more than 99% of GBS isolates. In the laboratory, these characteristics are used to select colonies for identification. Presumptive identification of GBS includes nonserologic methods such as Gram's stain, a negative catalase test, and the Christie-Atkins-Munch-Peterson (CAMP) test. In the CAMP test, GBS and Staphylococcus aureus are plated together. If synergistic hemolysis occurs, presumptive identification of GBS is made. Many laboratories use a latex agglutination test for definitive identification.
The site of culture is important in the isolation of GBS. The highest yield of GBS comes from cultures of the distal vagina and anorectum. The intestine is the presumed reservoir for GBS and obtaining a rectal culture increases the positive yield by 5% to 25%. The yield is lower from the cervix than from the distal vagina or rectum. Decisions regarding prophylaxis should be based on the presence of the organism anywhere in the anogenital tract. For convenience, most authorities recommend a single culture collected from the distal vagina and anorectum using paired swabs.
The culture technique also affects the yield of GBS. While the traditional culture medium is 5% sheep blood agar, many studies indicate that selective medium enhances the detection of GBS by approximately 50%. The preferred medium (recommended by the CDC and ACOG) is Todd-Hewitt broth with nalidixic acid plus gentamicin or colistin (for example, Lim or SBM broth). The addition of these antibiotics inhibits the growth of both normal lower genital tract flora and gram-negative Enterobacteriaceae that may interfere with recovery of GBS. The selective broth medium is inoculated with swabs from the distal vagina and anorectum. If selective broth medium is not available, swabs can be placed in transport media (for example, Amies) that will maintain GBS viability for up to 4 days. After 18 to 24 hours of incubation in selective broth medium, an aliquot is plated onto 5% sheep blood agar. After another 24 hours, potential GBS colonies are evaluated according to morphology and the presence of hemolysis, as discussed earlier.
In early pregnancy, many women undergo screening via urinary cultures. A woman should be considered to be colonized in the genital tract if GBS is cultured from her urine. Because this colonization is often heavy, positive urinary tract cultures are a risk factor for neonatal GBS sepsis; hence, a woman with a positive result is a candi-date for prophylaxis. In addition, women with urine cultures positive for GBS need to be treated because this infection can lead to cystitis or pyelonephritis. Do a follow-up culture after completion of treatment.
Completion of anorectal culture requires approximately 48 hours. In the culture-based scheme suggested by the CDC, cultures are obtained at 35 to 37 weeks' gestation. In ideal conditions, there is adequate time for culture processing and notification of the provider, the patient, and the labor and delivery unit. In some situations, however, such as when women present in labor without prenatal care or prior to the return of culture results, an effective rapid screening technique would be of great benefit.
In any discussion of rapid screening tests for GBS, three points must be kept in mind. First, the technique should be compared to the gold standard (that is, culture). Second, to be considered effective, a screening test must have high sensitivity, specificity, and predictive value. Third, predictive values change according to the prevalence of the organism in the population. In populations where GBS prevalence is low (5% to 10%), predictive values may be much lower than in those where the prevalence is high (30% and over).
Gram's stain. Because Gram's stain is widely available, inexpensive, and rapidly performed, investigators have considered its use for GBS detection. But compared with culture, Gram's stain has low sensitivity and low positive predictive value; therefore its usefulness as a screening tool for GBS infection is limited.
Rapid identification from culture. Methods for rapidly identifying GBS after a short incubation include those based on the starch serum medium and immunofluorescent antibodies.
The starch serum medium uses a colorimetric assay to detect GBS. This technique relies on GBS being the only streptococci to produce the carotenoid pigment that causes a color change from white to orange after 6 to 12 hours of incubation. While several studies of this method have demonstrated excellent sensitivity, specificity, and positive and negative predictive values (93% to 98%, 98% to 99%, 83% to 98%, and 98% to 99%, respectively), two important limitations remain. First, if the clinician uses this test during labor to determine whether a patient is a candidate for GBS prophylaxis, many women will have delivered by the time the result is available. Second, skilled laboratory personnel are needed to interpret the test, and they are not available 24 hours a day. (In a study that compared interpretations by microbiologic technicians with those by nurses, sensitivity and positive predictive values were 30% to 50% lower when tests were read by nursing staff.)
In the immunofluorescent antibody test, selective broth medium is mixed with type-specific GBS antisera conjugated with a fluorescein dye. The technologist then identifies organisms under fluorescent microscopy. Comparisons of immunofluorescence testing with vaginal culture demonstrate a sensitivity of only 49% after up to 6 hours of incubation in selective medium, though sensitivity increases considerably with incubation up to 18 hours. Again, the need for skilled technicians and the impracticality of longer incubation limit the usefulness of this test in laboring patients.
Direct antigen detection. Other tests detect GBS antigen directly from clinical specimens of the anogenital tract. Antigen detection techniques from vaginal and cervical swabs using commercially available kits include those based on coagglutination, latex particle agglutination, and enzyme immunoassay. In these techniques, incubation has been studied as an adjunct to direct detection from clinical specimens. While many tests are available, none of these techniques has been proved to be superior to another. Without incubation, sensitivities range from 4% to 88%. After incubation, sensitivities improve considerably. These assays work best in patients with heavy colonization; typically tests have higher sensitivity in women with heavy colonization compared with the overall group.
This characteristic of the tests is clinically relevant because an estimated 10% to 30% of cases of early-onset neonatal sepsis occur in the offspring of mothers who are lightly colonized. Thus, the ideal screening test for GBS must detect both heavily and lightly colonized patients.
Nucleic acid probes. The practical limitations of the aforementioned tests have led to the development of molecular biologic techniques for detection of GBS. Yancey and colleagues evaluated the use of a DNA probe complementary to a GBS RNA sequence to detect carrier status. While a 2.5-hour incubation period yielded a sensitivity of only 44%, sensitivity increased to 71% after 3.5 hours of growth. Specificity was 90%, positive predictive value 61%, and negative predictive value 94%. Rosa compared a commercially available DNA-based kit with GBS culture on a selective agar plate. The sensitivity of this test for GBS from a lower vaginal perineal swab was 8.3% without incubation and 81% after 16 to 24 hours of growth. Kircher examined another commercially available kit and compared results of vaginal and rectal swabs from 402 pregnant women during the third trimester with detection of GBS ribosomal RNA (rRNA). A 3-hour enrichment protocol produced a sensitivity of 73% and an 8-hour protocol achieved a sensitivity of 95%. Again, a significant incubation period to exclude light colonization limits the usefulness of this test for women in active labor.
Polymerase-chain-reaction techniques. Recently, Bergeron and colleagues have investigated the efficacy of two polymerase-chain-reaction (PCR) assays, one conventional and the other a new fluorogenic technique. They report that both PCR tests had a sensitivity of 97%, negative predictive value of almost 99%, and positive predictive value of 100%, compared with culture, for detecting GBS in pregnant women. The fluorogenic PCR assay provided results in 30 to 45 minutes. Whether these tests perform as well in routine practice remains to be seen. Another key question is whether they can be adapted to around-the-clock hospital laboratories that do not always have specialized personnel.
Culture for GBS remains the gold standard. The use of selective broth medium enhances the rate of detection of GBS by nearly 50%. Both the distal vagina and anorectum should be swabbed. On the whole, rapid screening techniques applied after a short incubation period lack the consistent sensitivity and positive predictive values necessary for clinical decision making during labor. The inability of these tests to detect women with light colonization is another major concern. The recently developed PCR assays are promising, and we need studies to determine their best use.
Bergeron MG, Ke D, Menard C, et al. Rapid detection of group B streptococci in pregnant women at delivery. N Engl J Med. 2000;343:175-179.
Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease: a public-health perspective. MMWR Morb Mortal Wkly Rep. 1996;46(RR-7):1-24.
Hager WD, Schuchat A, Gibbs R, et al. Prevention of group B streptococcal infection: current controversies. Obstet Gynecol. 2000;96:141-145.
Kircher SM, Meyer MP, Jordan JA. Comparison of a modified DNA hybridization assay with standard culture enrichment for detecting group B streptococci in obstetric patients. J Clin Microbiol. 1996;34:342-344.
Committee on Infectious Diseases and Committee on Fetus and Newborn. Guidelines for prevention of group B streptococcal (GBS) infection by chemoprophylaxis. Pediatrics. 1992;90:775-777.
Rosa C, Clark P, Duff P. Performance of a new DNA probe for the detection of group B streptococcal colonization of the genital tract. Obstet Gynecol. 1995;86:509-511.
Yancey MK, Armer T, Clark P, et al. Assessment of rapid identification tests for genital carriage of group B streptococci. Obstet Gynecol. 1992;80:1038-1047.
Yancey MK, Clark P, Armer T, et al. Use of a DNA probe for the rapid detection of group B streptococci in obstetric patients. Obstet Gynecol. 1993;81:635-640.
Robert McDuffie. OB/GYN Infection: Screening techniques for group B streptococcal infection. Contemporary Ob/Gyn 2000;12:105-116.