Rates of Early-Onset Group B Streptococcus Infection in the Neonate

Typically, in the United States, during a prenatal visit at around 35 to 37 weeks of pregnancy, the obstetric provider will bring up GBS testing. This test looks for the presence of a bacterium called Streptococcus agalactiae or group B Streptococcus in the genital tract or rectum of the pregnant person. In North America group B Streptococcus (GBS) colonizes (is present without causing an infection) the genital tract and rectum of about 23% of pregnant people (Russell, 2017 a). The test is usually performed by collecting the organisms from the birth canal and rectum via a swab, that is then sent in a transport medium to the laboratory for culturing (a specific process that facilitates, in this case, the growth of GBS if it is present). The pregnant person can request to collect the sample themselves if they prefer (CDC, 2010). If the person tests negative, there is nothing else to consider. If the person tests positive, they are typically told that they will need intravenous antibiotics during labor.

The reason why people are told that they will need intravenous antibiotics during labor even though the presence of GBS in the genital tract and rectum is not an infection, is because there is a small chance that their baby may become infected with GBS during labor. GBS infection can have an early-onset (within the first 7 days of life) or a late-onset (8 days to 3 months). Only the early-onset GBS (EOGBS) infection is reduced by antibiotic administration during labor. It has been demonstrated that in most cases EOGBS infection begins in the womb (Boyer & Gotoff, 1985; Dillon, 1987). Hence, giving antibiotics to the birthing person during labor is more effective at preventing infection in the baby than giving antibiotics to the baby once the baby is born. Therefore, the antibiotics are initiated at either the start of active labor or waters releasing (whichever happens first), and are repeated every four hours until the birth of the baby (CDC, 2010).

Of those babies who develop EOGBS infection, 84% to 96% will show symptoms within the first 24 hours of life (Boyer, 1983 I; Dillon, 1987; Yagupsky, 1991). The sick babies can develop pneumonia, sepsis, and meningitis. The factors that increase the risk of developing EOGBS infection are preterm birth (before 37 weeks), prolonged rupture of membranes (>18 hours from the time the water bag releases until birth, but some authors define this differently), or fever in the adult during labor (Benitz, 1999; Boyer, 1983 I; Dillon, 1987; Yagupsky, 1991).

When doing research on EOGBS infection, the statistics are rather inconsistent. Most commonly found statistics for EOGBS infection are 1 in 200 (CDC, 2018), 1.1% (Russell, 2017 b) or 1-2% (CDC, 2010). These statistics nearly double from one to the next. The variation in risk is often due to the fact that different groups of GBS positive people, have a different risk of having an infant with EOGBS infection. Unfortunately, though the risk is different for different populations, the person looking at the numbers rarely knows which population the risk refers to. Typically, the statistics represent populations that include preterm births, births with prolonged rupture of membranes (ROM), and births where the birther developed a fever (Russell, 2017 b).

When someone is positive for GBS, their decision regarding taking antibiotics in labor should be made with the understanding of different levels of risk depending on some important criteria. These criteria include whether the pregnancy is term or preterm, whether the water bag released before labor or has been broken for a prolonged period, or if a fever developed in labor (Benitz, 1999; Boyer, 1983 I; Dillon, 1987; Yagupsky, 1991). According to Benitz 1999 the risk of a preterm infant developing EOGBS infection is about 5 times higher than a term infant, if there is prolonged ROM the risk goes up about 7 times when compared to ROM at <18 hours, and in the presence of intrapartum fever the risk goes up about 4 times. The EOGBS infection rates confirmed by positive blood or CFS cultures are conservative and somewhat underestimate the actual number of babies infected by GBS, but without a positive culture there is no way to be certain if the infection was caused by GBS (Benitz, 1999; Russell, 2017 b).

Unfortunately, there is very little research on the population of GBS positive people who are term and have not been treated with antibiotics during labor. Even among people with GBS who have risk factors, the rate of EOGBS infection is very low; hence, a large group of people (over 800) would need to be examined get a good sense of the risk. At this point, adequate research on the outcomes for babies born at term to people who have tested GBS positive by the standard testing method utilized in the U.S. (CDC recommended) does not exist; hence, we will examine the available evidence.

In an article by Boyer & Gotoff (Boyer & Gotoff, 1985), the authors report their prior research conducted in the United States. According to this article, if a person tested positive for GBS prenatally, carried their pregnancy to term, had a labor uncomplicated by fever (>37.5˚C) or prolonged ROM (>12 hours in this article), and did not receive antibiotics, the risk of their baby developing EOGBS infection was 0.5%. That same report identified babies born in the presence of one or more risk factors (preterm, prolonged ROM, or adult fever) as having a 4% chance of developing EOGBS infection. Unfortunately, in the absence of details on data acquisition, it is difficult to comment on the validity of these statistics.

A study by Boyer et al. (Boyer, 1983 lll) conducted in the United States, examined 325 pregnant people who tested positive for GBS at some point during their pregnancy, had no risk factors (preterm labor, prolonged ROM, or fever), and experienced a birth free of antibiotics. There were 0 babies diagnosed with EOGBS infection born to these 325 GBS positive people. The fact that out 325 births, not one baby developed EOGBS infection bodes quite well for babies born to people colonized by GBS, as long as the birth is without risk factors. However, the timing of the test in this trial is earlier than the current CDC recommendation; hence, it is less accurate than a test done at 35-37 weeks, as is done today. Subsequently, only about a 67% intrapartum GBS retention rate would be expected (Boyer, 1983 ll), what that means is that only 67% of birthers would still be positive for GBS when labor commenced. Nevertheless, the risk of culture proven EOGBS infection is clearly less than 0.5%, according to this cohort.

In a meta-analysis by Russell et al. (Russell, 2017 b) the rates of EOGBS infection from multiple studies worldwide were pooled together to arrive at an average. The rate of EOGBS infection in babies born to GBS positive people who did not receive antibiotics during labor was 1.1%. The studies that were pooled to arrive at this number included babies who were preterm, low birth weight, born to people who developed fever during labor, or had prolonged ROM, all of which increase the risk of EOGBS infection (Benitz, 1999; Boyer, 1983 I; Dillon, 1987; Yagupsky, 1991). Additionally, some studies obtained a culture from high in the genital tract, did not obtain rectal cultures, or neglected to use selective enrichment medium. These techniques are at odds with the testing methods utilized in the U.S. and result in a lower rate of GBS detection (CDC, 2010). What that means is that these techniques identify people with heavy genital tract GBS colonization. Babies born to people with heavy genital tract GBS colonization are at much higher risk of being heavily colonized with GBS. The babies who are heavily colonized are 10 times more likely to develop EOGBS infection than lightly colonized babies (Boyer, 1983 ll; Dillon, 1987). Essentially, due to the inclusion of these studies by Russell et al., the review was skewed to over-represent people who are at a higher risk of having babies with EOGBS infection.

In a study from Sweden (Burman, 1992), only term pregnancies were examined; however, prolonged ROM and birthers with fever during labor were not excluded. Among the 409 term babies born to people who tested GBS positive and did not receive antibiotics during labor, one developed EOGBS infection. This suggests that the risk of EOGBS infection among term babies is about ¼ % or about 1 in 400, even if birthers with prolonged rupture of membranes and who developed a fever in labor are included. Two things of note about this trial; one, the babies were only evaluated for 4 days. Hence, it is possible that an infant who developed symptoms after 4 days was missed. However, by the 4th day, roughly 98% or more show symptoms if EOGBS infection is present (Bromberger, 2000; Dillon, 1987; Yagupsky, 1991). The other parameter to note about the study is that the specimen collection and culturing methods that were used, favored heavy GBS colonization; namely, a sample was taken high in the birth canal and no selective enrichment medium was used. This technique would miss people with light or moderate colonization and yield a group of only heavily colonized people, who are at higher likelihood of having a baby with EOGBS infection (Boyer, 1983 ll; Dillon, 1987).

A study by Dillon et al. (Dillon, 1987) conducted in the Unites States examined 1,523 birthing people who tested positive for GBS. Of the babies born to the 1,523 participants, 24 developed EOGBS infection. The babies in this trial had 1.6% chance of developing EOGBS infection. However, the methods of GBS testing employed in this trial were rather unorthodox; for instance, the decision whether to culture was done at the discretion of the clinician attending the birth, and the samples were obtained from only 61% of eligible participants. The culture was collected from the genital tract immediately after the birth, which introduced a bias. Specifically, the presence of risk factors during the birth (such as preterm, prolonged ROM, adult fever, or newborn symptoms) playing a role in the clinician’s decision to obtain culture. This likely increased the proportion of participants with risk factors. The comparison of people who were cultured with the general birthing population was not performed. Conversely, this trial defined EOGBS infection as occurring in the first 3 days of life, which may have omitted an occasional newborn who develops EOGBS infection beyond that period.

The babies from Dillon et al. who were heavily colonized with GBS at birth were 12 times more likely to develop EOGBS infection than babies who had lighter GBS colonization. The infants who weighed less than 2.5kg (5lbs 8oz) were 8 times more likely to develop EOGBS infection. Out of the 24 infants with EOGBS infection, 14 were born at less than 36 weeks gestation. Of the babies in this study, 20 had signs of infection within 3 hours of birth and 23 had symptoms with the first 12 hours of life. The vast majority of the EOGBS infected infants, 22 out of 24, were either preterm or had prolonged ROM (in this study is defined as >12 hours).

Based on reviewed evidence, the most relevant source, the Burman, 1992 study, suggests that term neonates born to people with heavy GBS colonization have about 1 in 400 chance of developing EOGBS infection. In the U.S. pregnant people are tested by the CDC endorsed method, which detects lighter GBS colonization in addition to heavy. About half of GBS positive people have lighter colonization density, and are 3 times less likely to have GBS colonized infants and 6 times less likely to have infants who are heavily colonized (Boyer, 1983 ll). Hence, those birthers who have lighter colonization density are 3 times less likely than those who are heavily colonized to have an infant who has any chance of EOGBS infection, since an infant must be colonized to develop the infection, and even their colonized infants have a much lower chance of developing EOGBS infection. If about half of GBS positive people are heavily colonized (1 in 400 chance of a baby with EOGBS) and half have lower colonization density (1 in 1200 chance of a baby with EOGBS, 3 times less), then on average those who test positive today in the U.S. have closer to 1 in 800 chance of having a baby with EOGBS infection if no antibiotics are administered.

In summation, there is not enough evidence to determine the risk of EOGBS infection in babies born without antibiotics at term to people who tested positive by the CDC method. Additionally, to make a fully informed choice regarding the benefit of antibiotic use during labor, one should consider the negative effects of antibiotics, primarily their effect on the microbiome of the baby and the consequences of a disturbed microbiome, which will be reviewed in the next post.


Edited 9/13/19