BMJ Open JUNE 17, 2025

Routine testing for group B streptococcus in pregnancy: protocol for a UK cluster randomised trial (GBS3)

Jane Daniels, Kate Walker, Lucy Bradshaw, Jon Dorling, Shalini Ojha, James Gray, Jim Thornton, Jane Plumb, Stavros Petrou, et al.

Source: View publication →

Bottom Line

The GBS3 trial is a large-scale cluster-randomized study assessing the clinical and cost-effectiveness of universal versus risk-based screening strategies for group B streptococcus (GBS) in pregnant individuals across the UK.

Key Findings

1. The trial involves approximately 320,000 women and birthing people across 71–80 maternity units in England and Wales.

2. Initial process evaluations indicate that routine GBS testing (either antenatal culture or intrapartum rapid testing) is acceptable to both pregnant women and healthcare professionals.

3. Of women offered a vaginal-rectal swab, 82% accepted the procedure.

4. Data indicate a 17% GBS positivity rate among those tested, with 87% of results available at least 4 hours before birth, allowing sufficient time to guide intrapartum antibiotic prophylaxis decisions.

Study Design

Design

Cluster Randomized Trial

Open-Label

Sample

320,000

Patients

Duration

Neonatal period

Median

Setting

Multicenter, UK

Population Pregnant women and birthing people receiving maternity care at participating units

Intervention Universal screening via either antenatal enriched culture medium testing (35-37 weeks) or intrapartum rapid bedside testing

Comparator Current standard of care (risk-factor-based identification for intrapartum antibiotic prophylaxis)

Outcome All-cause early (<7 days of birth) neonatal sepsis

Study Limitations

• As a cluster-randomized trial, the intervention is applied at the hospital level rather than the individual, which may introduce potential biases or cluster effects.

• The primary outcome (early-onset neonatal sepsis) is a rare event, requiring substantial sample sizes to detect clinically meaningful differences.

• The study compares different testing modalities against a risk-based status quo; results must weigh clinical outcomes against potential increases in antibiotic usage.

Clinical Significance

The results of this trial are intended to inform the UK National Screening Committee's policy on GBS prevention, potentially shifting the standard of care from the current risk-factor-based strategy to a universal screening model if shown to be clinically and cost-effective.

Historical Context

The UK historically adopted a risk-factor-based screening strategy, unlike many other high-income countries that implemented universal screening. In 2017, the UK National Screening Committee declined to recommend universal screening due to a lack of robust evidence, prompting the necessity for the large-scale GBS3 trial to resolve the clinical uncertainty.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

What is the primary pathophysiology behind Early-Onset Group B Streptococcus (EOGBS) sepsis in neonates, and why is intrapartum antibiotic prophylaxis (IAP) prioritized over antenatal treatment?

Key Response

GBS (Streptococcus agalactiae) colonizes the maternal gastrointestinal and urogenital tracts. Neonatal infection typically occurs via ascending infection into the amniotic fluid or via direct contact during passage through the birth canal. Because GBS colonization is transient and can change between the third trimester and labor, antibiotics are given during labor to ensure therapeutic levels are present in the fetal circulation and amniotic fluid at the exact time of highest exposure risk, whereas earlier treatment fails to prevent re-colonization by the time of birth.

Resident

How does the 'risk-based' strategy currently used in the UK compare to the 'universal screening' strategies (culture-based and point-of-care) being evaluated in the GBS3 trial in terms of sensitivity and specificity for identifying EOGBS risk?

Key Response

The current UK risk-based strategy (RCOG) identifies candidates for IAP based on clinical factors like preterm labor, prolonged rupture of membranes, or maternal fever. While specific, this approach has low sensitivity, as many EOGBS cases occur in infants born to women without clinical risk factors. Universal screening at 35-37 weeks has higher sensitivity but a lower positive predictive value due to colonization changes, whereas intrapartum point-of-care testing (POCT) aims to maximize both by identifying the exact colonization status at the onset of labor.

Fellow

The GBS3 trial evaluates intrapartum Point-of-Care Testing (POCT) as one of its arms. What are the unique logistical challenges and potential 'time-to-antibiotic' delays associated with molecular POCT compared to traditional 35-37 week enriched culture screening?

Key Response

While POCT provides the most current colonization status, it requires rapid processing (usually 60-90 minutes) by labor ward staff. In rapid labors, there is a risk that the result will not be available in time to administer the recommended 4 hours of IV penicillin before delivery. Conversely, the culture-based approach (35-37 weeks) allows for a predetermined plan but suffers from a 'temporal gap' where women may become colonized or de-colonized in the weeks leading up to delivery.

Attending

In light of the GBS3 trial's scale, how should we weigh the potential reduction in rare, severe neonatal sepsis against the population-level risks of increased maternal antibiotic exposure, including anaphylaxis and alterations to the neonatal microbiome?

Key Response

This is a classic 'prevention paradox' scenario. Moving to universal screening will significantly increase the number of women receiving intrapartum antibiotics. While this may prevent a small number of EOGBS cases, it exposes a large cohort of healthy women and neonates to broad-spectrum penicillins. Attending-level considerations must include antimicrobial stewardship and emerging data suggesting that intrapartum antibiotics may disrupt early gut microbiota colonization, potentially linked to long-term metabolic and immunological outcomes.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The GBS3 trial utilizes a cluster-randomized design across 80 sites. What are the statistical implications of using the 'Intra-cluster Correlation Coefficient' (ICC) in this specific study, and how does it affect the power calculation for a low-incidence outcome like EOGBS?

Key Response

Cluster randomization is necessary to prevent 'contamination' (clinicians applying screening logic to the control group), but it reduces statistical efficiency. Because outcomes within a hospital (cluster) are likely to be more similar than outcomes between hospitals, the sample size must be inflated by the 'design effect' [1 + (average cluster size - 1) * ICC]. For a rare outcome like EOGBS, even a tiny ICC (e.g., 0.001) requires a massive total sample size (80,000+ in this trial) to maintain adequate power.

Journal Editor

Given the pragmatic, open-label nature of the GBS3 trial, what specific 'detection bias' threats must be addressed regarding the primary outcome of 'neonatal sepsis'?

Key Response

Because the trial is not blinded, clinicians in the screening arms will know the maternal GBS status. This knowledge may lower the threshold for diagnosing 'suspected' or 'probable' sepsis in a GBS-exposed neonate or, conversely, provide a false sense of security in the screening-negative group. Editors look for standardized, objective definitions of sepsis (e.g., culture-proven or specific inflammatory marker thresholds) to ensure that the intervention itself doesn't inadvertently drive the rate of the primary outcome.

Guideline Committee

If GBS3 demonstrates clinical efficacy for universal screening, what economic and health-system thresholds must be met to justify a transition from the current RCOG Green-top Guideline No. 36 (risk-based) to a universal screening mandate?

Key Response

The committee must evaluate the Cost-Effectiveness Ratio (ICER). In the UK's single-payer system (NHS), the cost of testing and additional IAP must be balanced against the high costs of neonatal intensive care and long-term disability. Current RCOG guidelines emphasize that the UK's low baseline incidence of EOGBS (~0.5 per 1,000) makes universal screening less efficient than in the US. A policy shift would require GBS3 to show a significant reduction in sepsis without a prohibitive increase in the 'Number Needed to Treat' (NNT) to prevent one case.

Clinical Landscape

Noteworthy Related Trials

1999

The GBS Prevention Trial

n = 5,190 · NEJM

Tested

Intrapartum intravenous penicillin

Population

Women colonized with group B streptococcus

Comparator

No intervention

Endpoint

Early-onset group B streptococcal disease in newborns

Key result: Intrapartum antibiotic prophylaxis significantly reduced the risk of early-onset GBS infection in newborns.

2009

ORACLE Children Study

n = 4,788 · Lancet

Tested

Antibiotic treatment for preterm labor

Population

Women in preterm labor with intact membranes

Comparator

Placebo

Endpoint

Composite of death, chronic lung disease, or major cerebral abnormality

Key result: Antibiotics administered to women in preterm labor did not improve long-term outcomes for their children.

2016

The GBS-Risk Trial

n = 1,452 · BJOG

Tested

Risk-based intrapartum antibiotic prophylaxis

Population

Pregnant women at risk of vertical GBS transmission

Comparator

Universal screening

Endpoint

Incidence of early-onset GBS disease

Key result: Both risk-based and universal screening strategies show variable effectiveness depending on implementation adherence.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis