4 Research recommendations

The Guideline Development Group has made the following recommendations for research, based on its review of evidence, to improve NICE guidance and patient care in the future. The Guideline Development Group's full set of research recommendations is detailed in the full guideline (see section 5).

4.1 Screening and intrapartum antibiotic prophylaxis for group B streptococcal colonisation

What is the clinical and cost effectiveness of intrapartum antibiotic prophylaxis targeting group B streptococcus and guided by routine antenatal screening?

Why this is important

Antenatal care (NICE clinical guideline 62) considers the clinical and cost effectiveness of routine antenatal screening for group B streptococcus separately from the clinical and cost effectiveness of intrapartum antibiotic prophylaxis once group B streptococcus has been identified. This guideline considers the clinical and cost effectiveness of intrapartum antibiotic prophylaxis separately from the clinical and cost effectiveness of routine antenatal screening to identify women colonised with group B streptococcus.

Further research is needed to evaluate the clinical and cost effectiveness of routine antenatal screening for group B streptococcus combined with intrapartum antibiotic prophylaxis in women identified as carriers. The research could take the form of health economic modelling based on published studies or new studies (for example, randomised controlled trials or observational studies) comparing outcomes from different screening and treatment strategies. The research should also consider the gestational age at which screening should occur.

4.2 Risk factors for early-onset neonatal infection and symptoms and signs

Which risk factors for early-onset neonatal infection, clinical symptoms and signs of infection, and laboratory investigations should be used to identify babies who should receive antibiotics?

Why this is important

The evidence reviewed for the guideline included several risk scoring models designed to identify babies at risk of developing an early-onset neonatal infection and in whom antibiotic treatment should be started. The models – which incorporated maternal and fetal risk factors for infection, clinical symptoms and signs of infection, and the results of laboratory investigations (such as C-reactive protein concentrations) – were intended for use before or after birth of the baby. However, the models were suboptimal because they were not specific to early-onset neonatal infection, or they were based on data collected using a case–control design (which tends to overestimate predictive accuracy because it includes extremes of the risk spectrum but not the harder to classify patients who are not obviously free from infection or confirmed as having an infection), or they did not examine predictive accuracy in independent training and validation sets.

Further research is needed, particularly to examine risk scoring models that incorporate measurements from novel laboratory investigations, such as molecular diagnostics (polymerase chain reaction and 16S approaches). The ideal study design would be a randomised controlled trial that compares clinical outcomes associated with particular investigation and treatment initiation strategies. The next best design would be a prospective cohort study to determine the predictive accuracy of an investigation strategy or a risk scoring model evaluated in a clinically relevant group of babies that is independent of the study population used to derive the risk scoring model.

4.3 Intrapartum antibiotic prophylaxis in preterm labour

What is the clinical and cost effectiveness of intrapartum antibiotic prophylaxis using benzylpenicillin in women with preterm labour?

Why this is important

In the absence of unequivocal evidence of clinical and cost effectiveness of intrapartum antibiotic prophylaxis to prevent early-onset neonatal infection in the babies of all women with preterm labour, the recommendation to consider intrapartum antibiotic prophylaxis for women with preterm labour and either prelabour rupture of membranes or confirmed or suspected rupture of membranes of 18 hours' duration or longer was based on the Guideline Development Group's consensus view and knowledge of current practice.

Further research is needed to evaluate the clinical and cost effectiveness of intrapartum antibiotic prophylaxis using benzylpenicillin compared with placebo in women with preterm labour (including women with intact membranes and those with ruptured membranes). The research should be conducted through multicentre randomised controlled trials, including some UK centres to allow subgroup analysis of UK data. The primary outcome for evaluating the clinical effectiveness of benzylpenicillin should be the incidence of early-onset neonatal group B streptococcal infection (infection within 72 hours of birth). Secondary outcomes should include long-term outcomes in the baby. The research should include subgroup analyses for women in spontaneous preterm labour with intact membranes and those with membranes that rupture before or during labour.

4.4 Investigations during antibiotic treatment

What is the clinical and cost effectiveness of laboratory investigations used individually or in combination to exclude early-onset neonatal infection in babies receiving antibiotics for suspected infection?

Why this is important

The systematic reviews conducted for the guideline identified limited evidence relating to investigations used to guide the decision to stop antibiotic treatment in babies receiving antibiotics for suspected early-onset neonatal infection. One study evaluated procalcitonin-guided decision making for identifying babies in whom antibiotic treatment could safely be stopped, but the approach used was at an early stage of development and had not been evaluated fully.

The guideline recommendations reflected uncertainty about the diagnostic test accuracy of laboratory investigations used individually or in combination, and further research involving sufficiently powered studies is needed to evaluate this. The ideal study design would be a randomised controlled trial that compares clinical outcomes associated with particular investigation and treatment termination strategies. The next best design would be a prospective cohort study to determine the diagnostic test accuracy of an investigation strategy evaluated in a clinically relevant group of babies. The research should examine clinical effectiveness or diagnostic test accuracy in preterm and term babies separately.

4.5 Duration of antibiotic treatment

What is the optimal duration of treatment (course length) in babies who receive antibiotics for confirmed early-onset neonatal infection?

Why this is important

The Guideline Development Group identified no evidence to inform the choice of duration of antibiotic treatment (course length) for confirmed early-onset neonatal infection. In the absence of evidence, the Guideline Development Group based its recommendations on its knowledge of current clinical practice. Further research is needed to evaluate different course lengths in the following clinical circumstances:

  • babies with group B streptococcal bacterial meningitis

  • babies with group B streptococcal septicaemia

  • babies with Gram-negative bacterial meningitis (such as Escherichia coli meningitis)

  • babies with Gram-negative septicaemia.

The research should ideally take the form of multinational randomised controlled trials. The primary outcome should be relapse within 10 days of stopping treatment. Secondary outcomes should include long-term neurodevelopment.

  • National Institute for Health and Care Excellence (NICE)