Meningitis in a Newborn: What Doctors Now Look For in the First Week

Four studies published in 2026 — from Australia and China — on which babies are most at risk of lasting harm, and why a second spinal tap can matter

A small infection with a long shadow

Meningitis is an infection of the thin layers that wrap around the brain and spinal cord. Newborn babies catch it more often than people of any other age, and antibiotics usually clear the bacteria. The harder problem is the swelling and inflammation left behind, which can injure the brain even after the infection is gone [1].

That is why a baby who seems to recover well can still, months later, be found to have hearing loss, seizures, or slower development. Long-term studies that followed thousands of children after infant meningitis found that the extra risk of these problems lasts for years [2]. Four research papers published in 2026 tried to answer a practical question that matters enormously to families: can doctors tell, during the hospital stay, which babies are heading for trouble — early enough to do something about it [3], [4], [5], [6]?

How families and doctors used to face this

For most of the past thirty years, progress in newborn meningitis came from stopping it rather than treating it. In the 1990s, hospitals in many countries began testing pregnant women for a common bacterium called Group B Streptococcus and giving antibiotics during labour to those who carried it. That worked well for infections that appear in a baby's first few days [7]. It did nothing, however, for infections that appear two or three weeks later, because those come from bacteria the baby picks up after birth rather than during delivery.

Treatment, meanwhile, stayed much the same. Doctors settled on standard antibiotic combinations and standard lengths of treatment, and there it stopped. In older children with one particular type of meningitis, adding a steroid drug reduces hearing loss — but the studies in newborns were too few and too small to show whether it helps them, so it never became routine, and that uncertainty has lasted decades [8].

So for a long time, the conversation with parents went something like this: the antibiotics are working, your baby is improving, and we will keep an eye on things. What "keeping an eye on things" meant varied a great deal between hospitals. Some scanned every baby's brain in detail before sending them home. Others did a simpler bedside ultrasound and, if it looked normal, stopped there. Some repeated the spinal tap a few days into treatment; many did not, because it is uncomfortable for the baby and rarely changes which antibiotic is used. Families were often left waiting for a developmental check-up months later to learn whether anything had been missed. The four 2026 studies were designed to narrow that waiting.

What the four studies did

The first, from three large children's hospitals in New South Wales, Australia, looked back over five years at every baby up to six months old who had bacteria grown from a sample of the fluid around their brain and spinal cord [3]. There were 21 genuine cases.

The second, from a children's hospital in Hangzhou, China, followed 68 newborns with meningitis and recorded which of them developed visible damage inside the skull. Every brain scan was read by two specialist radiologists who did not know how the baby had fared, so their judgement could not be swayed by the outcome [4].

The third, across three hospitals in Changsha and Guangzhou, China, compared 73 babies who developed hydrocephalus — a damaging build-up of fluid inside the brain — with 154 who did not, from 227 babies with meningitis. This group did something unusual: they repeated the spinal tap on a fixed schedule, so the fluid could be tracked over time rather than measured once [5].

The fourth, from a children's hospital in Suzhou, China, took 214 babies treated over 13 years and built a scoring chart designed to estimate each baby's risk from information available in the first day or two [6].

What they found

The Australian study is a reminder that this infection has not become gentler. Group B Streptococcus was the commonest cause, in 6 of 21 babies, followed by two other bacteria in 4 babies each. Four babies died — one in five. All four had been born a few weeks early, all became ill after their first week, and three of them had an infection around a tube that had been surgically placed to drain fluid from the brain. Infections caused by one broad family of bacteria, called Gram-negative, killed 3 of 9 affected babies, a rate no better than an Australian study from forty years earlier.

Two details from that study are worth knowing as a parent. All six babies with Group B Streptococcus infection became ill after their first week and arrived from home rather than from the hospital ward — and five of their mothers had been tested during pregnancy, with only two testing positive. A negative test in pregnancy is genuinely reassuring about the first few days of life; it is not a guarantee for a three-week-old who has become unwell. And the early signs were vague. The typical baby had about nine different symptoms, most commonly being unsettled or irritable (two-thirds) and unusually sleepy or floppy (just over half). A bulging soft spot on the head — the sign many people expect — was present in fewer than half, and seizures in only a third. Among the babies who were sickest of all, the obvious signs were actually less common. The sickest babies looked the quietest.

The Hangzhou study found that damage inside the skull was common: just under a third of the 68 babies, 29 in every 100, had some form of it. The most frequent was softening of brain tissue, followed by fluid collections, hydrocephalus, abscesses and inflammation of the lining inside the brain's fluid spaces. Two things marked out the babies this happened to: having seizures, and having Group B Streptococcus as the cause. Nothing about birth weight, how early the baby was born, or the delivery told the two groups apart — and neither did the first spinal fluid sample.

That last point is what makes the Changsha and Guangzhou study so useful. Because they repeated the spinal tap, they could see something a single test cannot show: whether the infection is settling. In babies who went on to develop hydrocephalus, the second sample — taken within a week of the first — still showed high inflammation, while in babies who did well it was falling. By the third sample, the difference was stark. The number that mattered was not how bad the fluid looked on day one, but whether it was better on day five.

The reason to care is what followed. Of the 73 babies with hydrocephalus, about half needed a small device implanted to drain fluid and a third needed a permanent shunt. Between six months and two years of age, 37 in 100 had developed epilepsy compared with 4 in 100 of the other babies; 53 in 100 had hearing problems compared with 16; and 40 in 100 had brain tissue damage compared with under 5. Six of them died, compared with one baby in the much larger group without hydrocephalus.

Finally, the Suzhou team turned these threads into a chart. Using only how early the baby was born, whether they had seizures, and two measurements from the spinal fluid, it sorted babies into lower-, middle- and higher-risk groups. Among babies in the lowest group, about 15 in 100 had a poor outcome at discharge; in the middle group, 38 in 100; in the highest group, 69 in 100. The researchers were careful to say this chart is not ready to be used on real babies yet, because it was built in a single hospital and has not been tested anywhere else.

What this means for you and your baby

If your baby is being treated for meningitis, the most useful thing to understand is why the team may want to repeat a test that has already been done. A second spinal tap a few days in is not a sign that something has gone wrong or that the first one failed. It is how the team checks that the inflammation is actually going down — and that single piece of information turns out to predict complications better than anything measured on the first day.

If your baby has had a seizure, you will probably notice the team becoming more watchful rather than more worried in front of you. A seizure is taken as a signal that the brain has been more involved, so it usually triggers closer monitoring, brain scans, and sometimes continuous recording of brain activity. Many newborn seizures cause no visible movement at all, which is why guidelines recommend that kind of continuous recording rather than relying on what can be seen at the cotside [9], [10]. Importantly, the researchers were clear that a seizure is a sign of injury already happening, not the cause of it — so stopping the seizure is worth doing, but it is not the same as preventing the damage.

You may also be offered two different kinds of brain imaging. A bedside ultrasound is quick, painless and good at spotting problems that need urgent surgery. A magnetic resonance scan takes longer but shows far more detail, and it is the better test for understanding how much of the brain has been affected [11], [12].

Lastly, if your baby has developed hydrocephalus, ask for hearing and development follow-up to be arranged before you go home rather than afterwards. In these studies, hearing and developmental problems were common enough in that group that finding them early — when support helps most — is the whole point [13].

What researchers are working on next

Three efforts are under way. The first is testing the risk chart in other hospitals and other countries, so that the numbers mean something outside the place they came from. The second is a question that has been open for decades: whether adding a steroid or an immune treatment to antibiotics protects the newborn brain. Studies like these four cannot answer it, because they only watch what happened rather than compare treatments fairly [3]. The third is prevention. A vaccine given to mothers during pregnancy could protect against the late Group B Streptococcus infections that antibiotics in labour cannot reach, and work on such a vaccine is continuing [7]. There is also growing attention to infections around surgically placed brain drains, which caused most of the deaths in the Australian study and affect a growing number of babies born very early [14].

It is also worth naming where the greatest need is. All four of these studies come from well-resourced hospitals. Research modelling the burden of disease in eastern sub-Saharan Africa shows newborn meningitis remaining common and deadly across the past three decades, in places where repeat spinal taps and detailed brain scans are simply not available [15]. Making the careful watching described here possible everywhere is as important as refining it further.

None of this removes the uncertainty that comes with a newborn in hospital. But it does change what that week looks like. Instead of waiting to see what happens, the team now has two concrete questions it can ask on day five — has this baby had a seizure, and is the inflammation clearing — and reasonable evidence that the answers matter [16].

References

  1. van der Flier M. Neonatal meningitis: small babies, big problem. Lancet Child Adolesc Health. 2021;5(6):386–387. doi:10.1016/S2352-4642(21)00092-400092-4) ↩
  2. Snoek L, Gonçalves BP, Horváth-Pühó E, et al. Short-term and long-term risk of mortality and neurodevelopmental impairments after bacterial meningitis during infancy in children in Denmark and the Netherlands: a nationwide matched cohort study. Lancet Child Adolesc Health. 2022;6(9):633–642. doi:10.1016/S2352-4642(22)00155-900155-9) ↩
  3. Duffy N, Koirala A, Williams PCM. Epidemiology and Outcomes of Culture-proven Neonatal Meningitis in Australian Tertiary Hospitals: A Multicenter Retrospective Cohort Study. Pediatr Infect Dis J. 2026;45(6):e193–e195. doi:10.1097/INF.0000000000005131 ↩
  4. Zhang F, Yu Z, Pan J, Yuan T. Risk factors for acute neurological complications in neonatal bacterial meningitis: a retrospective cohort study. J Pediatr (Rio J). 2026;102(3):101533. doi:10.1016/j.jped.2026.101533 ↩
  5. Xu J, Cheng T, Yu Z, et al. Clinical features, risk factors and outcomes of hydrocephalus in neonatal bacterial meningitis: a retrospective case–control study. Sci Rep. 2026;16(1):20218. doi:10.1038/s41598-026-50052-w ↩
  6. Yu M, Fang Y, Yang X. Development and internal validation of a prognostic nomogram for poor outcomes in neonatal bacterial meningitis. Transl Pediatr. 2026;15(5):177. doi:10.21037/tp-2026-0206 ↩
  7. Shane AL, Sánchez PJ, Stoll BJ. Neonatal sepsis. Lancet. 2017;390(10104):1770–1780. doi:10.1016/S0140-6736(17)31002-431002-4) ↩
  8. Hasbun R. Progress and Challenges in Bacterial Meningitis: A Review. JAMA. 2022;328(21):2147–2154. doi:10.1001/jama.2022.20521 ↩
  9. Wusthoff CJ, Numis AL, Pressler RM, et al. The American Clinical Neurophysiology Society Guideline on Indications for Continuous Electroencephalography Monitoring in Neonates. J Clin Neurophysiol. 2025;42(1):1–11. doi:10.1097/WNP.0000000000001120 ↩
  10. Mehta N, Shellhaas RA, McCulloch CE, et al. Seizure Burden, EEG, and Outcome in Neonates With Acute Intracranial Infections: A Prospective Multicenter Cohort Study. Pediatr Neurol. 2022;137:54–61. doi:10.1016/j.pediatrneurol.2022.09.001 ↩
  11. Licht-van der Stap RG, de Vries LS, Alarcon A, et al. Cranial ultrasound in neonatal brain infections. Dev Med Child Neurol. 2025;67(8):986–1003. doi:10.1111/dmcn.16279 ↩
  12. Oliveira CR, Morriss MC, Mistrot JG, et al. Brain magnetic resonance imaging of infants with bacterial meningitis. J Pediatr. 2014;165(1):134–139. doi:10.1016/j.jpeds.2014.02.061 ↩
  13. Haffner DN, Machie M, Hone E, et al. Predictors of Neurodevelopmental Impairment After Neonatal Bacterial Meningitis. J Child Neurol. 2021;36(11):968–973. doi:10.1177/08830738211026053 ↩
  14. Strunk T, Molloy EJ, Mishra A, Bhutta ZA. Neonatal bacterial sepsis. Lancet. 2024;404(10449):277–293. doi:10.1016/S0140-6736(24)00495-100495-1) ↩
  15. Tadesse S, Gemechu T, Lemma K, et al. Incidence of Neonatal Meningitis and the Premature Mortality in Eastern Sub-Saharan Africa, From 1990 to 2023: A Global Burden of Diseases Study. Sage Open Pediatr. 2026;13:30502225261476890. doi:10.1177/30502225261476890 ↩
  16. Sánchez PJ, Duncanson AS, Gamer JT, Haffner DN. Neonatal bacterial meningitis: the elusive search for markers of severity. J Pediatr (Rio J). 2026;102(4):101567. doi:10.1016/j.jped.2026.101567 ↩