Outcomes of meningitis in infants born preterm


In a recent study, it was found that lumbar punctures are not as common for late-onset sepsis, potentially leading to an underestimated figure when measuring cases and outcomes of late-onset meningitis.

Late-onset meningitis (LOM) incidence and outcomes in children born extremely preterm has been underestimated, according to a study published in JAMA Network Open.

Risk of meningitis is greater in newborns than in any other age group, because of an immature immune system and a blood barrier with increased permeability. Meningitis can be determined when a lumbar puncture (LP) is used to obtain cerebrospinal fluid (CSF) for a late-onset sepsis (LOS), but use of this method is debated.

Neonatal meningitis is confirmed in about 0.3 per 1000 live births, but the variance of instances where CSF is obtained could make this number inaccurate. According to a prospective study, CSF is 5 times more likely to be positive in LOS than early-onset sepsis.

To determine the detected incidence of LOM, neurodevelopmental outcomes in infants with LOM, and how often LP is used during LOS evaluations, investigators conducted a secondary analysis of children born with a gestational age of 22 to 26 weeks.

Infants were cared for at 25 Neonatal Research Network (NRN) centers from 2003 to 2017, with each NRN maintaining databases which were used for data collection. Data from 2003 to 2017 was analyzed for LOM incidence and outcomes, while data from 2011 to 2017 was analyzed for LP performance.

Isolation of bacteria or fungal pathogen in blood taken more than 72 hours after birth was used to determine LOS. Treatment lasted for 5 days or until death. Bacteria of fungal pathogen isolated from CSF culture gathered more than 72 hours after birth was used to determine LOM. Treatment lasted for 7 days or until death. 

If CSF and blood cultures were gathered within 1 week of each other, they were considered concurrent, while infections were defined as polymicrobial if there were different organisms cultured within 7 days. If one organism was cultured after the 7-day period since a different organism, it was a new infection. An instance of the same organism more than 10 days after the first was also considered a new infection.

Exclusion criteria included meningitis less than 72 hours after birth, missing LOS or LOM data, polymicrobial LOM without concurrent LOS with 1 of the same pathogens, viral LOM, ventricular shunt or reservoir presence, death before 72 hours after birth, and major congenital anomalies.

Primary outcomes included LOS, LP performance during LOS, and neurodevelopmental impairment (NDI) or death prior to follow-up. Follow-up took place when infants were aged 18 to 22 months if born before July 2012 and aged 22 to 26 months when born during or after July 2012.

There were 13,372 infants with a median gestational age of 25.4 weeks, of which 167 had LOM, 4564 had LOS without LOM, and 8641 had neither LOS nor LOM. Infants with LOM were more often born to younger mothers who were less likely to have a college degree, private insurance, or a hypertensive disorder. Mothers of this group were also less likely to undergo cesarean section.

Gestational age was lower for infants with LOS or LOM than those with neither. Infants with LOM were more likely to experience morbidities such as periventricular leukomalacia and severe intracranial hemorrhage.

Cases of LOM and LOS decreased between 2003 and 2017. LOM without concurrent positive blood culture was seen in 27 cases, 18 of which saw no positive blood culture during hospitalization. Diagnosis for LOM occurred at a median 16 days.

In 2011, LP occurred in 36% of LOS evaluations. This decreased to 24% by 2017. There were 5865 LOS evaluations performed during this time, and 1771 included LP. Of the 1771 infants who experienced LP, 47 were diagnosed with LOM. While incidence of LOS was consistent, LP performance decreased over time.

Of the 13,372 infants analyzed, 12,023 had data taken during follow-up assessments at 18 to 26 months after discharge. There were 49 deaths in children with LOM, while NDI was present in 42% of survivors. Among children with LOS but not LOM, 43% were present with NDI, compared to 33% of children with neither.

While LOM and LOS were associated with increased risk of death and NDI, LP performance decreased over time. Investigators concluded that current diagnostic practices may underestimate the full association of LOM with outcomes of children born extremely preterm.


Brumbaugh JE, Bell EF, Do BT, Greenberg RG, Stoll BJ, DeMauro SB, et al. Incidence of and neurodevelopmental outcomes after late-onset meningitis among children born extremely preterm. JAMA Netw Open. 2022;5(12):e2245826. doi:10.1001/jamanetworkopen.2022.45826

Related Videos
Carissa Baker-Smith
Perry Roy, MD
Perry Roy, MD | Image Credit: Carolina Attention Specialists
Angela Nash, PhD, APRN, CPNP-PC, PMHS | Image credit: UTHealth Houston
Allison Scott, DNP, CPNP-PC, IBCLC
Joanne M. Howard, MSN, MA, RN, CPNP-PC, PMHS & Anne Craig, MSN, RN, CPNP-PC
Juanita Mora, MD
Natasha Hoyte, MPH, CPNP-PC
Lauren Flagg
© 2024 MJH Life Sciences

All rights reserved.