Klein, Jonathan M.

ImportanceLate-onset meningitis (LOM) has been associated with adverse neurodevelopmental outcomes in children born extremely preterm.ObjectiveTo report the incidence of LOM during birth hospitalization and neurodevelopmental outcomes at 18 to 26 months’ corrected age.Design, Setting, and ParticipantsThis cohort study is a secondary analysis of a multicenter prospective cohort of children born at 22 to 26 weeks’ gestation between 2003 and 2017 with follow-up from 2004 to 2021. The study was conducted at 25 Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network centers.ExposuresCulture-confirmed LOM.Main Outcomes and MeasuresIncidence and microbiology of LOM (2003-2017); lumbar puncture (LP) performance in late-onset sepsis (LOS) evaluations (2011-2017); composite outcome of death or neurodevelopmental impairment (NDI; 2004-2021).ResultsAmong 13 372 infants (median [IQR] gestational age, 25.4 [24.4-26.1] weeks; 6864 [51%] boys), LOM was diagnosed in 167 (1%); LOS without LOM in 4564 (34%); and neither LOS nor LOM in 8641 (65%). The observed incidence of LOM decreased from 2% (95% CI, 1%-3%) in 2003 to 0.4% (95% CI, 0.7%-1.0%) in 2017 (P < .001). LP performance in LOS evaluations decreased from 36% (95% CI, 33%-40%) in 2011 to 24% (95% CI, 21%-27%) in 2017 (P < .001). Among infants with culture-confirmed LOS, LP performance decreased from 58% (95% CI, 51%-65%) to 45% (95% CI, 38%-51%; P = .008). LP performance varied by center among all LOS evaluations (10%-59%, P < .001) and among those with culture-confirmed LOS (23%-79%, P < .001). LOM occurred in the absence of concurrent LOS in 27 of 167 cases (16%). The most common LOM isolates were coagulase-negative Staphylococcus (98 [59%]), Candida albicans (38 [23%]), and Escherichia coli (27 [16%]). Death or NDI occurred in 22 of 46 children (48%) with LOM due to coagulase-negative Staphylococcus, 43 of 67 (64%) due to all other bacterial pathogens, and 26 of 33 (79%) due to fungal pathogens. The adjusted relative risk of death or NDI was increased among children with LOM (aOR, 1.53; 95% CI, 1.04-2.25) and among those with LOS without LOM (aOR, 1.41; 95% CI, 1.29-1.54) compared with children with neither infection.Conclusions and RelevanceIn this cohort study, LP was performed with decreasing frequency, and the observed incidence of LOM also decreased. Both LOM and LOS were associated with increased risk of death or NDI; risk varied by LOM pathogen. The full association of LOM with outcomes of children born extremely preterm may be underestimated by current diagnostic practices.

AbstractAimTo characterise the relationship between bronchopulmonary dysplasia (BPD) severity and cognition in the post‐surfactant era.MethodsThis was a single‐centre retrospective analysis of a cohort of infants born 2009–2012. Inclusion criteria were as follows: admission within 48 hours of birth, gestational age 22‐0/7–31‐6/7 weeks, birthweight 400–1500 g and Bayley Scales of Infant and Toddler Development‐III testing at 18–26 months corrected age. Infants (n = 151) were classified by BPD severity with the NIH Workshop definition. Generalised linear modelling and multivariate logistic regression were performed.ResultsBayley cognitive score was not associated with BPD severity in univariate (p = 0.053) or multivariate (p = 0.503) analysis. About 27% of infants with no/mild BPD, 33% of infants with moderate BPD and 40% of infants with severe BPD had a cognitive score <85. There was no difference in the odds of cognitive score <85 based on BPD severity in univariate (p = 0.485) or multivariate analysis (p = 0.225). All infants with cognitive score <70 had severe BPD, although the association between cognitive score <70 and BPD severity was not significant.ConclusionWe found no independent effect of BPD severity level on cognition. The likelihood of a cognitive score <85 was not associated with BPD severity.