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

Abstract

Background: Few studies have reported the long-term consequences of bacterial meningitis during infancy, and studies that have been done usually do not include a comparison cohort. We aimed to assess short-term and long-term risk of mortality, neurodevelopmental impairment (NDI), and health-care use and household income in cohorts of children with and without a history of bacterial meningitis during infancy in Denmark and the Netherlands.

Methods: In this nationwide cohort study, infants with a history of bacterial meningitis before age 1 year were identified through the Danish Medical Birth Registry and Danish National Patient Registry using International Classification of Diseases (ICD)-10 codes and through the Netherlands Reference Laboratory for Bacterial Meningitis. Infants were matched (1:10) by sex and birth month and year to a comparison cohort of the general population without a history of bacterial meningitis. We analysed mortality using Cox proportional hazards regression. In Denmark, diagnoses of NDIs were based on ICD-10 codes; in the Netherlands, special educational needs were used as a functional NDI outcome. Risk ratios (RRs) of NDIs were estimated using modified Poisson regression. We also analysed long-term health-care use in Denmark and household income in both countries. All regression analyses were adjusted for sex and year of birth, and stratified by pathogen whenever sample size allowed.

Findings: We included 2216 children with a history of bacterial meningitis (570 [25·7%] in Denmark between Jan 1, 1997, and Dec 31, 2018, and 1646 [74·3%] in the Netherlands between Jan 1, 1995, and Dec 31, 2018), matched to 22 127 comparison cohort members. Median age at diagnosis was 2·8 months (IQR 0·4-7·1) in Denmark and 4·3 months (0·7-7·4) in the Netherlands. Mortality risks within 3 months after disease onset were 3·9% (95% CI 2·6-5·8%) in Denmark and 5·9% (4·7-7·0) in the Netherlands, compared with 0·0% (p<0·0001) and 0·1% (p<0·0001) in the comparison cohorts. Survivors had an increased risk of moderate or severe NDIs at age 10 years (RR 5·0 [95% CI 3·5-7·1] in Denmark and 4·9 [4·0-6·2] in the Netherlands) compared to children in the comparison cohort, particularly after pneumococcal and group B streptococcal meningitis. In Denmark, a history of bacterial meningitis was associated with increased health-care use in the 10 years following diagnosis (rate ratio 4·5 [95% CI 3·9-5·2] for outpatient visits and 4·1 [3·6-4·7] for hospital admissions).

Interpretation: Our study shows increased risk of mortality in the short and long term, a five times increase in risk of NDIs, and increased health-care use after bacterial meningitis during infancy. Together with context-specific incidence data, our results can advance pathogen-specific estimation of the meningitis burden and inform service provision at the individual and population level.

Funding: Bill & Melinda Gates Foundation, the Stichting Remmert Adriaan Laan Fonds, and the Netherlands Organisation for Health Research and Development.

Figure 1

Participant selection Flowchart of the inclusion of children in the bacterial meningitis cohort and comparison cohort in Denmark (A) and the Netherlands (B). ICD=International Classification of Diseases. NDI=neurodevelopmental impairment.

Figure 2

Median age of onset of meningitis Median age of onset in the cohort with a history of meningitis from Denmark (A) and the Netherlands (B), by causative agent. Causative agents include Streptococcus agalactiae, Escherichia coli, Streptococcus pneumoniae, Neisseria meningitidis, Haemophilus influenzae, and other (ie, less common, gram-positive, and gram-negative bacteria). Error bars represent minimum and maximum range.

Figure 3

Risk of NDIs in children with and without a history of bacterial meningitis in Denmark and the Netherlands Risk of NDIs in Denmark (A) and in the Netherlands (B), by causative agent. Causative agents include Streptococcus agalactiae, Streptococcus pneumoniae, and Neisseria meningitidis. Due to low event numbers in relation to data protection regulations, data on NDIs in children with N meningitidis meningitis at age 5 years in the Netherlands and in children with Haemophilus influenzae and Escherichia coli meningitis at age 5 years and 10 years in both countries could not be reported. NDI=neurodevelopmental impairment.

Figure 4

Domain-specific risk of NDIs in children with and without a history of meningitis in Denmark Risk of NDIs at age 5 years (A) and 10 years (B), by domain. NDI=neurodevelopmental impairment.