Article Text
PDF
Statistics from Altmetric.com
Gale C, Statnikov Y, Jawad S, et al. Neonatal brain injuries in England: population-based incidence derived from routinely recorded clinical data held in the National Neonatal Research Database. Arch Dis Childhood – Fetal and Neonatal Edition 2018;103:F301–6.
The authors have identified an error in data extraction that resulted in the following: (1) the inclusion of babies with Grade 1 HIE who did not receive therapeutic hypothermia; (2) the exclusion of babies with central nervous system infection after 48 hours after birth. A data transcription error resulted in incorrect numbers of exclusions. Therefore the following sections are corrected as outlined below.
Abstract results
The results section of the Abstract should be as follows: ‘In 2010, the lower estimate of the rate of ‘Brain injuries occurring at or soon after birth’ in England was 3.99 and the upper estimate was 4.66; in 2015, the rate was 4.52 (4.36 to 4.68). For preterm infants, the population incidence in 2015 was 24.45 (23.12 to 25.85) and for term infants 2.91 (2.78 to 3.05). Hypoxic ischaemic encephalopathy was the largest contributor to term brain injury and intraventricular/periventricular haemorrhage was the largest contributor to preterm brain injury.’
Results
The first and second paragraph of the Results section of the manuscript are corrected to ‘The number of infants admitted to neonatal units contributing data to the NNRD increased from 64 375 in 2010 to 88 785 in 2015; the annual rate of brain injuries after exclusions in England in 2010 was between 3.87 and 4.44 per 1000 live births; in 2015 to the rate was 4.52 (95% CI 4.36 to 4.68) per 1000 live births (table 1); annual data for 2010 to 2015 are presented table 1.’
Infants in England (all gestational ages) with a diagnosis of brain injury to before and after exclusion of conditions leading to brain injury prior to birth; data for 2010 and 2011 are adjusted to account for the incomplete coverage of the NNRD during those years; from 2012 onwards the NNRD has complete population coverage of neonatal admissions in England so no adjustment was necessary and data are presented as a rate (95% CI)
‘The annual rate of brain injuries among term infants (≥37 gestational weeks) in England in 2015 was 2.91 (95% CI 2.78 to 3.05) per 1000 live term births; data for term infants born over the period 2010 to 2015 are presented in table 2. The annual rate of brain injuries among preterm infants (<37 gestational weeks) in England in 2015 was 24.45 (95% CI 23.12 to 25.85) per 1000 live preterm births; data for preterm infants between 2010 and 2015 can be found in table 3.’
Term (≥37 gestational weeks) infants in England with a diagnosis of brain injury to before and after exclusion of conditions leading to brain injury prior to birth; data for 2010 and 2011 are adjusted to account for the incomplete coverage of the NNRD during these years; from 2012 onwards the NNRD has complete population coverage of neonatal admissions in England so no adjustment was necessary and data are presented as a rate (95% CI)
Preterm (<37 gestational weeks) infants in England with a diagnosis of brain injury to before and after exclusion of conditions leading to brain injury prior to birth; data for 2010 and 2011 are adjusted to account for the incomplete coverage of the NNRD during these years; from 2012 onwards the NNRD has complete population coverage of neonatal admissions in England so no adjustment was necessary and data are presented as a rate (95% CI)
Infants in England with conditions leading to brain injury at or soon after birth; infants can be diagnosed with more than one condition so the sum of conditions for each year will not match data given in tables 1–3; P/IVH: periventricular/intraventricular haemorrhage, CI: confidence interval
Discussion
The third paragraph of the Discussion section of the manuscript is corrected to: ‘The novelty of the measure brain injuries that occur at or soon after birth, and the source of the data, the NNRD which is formed from routinely recorded clinical information makes it necessary to consider how incidence rates of individual conditions we report compare with other published data. The annual incidence rates for moderate and severe HIE of between 1.6 and 1.9 per 1000 live births are consistent with other reported rates of neonatal encephalopathy of between 0.77–3.8 per 1000 live births in low neonatal mortality regions such as the United Kingdom and the United States of America. When considering neonatal intracranial haemorrhage, published data commonly reported incidence separately for term and preterm infants. A 30-year-old, single centre study from the USA reported a regional incidence of 2.7 per 10 000 live births for symptomatic intracranial haemorrhage in term infants, which is comparable with the population incidence of 1.5 to 1.9 per 10 000 term births that we report. For preterm infants born at 22–31 weeks gestational age, comparable population-level incidence data for intraventricular/periventricular haemorrhage from the national French EPIPAGE cohort are 3.8% for grade 3% and 3.3% for grade four intra/periventricular haemorrhage. In the same gestational age band, we report annual incidence rates between 5.8% and 6.1% for a composite including grades 3 and 4 intraventricular/periventricular haemorrhage. We report annual incidence rates for neonatal or perinatal stroke of between 0.11 and 0.15 per 1000 live births, which is similar to the estimated minimum incidence rate of 0.10 per 1000 live births reported by a prospective, population-based study from Canada. The annual incidence of neonatal central nervous system infection we report of 0.56 to 0.82 per 1000 live births is similarly in agreement with the population-level incidence rate for neonatal meningitis in England and Wales of 0.39 per 1000 live births (1996–1997) reported by the British Paediatric Surveillance Unit (BPSU). Similarly, annual rates of bilirubin encephalopathy reported here of between 0.3 and 1.2 per 100 000 live births are comparable with BPSU population surveillance rates of 0.9 per 100 000 live births (2003–2005). EPIPAGE 2 (2011) report an incidence among 23 to 34 gestational week infants of 1.8% for cystic periventricular leukomalacia; in the same gestational age group, we report annual rates of between 1.13% and 1.27%. Finally population-based studies of neonatal seizures over the last 30 years report incidence rates between 1.8 and 3.5 per 1000 live births, results that are again in keeping with our annual rates of 1.9 to 2.2 per 1000 live births.’
The following 4 tables (tables 1–4) are corrected as below.