Evid Based Med 18:37-38 doi:10.1136/ebmed-2012-100651
  • Aetiology
  • Systematic review

Children surviving intrauterine and neonatal insults have a significant risk of long-term adverse neurodevelopmental outcomes

  1. Shripada C Rao1,2,3
  1. 1Department of Neonatal paediatrics, KEM Hospital for Women, Perth, Australia
  2. 2Department of Neonatal Paediatrics, Princess Margaret Hospital for Children, Perth, Australia
  3. 3Centre for Neonatal Research and Education, University of Western Australia, Perth, Australia
  1. Correspondence to: Sanjay Patole
    Department of Neonatal Paediatrics, KEM Hospital for Women, 374, Bagot Road, Subiaco, Perth, WA 6008, Australia; sanjay.patole{at}

Commentary on: [CrossRef][Medline][Web of Science]Google Scholar


Children who survive intrauterine and neonatal insults are known to be at risk for long-term adverse neurodevelopmental outcomes. However, the extent to which this occurs, especially in resource-poor nations, is not well documented. The systematic review by Mwaniki et al focuses on this important issue.


The authors searched Medline, CINAHL, EMBASE and the Cochrane Library for studies until June 2011 with no language restrictions. Unpublished studies and grey literature were searched using Dissertation Abstracts International and the WHO library. Reference lists of identified reports were searched for additional studies. Inclusion criteria were the following: occurrence of the insult during the intrauterine or neonatal period; insults verified with appropriate diagnostic criteria; follow-up of at least 6 months after neonatal insult; use of standardised tests/controls for neurodevelopmental assessment; and follow-up of at least 80% of survivors. Outcomes of interest were neurodevelopmental sequelae and their severity using well-accepted definitions. Pooled weighted mean risks and 95% CIs were calculated using a random-effect model. Heterogeneity was assessed using the I2 statistic and explored with meta regression analysis.


A total of 153 publications (N=22 161 surviving children) were included. Nearly 80% were from North America or Europe; 124 were prospective studies and 29 retrospective. The duration of follow-up varied from 6 months to >60 months. A total of 6851 (31%) of children had at least one impairment. Meta-analysis found that the pooled risk estimate of development of at least one sequelae in any domain was 37.0% (95% CI 27.0 to 48.0). Meta regression analysis found that the risks of sequelae did not differ with any of the four covariates tested: WHO region, period of data collection, duration of follow-up and study design.


The review addressed a clear question with appropriate inclusion criteria and outcomes of interest. Search terms were reported, but not in a reproducible format. The search terminologies appeared inadequate; many publications use ‘low birthweight’ instead of ‘preterm birth’ or ‘prematurity’. For example, a search of PubMed using the following MeSH terms, (‘Infant, Very Low Birth Weight’ OR ‘Infant, Low Birth Weight’ OR ‘Infant, Extremely Low Birth Weight’ OR ‘Infant, Premature’ OR ‘Infant, Premature, Diseases’) AND (‘Language Development Disorders’ OR ‘Developmental Disabilities’ OR ‘Learning Disorders’), revealed at least three additional large eligible studies (N=4000).1 , ,3 In addition, some studies on hypoxic ischaemic encephalopathy were missing,4 and it is not clear if they were identified but excluded. Inclusion of additional perinatal insults, such as meconium aspiration syndrome,5 persistent pulmonary hypertension of the newborn6 and neonatal surgery7 would have strengthened the review considering their adverse impact on long-term neurodevelopment. It is not clear if the authors of included studies were contacted for additional data/clarifications. Formal assessment for publication bias was not done; the quality of included studies was not assessed. Some key recommendations of the preferred reporting items for systematic reviews and meta-analysis (PRISMA) statement8 were not followed. For example, characteristics of the included studies were not provided. Reference citations of the studies that were excluded at the final stage would have been informative.

The findings of this systematic review may not be generalisable because the majority of the publications were from resource-rich countries. However, many emerging economies are reporting improved survival of high-risk neonates, and would find this review useful. In spite of some of the methodological limitations, this systematic review highlights the challenges of long-term child survival that accompany a decrease in neonatal mortality and can help policy makers identify and allocate adequate resources.


  • Competing interests None.