Long-term outcomes of survivors of neonatal insults: A systematic review and meta-analysis

Abstract

Background: The Millennium Developmental Goals ensured a significant reduction in childhood mortality. However, this reduction simultaneously raised concerns about the long-term outcomes of survivors of early childhood insults. This systematic review focuses on the long-term neurocognitive and mental health outcomes of neonatal insults (NNI) survivors who are six years or older.

Methods: Two independent reviewers conducted a comprehensive search for empirical literature by combining index and free terms from the inception of the databases until 10th October 2019. We also searched for additional relevant literature from grey literature and using reference tracking. Studies were included if they: were empirical studies conducted in humans; the study participants were followed at six years of age or longer; have an explicit diagnosis of NNI, and explicitly define the outcome and impairment. Medians and interquartile range (IQR) of the proportions of survivors of the different NNI with any impairment were calculated. A random-effect model was used to explore the estimates accounted for by each impairment domain.

Results: Fifty-two studies with 94,978 participants who survived NNI were included in this systematic review. The overall prevalence of impairment in the survivors of NNI was 10.0% (95% CI 9.8-10.2). The highest prevalence of impairment was accounted for by congenital rubella (38.8%: 95% CI 18.8-60.9), congenital cytomegalovirus (23.6%: 95% CI 9.5-41.5), and hypoxic-ischemic encephalopathy (23.3%: 95% CI 14.7-33.1) while neonatal jaundice has the lowest proportion (8.6%: 95% CI 2.7-17.3). The most affected domain was the neurodevelopmental domain (16.6%: 95% CI 13.6-19.8). The frequency of impairment was highest for neurodevelopmental impairment [22.0% (IQR = 9.2-24.8)] and least for school problems [0.0% (IQR = 0.0-0.00)] in any of the conditions.

Conclusion: The long-term impact of NNI is also experienced in survivors of NNI who are 6 years or older, with impairments mostly experienced in the neurodevelopmental domain. However, there are limited studies on long-term outcomes of NNI in sub-Saharan Africa despite the high burden of NNI in the region.

Trial registration: Registration number: CRD42018082119.

Fig 1. Flow chart showing the selection of studies on long-term outcomes of neonatal insults.

Fig 2. Individual and pooled estimates and 95% confidence intervals for random-effects model examining the long-term outcomes of fetal growth restriction.

I2- heterogeneity statistic; ES- effect size; %—percent; sub-groups with (I2 = . %, p = .) indicate that the number of studies were too few for the estimates to be calculated.

Fig 3. Individual and pooled estimates and 95% confidence intervals for random-effects model examining the long-term outcomes of Preterm Birth.

I2- heterogeneity statistic; ES- effect size; %—percent; sub-groups with (I2 = . %, p = .) indicate that the number of studies were limited for the estimates to be calculated.

Fig 4. Individual and pooled estimates and 95% confidence intervals for random-effects model examining the long-term outcomes of Neonatal Jaundice.

Fig 5. Individual and pooled estimates and 95% confidence intervals for random-effects model examining the long-term outcomes of Congenital Rubella.

I2- heterogeneity statistic; ES- effect size; %—percent; sub-groups with (I2 = . %, p = .) indicate that the number of studies were limited for the estimates to be calculated.

Fig 6. Individual and pooled estimates and 95% confidence intervals for random-effects model examining the long-term outcomes of Hypoxic-ischemic encephalopathy.