The ability of early serial developmental assessment to predict outcome at 5 years following neonatal hypoxic-ischaemic encephalopathy
Introduction
Hypoxic-ischaemic encephalopathy (HIE) occurs following approximately 3 per 1000 term births. It remains one of the leading causes of neonatal death, with survivors at risk of significant neurological disability including cerebral palsy, epilepsy, intellectual disability, and sensory loss [1]. Longer term outcome studies have reported more specific neurodevelopmental sequelae that may not emerge until after 24 months [2], [3], [4]. These include delayed educational attainment and specific deficits in language, sensorimotor and selected memory domains [5], [6]. In particular, working memory, long-term episodic and specifically verbal learning and retention are lower in children with moderate and severe HIE [7], [8].
The challenge lies in the early identification of children for whom later-emerging deficits are more likely to occur. Research studies have outlined the importance of follow-up of affected children into their school years to capture the full impact of early injury [9].
The ability of early standardised developmental assessment to predict later cognitive outcome is clear at the extremes of developmental progress because a severe insult has a suppressor effect across multiple development domains [10]. However the prediction of more subtle deficits, most amenable to intervention, poses a greater challenge. Prediction studies observing clinical samples of infants tend to report low sensitivity and high specificity for later intellectual difficulties [11], [12], [13]. Development progresses rapidly in the first years of life and the range of “normal development” is wide. Little information is available regarding the temporal stability of repeated early developmental assessments [14]. Stability is likely to depend upon the patterns of development in the underlying domains of motor skills, language and cognition. The optimal time to assess these domains in early childhood remains unclear and the importance of serial assessment is acknowledged [15].
The use of the Bayley Scales [16], predominates in this field, and knowledge of predictive properties of other standardised development assessments such as the Griffiths scales is less well understood, especially for infants following neonatal HIE.
This study sought to record the temporal stability of serial developmental assessment in children following HIE at 6, 12 and 24 months and to determine the role of age at assessment for the prediction of (i) cognitive abilities and (ii) overall outcome at five years.
Section snippets
Participants
Infants with hypoxic-ischaemic encephalopathy (HIE) were prospectively recruited between May 2003 and December 2005, from a maternity service with approximately 6000 deliveries per annum [17]. Children were recruited prior to the introduction of therapeutic hypothermia and were therefore not cooled. Ethical approval was granted from the Clinical Research Ethics Committee of the Cork Teaching Hospitals, Cork, Ireland. Inclusion criteria were infants of ≥ 37 week gestations who met at least two of
Results
In total, 60 children were recruited to the study. The modified Sarnat grades assigned at 24 h were: 27 mild, 21 moderate, and 12 severe. Of these, by 24 months, 6 (all severe) had died, 1 (mild) was lost to follow-up and two had been excluded due to confounding medical conditions – one (mild) with a repaired congenital diaphragmatic hernia and the other (moderate) with a suspected genetic syndrome due to dysmorphic features and generalised hypotonia. The remaining 51 children were contacted at
HIE birth cohort - outcome at five years
Of the original cohort of 60 children born with HIE, the outcome of 53 children was known at 5 years, with 47.2% (25/53) having an abnormal outcome. Six died (5 in the neonatal period and 1 before 24 months), eight had cerebral palsy (3 with intellectual disability), one had hearing impairment, four had an IQ 1SD below the mean (3 with concurrent diagnoses: 1 ADHD, 2 Lang D/O) four further diagnosed neurodevelopmental disorders (1 ASD, 1 ADHD, 1 DCD, 1 Lang D/O) and two attending MD EI teams.
Correlation between early developmental assessment and five year outcome
We next examined the strength of association between the Griffiths' developmental assessments undertaken at 6, 12 and 24 months with IQ outcome at five years. GDQ at 6, 12 and 24 months correlated significantly with FSIQ, VIQ and PIQ at five years (Table 2), and the strength of correlation increased over time. The highest correlation was found between the composite Griffiths DQ at 24 months and FSIQ at 5 years (R = 0.64, p < 0.001), explaining 41% of the shared variance.
Temporal stability of early developmental assessments
The stability of Griffiths developmental quotients (GDQ) gathered at 6, 12 and 24 months for each infant were tracked across time, and are presented with both five-year IQ and overall outcome in Table 3.
Overall, twenty-nine children tested normal on the Griffiths - either on all three occasions (n = 16) or on all of the occasions that they attended (n = 13) in early childhood. This was 100% predictive of a normal cognitive (IQ) outcome. However despite a normal FSIQ, 3/29 (10.3%) of these had an
Predictive power of early childhood assessment at each time point for 5 year IQ
We examined the predictive power of a normal GDQ score for a normal five-year IQ at each time point (Table 4a). As expected, the best prediction was seen at 24 months with an AUROC = 0.941 (p = 0.001). The 24 month normal GDQ was 100% predictive of normal 5 year IQ. At all time-points the early GQ scores displayed moderate to high AUROC accuracy values (see Fig. 2), with higher sensitivity and lower false negative rates. A higher proportion of children with a normal five-year IQ, had normal early
Predictive Power of early childhood assessment at each time point for overall outcome
For overall outcome, all children with a normal five year outcome had normal 6 and 12 month GDQ. Furthermore, an abnormal 6 and 12 month GDQ score was 100% predictive of abnormal outcome. Surprisingly GDQ at 12 months was most closely related to overall outcome suggesting that this was a more global measure of injury (see Table 4b and Fig. 3), whilst the 24 month assessment was more closely linked to 5 year IQ.
Discussion
We have described the five year IQ and overall outcome in a cohort of five-year-old children following neonatal HIE. We confirmed the importance of undertaking repeated developmental assessments in the early years since one-third of children with HIE in our cohort had unstable development scores across early childhood. We have also confirmed the utility of early developmental assessment in the prediction of overall outcome at five years, and the importance of long-term monitoring for those with
Conflict of Interest
DM, CAR and GB nothing to declare; COC is currently a member of the Executive Board and Trustee of ARICD which is a charity which develops the Griffiths Mental Development Scales. The decision to use the GMDS for two year follow-up predated any involvement by COC in this project.
Acknowledgements
Funding for this research was supported by a grant from the Health Research Board (HRB) of Ireland [grant number RP/2008/238]. The HRB had no involvement in the study design, collection or interpretation of data. Dr. Deirdre Murray received the Denis O′Sullivan Research Fellowship from University College Cork. The authors would like to thank Dr. Evonne Low and Dr. Louise Gibson for completing a number of the neurological assessments at age five. The authors thank sincerely all of the children
References (35)
- et al.
School performance of survivors of neonatal encephalopathy associated with birth asphyxia at term
J. Pediatr.
(1989) - et al.
Hypoxic-ischemic encephalopathy: challenges in outcome and prediction
J. Pediatr.
(2011) - et al.
Instability of delay classification and determination of early intervention eligibility in the first two years of life
Res. Dev. Disabil.
(2014) - et al.
Structure and continuity of intellectual development in early childhood
Intelligence
(2009) - et al.
The incidence and severity of post-asphyxial encephalopathy in full-term infants
Early Hum. Dev.
(1985) - et al.
Childhood growth and development associated with need for full-time special education at school age
Eur. J. Pediatr. Neurol.
(2009) - et al.
Intra-individual stability of rate of gross motor development in full-term infants
Early Hum. Dev.
(1998) Neurology of the Newborn
(2008)- et al.
Neuropsychological and educational problems at school age associated with neonatal encephalopathy
Arch. Dis. Child. Fetal Neonatal Ed.
(2005) - et al.
Long-term follow-up of term neonates with perinatal asphyxia
Clin. Perinatol.
(1993)
Long-term outcome after neonatal hypoxic-ischaemic encephalopathy
Arch. Dis. Child. Fetal Neonatal Ed.
Neonatal watershed brain injury on magnetic resonance imaging correlates with verbal IQ at 4 years
Pediatrics
Specific memory impairment following neonatal encephalopathy in term-born children
Dev. Neuropsychol.
Brain and cognitive-behavioural development after asphyxia at term birth
Dev. Sci.
Presidential address. Prediction of function from infancy to early childhood: implications for pediatric psychology
J. Pediatr. Psychol.
Prediction of cognitive abilities at the age of 5 years using developmental follow-up assessments at the age of 2 and 3 years in very preterm children
Dev. Med. Child Neurol.
Predictive validity of early developmental assessments in identifying school-age cognitive deficits in children born preterm or very low birthweight: systematic review and meta-analysis
Arch. Dis. Child.
Cited by (22)
Motor function and white matter connectivity in children cooled for neonatal encephalopathy
2021, NeuroImage: ClinicalCitation Excerpt :Neonatal hypoxic-ischaemic encephalopathy (HIE), secondary to perinatal asphyxia, increases the risk of death and disability, including cerebral palsy (CP) (Marlow, 2005; O’Connor et al., 2017; Robertson et al., 1989).
Disrupted brain connectivity in children treated with therapeutic hypothermia for neonatal encephalopathy
2021, NeuroImage: ClinicalCitation Excerpt :Neonatal encephalopathy (NE), which often results from perinatal asphyxia, leads to a high risk of death or disability, including cerebral palsy (CP) (Azzopardi et al., 2014; Marlow, 2005; O’Connor et al., 2017; Robertson et al., 1989).
Sarnat Grading Scale for Neonatal Encephalopathy after 45 Years: An Update Proposal
2020, Pediatric NeurologyCitation Excerpt :The longevity of the study in part is because it was the first study of this type, is clinically simple as part of the physical examination that could be done by any bedside physician without special equipment, and has been repeated and validated by numerous independent clinical investigators in other centers in the United States and worldwide. This study also has been used as a standard in subsequent studies including prognosis2-6; it has been applied in electroencephalography (EEG) monitoring studies of HIE,7 in correlating neuroimaging studies,8 and as a patient selection criterion for therapeutic hypothermia.9,10 A recommendation on clinical criteria of neonatal encephalopathy was sponsored jointly by the American College of Obstetrics and Gynecology and the American Academy of Pediatrics in 2003 and again in 2014.11-13
Prospective assessment of early developmental markers and their association with neuropsychological impairment
2023, European Journal of Pediatrics