Predicting cerebral palsy and 18-month neurodevelopmental outcome in infants with presumed hypoxic ischaemic encephalopathy : role of general movements assessment and early neurological examination

Abstract

INTRODUCTION : General movements assessment (GMA), including the Motor Optimality Score—Revised (MOS-R) and the Hammersmith Infant Neurological Examination (HINE), has been shown in different settings to predict cerebral palsy (CP) and delayed neurodevelopment with high accuracy. However, their combined predictive ability has not been fully evaluated in infants with presumed hypoxic–ischaemic encephalopathy (HIE). OBJECTIVE : This study aimed to assess the predictive ability of combined GMA, MOS-R, and HINE at 3 months in term or near-term infants diagnosed with presumed HIE, for neurodevelopmental outcome at 18 months. METHODS : A cohort of presumed HIE infants treated with therapeutic hypothermia (TH) underwent GMA, MOS-R, and HINE at 12–15 weeks, and neurodevelopmental assessments using the Bayley Scales of Infant and Toddler Development Third Edition (BSID-III) at 9–12 and at 18–24 months of age. Combined early assessments were analysed for their predictive ability across different domains on the BSID-III. RESULTS : Twenty-four infants were included; 7 (29%) had both 12-month and 18-month BSID-III assessments, 12 (50%) were seen only at 12 months, and 5 (21%) only at 18 months. Two infants with absent fidgety movements (FMs) and poor motor repertoire were later diagnosed with CP or showed delays in two domains on the BSID-III assessment at 18 months. While most infants had some abnormality in the MOS-R categories, only absent FMs and abnormal finger variability showed some association with the 18-month BSID-III assessment on univariate analysis. Of the four infants classified as at risk for CP on the HINE at 3 months, two had some motor abnormalities at 18 months. Combining the GMA, MOS-R, and HINE had high sensitivity and negative predictive value (100%), but low specificity (0–17.6%) and positive predictive value (6.2%–25%) for the BSID-III outcome. CONCLUSION : Combining GMA, MOS-R, and HINE was highly sensitive in this cohort, but had low specificity. This may lead to overdiagnosis, but it may be a useful screening tool for identifying typically developing infants who do not need intensive follow-up.