Early Detection of Cerebral Palsy Among a High-risk Cohort in Low Resource Settings

ABSTRACT

Background and Objective(s) Delayed diagnosis of cerebral palsy (CP) limits access to early interventions when the infant brain has the most neuroplastic potential, particularly in low-and middle-income countries (LMICs). The 2017 clinical guideline on the early diagnosis of CP outlines best practice tools to support diagnosis. We aimed to assess the feasibility of implementing these tools for early detection of CP in Bangladesh. Study Design: Prospective cohort study. Study Participants & Setting: Neonates admitted to a regional tertiary hospital neonatal intensive care unit (NICU) in Bangladesh with major risk factors for CP (preterm birth, hypoxic ischemic encephalopathy/neonatal encephalopathy (HIE/NE), neonatal sepsis and/or severe jaundice/kernicterus) were enrolled. Materials/Methods: A physician identified eligible neonates via physical assessment, medical record review and parent interview using a risk factor questionnaire developed for this study. General Movements Assessment (GMA) were completed at the time of recruitment (writhing period) and 13 weeks corrected age (fidgety age);neuroimaging data collected from the NICU;and Hammersmith Infant Neurological Examination (HINE) conducted at 12 months corrected age. Due to the impact of COVID-19, a proportion of the cohort were not able to have GMA fidgety videos completed and the first HINE assessment was delayed to 12 months. GMA data is not currently reported in this . Result(s) A total of 227 high risk neonates were recruited between November 2019 to March 2020. All neonates had evidence for prematurity and infection/sepsis on physical examination, 83.7% (n=190) had HIE/NE and 14.5% (n=33) had severe jaundice/kernicterus. Only 1.8% (n=4) had cranial ultrasound and none had magnetic resonance imaging. Of the surviving (76.7%, n=174) infants, 77.0% (n=134) were assessed at 12 months. Among them, writhing videos and fidgety videos were previously collected for 100% (n=134) and 29.9% (n=40) respectively. At 12 months, 32.1% (n=43) infants were identified to have CP of whom 90.7% (n=39) infants had global HINE score <66 (sensitivity 90.7% and specificity 97.8%). Conclusions/Significance: Despite study attrition and the impact of COVID-19, it was feasible to collect GMA videos in inpatient setting and infants at risk of CP were diagnosed as early as 12 months in a LMIC. Use of the structured risk factor questionnaire and adherence to best practice guidelines ensured a highly sensitive screening process and diagnostic outcomes. Our interim findings demonstrate the scope of this simple and scalable protocol in supporting clinicians for the early identification of infants with CP to facilitate early intervention and shared decision-making with families for best outcomes in LMICs.