부산대학교 도서관

In sub-Saharan Africa, the timing and nature of brain injury and their relation to mortality in neonatal encephalopathy (NE) is unknown. AIM: To evaluate cranial ultrasound (cUS) scans from term Ugandan infants with NE and unaffected controls for evidence of established/evolving brain injury. METHODS: Infants were recruited to a case-control study at Mulago Hospital, Kampala. Cases had NE (score >5). Controls were randomly selected unaffected term infants. All had cUS scans <36 h; surviving cases had day 4/5 scans. Minimally 11 views were reported blind to NE status. Scan injury definitions were evolving: extensive white matter (WM) and/or bilateral basal ganglia-thalamic (BGT) echogenicity and established: parenchymal cysts, atrophy, organising haemorrhage. RESULTS: Early scans showed significant abnormality in 20.4% (35/172) cases vs 1.0%(1/99) controls (p < 0.0001, mean age 11.2 h/8.1 h, respectively). Case scans at <18 h (120) still showed a high prevalence of abnormalities (16.7%). Case infants had 9.3% WM injury alone, 6.4% BGT injury alone and 4.7% BGT/WM injury. No established injury was seen. Fatality was significantly higher in cases with early cUS abnormalities vs those without (57.1%(20/35) vs 25.9%(35/135); OR 3.81(95% CI 1.76–8.25)). Serial scans were done in 88 mostly surviving NE infants; 74 had normal early scans with 35(47.3%) becoming abnormal by day 4/5. CONCLUSIONS: The high proportion of infants with early cUS abnormality plus absence of established antepartum injury implies the injury pathway often begins several hours before delivery in this population, with implications for therapeutic interventions considered effective early in the injury pathway. Early cUS abnormality in NE was a significant predictor for death. REFERENCES