Abstract
The incidence and aetiologies of neonatal convulsions within a hospital population were reviewed over 10 years (1971-80 inclusive). Three periods, 1971-4 (12 904) live births). 1975-7 (13 304 live births), and 1978-80 (14 849 live births) during this decade are compared. The overall incidence of neonatal convulsions rose from 2-6/1000 and then to 8.6/1000 live births. Most convulsions were associated with cerebral hypoxia and 'fifth day fits'. Increasingly thorough investigations reduced the number of infants in whom no cause was identified to 11%. Although there was an appreciable overall reduction in mortality associated with convulsions, mortality among babies whose fits were caused by cerebral hypoxia remained over 50%.