[Hemorrhages in the newborn].

Several considerations serve to exclude the study of retinal haemorrhage from the routine neurological assessment of the newborn: (1) the lack of any clearly demonstrated correlation between cerebral trauma and the incidence of retinal haemorrhage; (2) or even between subsequent ocular abnormalities and retinal haemorrhage; (3) the speedy absorption of retinal haemorrhage beginning before the infant becomes neurologically 'stable'; and (4) the wide variation in the reported incidence of retinal haemorrhage. Even so, these haemorrhages cannot be totally ignored in the pathophysiology of childbirth, and, as our knowledge of specific developmental defects increases, it is tempting to seek new correlates and to re-open an aspect of neuropaediatrics which has fallen into disrepute in the past. The principal method of examination has been direct ophthalmoscopy of the posterior pole. The data and the method itself are reviewed historically and in the light of personal experience.