Objective-To analyse the occurrence and influence on outcome of secondary brain insults during neurointensive care of patients with subarachnoid haemorrhage. Methods-Sixty one seriously ill patients with subarachnoid haemorrhage with a poor neurological grade, acute hydrocephalus, or intraventricular and/or intracerebral haemorrhages, who were referred as emergency cases to the neurosurgical intensive care unit during 1990 and 1991, comprised the study patients. Results-The follow up performed according to the Glasgow outcome scale 14 months (median) later showed 23 patients with good recovery (38%), 11 with moderate disability (18%), seven with severe disability (11%), and two in a vegetative state (3%); 18 patients had died (30%). Clinical outcome was significantly related to the CT modified Hunt and Hess grade (P = 0.006). In total, 164 secondary brain insults (potentially avoidable factors) of various types were seen at the unit during the first seven days after the haemorrhage. Patients with a favourable outcome had significantly fewer secondary insults than patients with an unfavourable outcome (P = 0.0008). The occurrence of insults in each patient was related to the neurological grade (CT modified Hunt and Hess grade, P = 0.05), Multivariate analysis with the CT modified Hunt and Hess grade and the number of secondary brain insults during the first week as explanatory variables and favourable outcome as the dependent variable, showed that the number of complications was a significant independent predictor of favourable outcome (beta = -0.38, SE(beta) = 0.17, P = 0.03), whereas the CT modified Hunt and Hess grade did not reach significance (beta = -1.2, SE(beta = 0.81, P = 0.14). Conclusions-The clinical outcome after subarachnoid haemorrhage is at least paltry determined by the number of secondary insults. Therefore, vigorous attempts should be made to avoid all events that may potentially increase the risk of secondary cerebral ischaemia. Prospective studies must be initiated to define the Pole of ''priming'' of the brain and the impact of specific individual secondary insults in patients with subarachnoid haemorrhage.
