Intraoperative applications of intracranial pressure monitoring in patients with severe head injury.

From December 2002 to January 2004, 30 patients (20 men and 10 women; mean age 36.8 years [+/- 14.9 years]) with preoperative Glasgow Coma Scale scores of 8 or less underwent emergency haematoma evacuation surgery and continuous intracranial pressure (ICP), cerebral perfusion pressure (CPP) and mean arterial blood pressure monitoring to determine ICP and CPP thresholds to predict patient outcomes. Receiver-operating characteristic (ROC) curves were plotted. Using the ROC curve, the diagnostic accuracy is given by the area under the curve and at the point on the curve farthest from the diagonal, which indicates the threshold value. The results showed that the initial ICP for unfavourable outcomes was 47.4 +/- 21.4 mmHg, resulting in a CPP of 22.8 +/- 12.83 mmHg. The initial ICP for favourable outcomes was 26.4 +/- 10.1 mmHg, resulting in a CPP of 48.8 +/- 13.4 mmHg. The CPP had the largest area under the ROC curve in all stages of the operation, corresponding to intraoperative CPP thresholds of 37 mmHg (initial), 51.8 mmHg (intraoperative) and 52 mmHg (after scalp closure). The ROC curve analysis showed that CPP was a better predictor of outcome than ICP.