Predictive value of initial intracranial pressure for refractory intracranial hypertension in persons with traumatic brain injury: A prospective observational study

Objective: To prospectively investigate the predictive value of initial intracranial pressure (ICP) for refractory intracranial hypertension and outcomes in persons with diffuse traumatic brain injury (TBI). Methods: A prospective observational study was conducted in 107 adult persons with diffuse TBI (Marshall CT Class II-IV). Initial ICP was defined as the first ICP recorded in the operating room. Refractory intracranial hypertension was defined as ICP increases to more than 30mmHg and/or reduces in cerebral perfusion pressure to less than 60mmHg for a period longer than 15 minutes and failure to respond to the maximum medical treatment. Baseline demographics and injury-specific data were recorded. Multiple logistic regression models were used to determine independent risk factors for refractory intracranial hypertension and unfavourable outcomes. A receiver-operating characteristic (ROC) curve was then drawn. Results: The initial ICP allowed for a better refractory intracranial hypertension prediction (ROC area 0.868; 95% CI 0.799-0.937) than the Marshall Classification (ROC area 0.670; 95% CI 0.569-0.772) or Rotterdam Classification scores (ROC area 0.679; 95% CI 0.5770.780). An initial ICP value higher than 20mmHg had 83% sensitivity and 83% specificity, whereas an initial ICP value higher than 25mmHg had 64% sensitivity and 92% specificity for refractory intracranial hypertension. A multivariable logistic regression model showed that any 5mmHg pressure increase in a patient with initial ICP led to 2.884-times higher odds of refractory intracranial hypertension (95% CI 1.893-4.395; p<0.001). Head Abbreviated Injury Scale score, initial Glasgow Coma Scale (GCS) and initial GCS motor scores were not predictive of refractory intracranial hypertension (p>0.05). Conclusion: For persons with diffuse TBI, the initial ICP provides great prognostic discrimination and is an independent predictor of refractory intracranial hypertension.