Patients with EDH and GCS score <9 should undergo surgical evacuation. Surgery is also recommended for patients with EDH volume ≥30 cc, thickness ≥15 mm or midline shift of ≥5 mm regardless of GCS. Patients managed nonoperatively should undergo serial CT imaging and frequent neurological monitoring where emergent neurosurgical treatment is available. Temporal location increases the risk of deterioration and therefore, threshold for surgery should be kept lower for patients with temporal EDH. Time from neurological worsening to EDH evacuation correlates more with outcomes than time from injury to evacuation. Therefore, in patients undergoing initial non-surgical treatment, neurological deterioration as defined by worsening level of consciousness, abnormal papillary reflex, and appearance of new focality or worsening of existing focal deficits should prompt urgent surgical evacuation.

The indications for surgery in acute traumatic SDH include clot thickness >10 mm or midline shift >5 mm on initial CT imaging regardless of GCS. In patients with smaller SDH, GCS <9 or decrease in GCS by 2 points or more since presentation, presence of papillary abnormality or persistent elevation of ICP >20 mmHg should prompt surgical evacuation.

Patients with GCS of <9 with frontal or temporal hematoma of >20 cc with midline shift of ≥5 mm or cisternal compression and those with any hematoma of >50 cc should undergo evacuation. Also, patients with neurological deterioration thought to be related to intraparenchymal hematoma should also be treated surgically [4]. Neurologically stable patients with parenchymal contusion showing no significant mass effect on CT scan can be managed nonoperatively with close monitoring and serial imaging. Patients with medically refractory intracranial hypertension from diffuse cerebral edema may be treated with bifrontal or hemispheric decompressivecraniectomy.

Invasive ICP monitoring is not routinely indicated in all cases of TBI and the risks of monitoring such as infection and bleeding must be weighed against the benefit of additional information obtained by such monitoring. Patients with severe TBI with abnormal CT scan and GCS <9 should undergo ICP monitoring as they have high likelihood of transient or persistent elevation in ICP resulting in compromise in cerebral perfusion pressure and inability of serial clinical examinations to identify subtle changes related to intracranial hypertension due to poor baseline neurological status. Normal neuroimaging especially during early hours of TBI does not rule out intracranial hypertension. Patients with two or more of the following factors should undergo ICP monitoring in the absence of abnormal findings on CT scan of the head: age above 40 years, motor posturing, systolic blood pressure <90 mmHg. The ICP threshold above which interventions aimed at lowering the ICP should be implemented is unclear at this time. ICP of above 20 mmHg is generally recommended as the treatment threshold by several guidelines. Cerebral perfusion pressure (CPP = MAP-ICP) based threshold may provide for a more physiologically rational parameter and should be used in conjunction with ICP parameter with goal CPP >60–65 mmHg. Different devices used for ICP monitoring use any of available technologies such as fiberoptic sensor, microchip with internal strain gauge, air pouch or fluid filled catheter connected to pressure transducer. Devices also differ in terms of location of tip such as subdural, epidural, subarachnoid, intraparenchymal and intraventricular. Of these, fluid-filled transduced ventriculostomy catheter provides the most accurate ICP value and also allows for therapeutic CSF drainage and is therefore, preferred over other modalities.