Background: Patients with severe traumatic brain injury (TBI) require aggressive management to prevent secondary brain injury. "Preemptive" craniectomy (CE)--craniectomy performed as a primary procedure in conjunction with craniotomy--has been used as prophylaxis for secondary injury, but the indications and outcomes of craniectomy used for this purpose are not well defined.
Methods: To evaluate the role of CE in the management of TBI, we retrospectively reviewed 62 consecutive patients who underwent CE in a 78-month period at our level I trauma center. A cohort of patients who underwent craniotomy only (CO) during this period was compared with the CE group for TBI patterns, indications for operation, and outcomes. Multivariable logistic regression and matched propensity score analysis were used to test the association between CE and survival. The rate of CE was determined by individual neurosurgeons.
Results: Of 197 patients with brain injuries who underwent craniotomy, 62 (31.5%) had CE and 135 (68.5%) had CO. Mean age for CE versus CO was 41 years versus 51 years (p < 0.01). Mean admission Glasgow Coma Score was lower in CE versus CO (7.6 vs. 11.8, p < 0.001); Injury Severity Score was higher (30.2 vs. 26.3, p < 0.01). The indication for operation for CE compared with CO was subdural hematoma in 41 (66.1%) versus 87 (64.4%, p = 0.82), epidural hematoma in 2 (3.2%) versus 26 (19.3%, p < 0.01), and cerebral contusion or hematoma in 15 (24.2%) versus 8 (5.9%, p < 0.001). Postoperative intracranial pressure was monitored in 48 (77.4%) CE and 44 (32.6%) CO patients (p < 0.001). Intracranial pressure <20 was maintained in 26 (54.2%) after CE and in 31 (70.5%) after CO (p = 0.12). In the CE group, 26 (42%) died compared with 31 (26%, p < 0.01) in the CO group. When adjusted for severity of injury, however, there was no significant difference in mortality between the two groups (p = 0.134). The CE rate obtained by a neurosurgeon varied from 8.6% to 75.0% (p < 0.001).
Conclusion: CE was used in patients with more severe injuries, and particularly in those with more severe head injuries. When adjusted for injury severity, CE was not associated with worsened survival, and therefore may reasonably be included in the armamentarium of neurotrauma care. Use of CE by our neurosurgeons, however, varied significantly. These findings underscore the need for practice guidelines based on randomized trials to fully evaluate the role of CE in the management of TBI.