Role of decompressive craniectomy in the management of severe head injury with refractory cerebral edema and intractable intracranial pressure. Our experience with 48 cases

Abstract

Background: The effects of decompressive craniectomy in the treatment of severe head injury remain unclear. Only very few randomized studies relating to this topic exist in the literature, including a very small number of patients with no class I evidence.

Methods: We rretrospectively reviewed a series of 221 patients operated on for a head injury during a 25-month period. Of these, 48 patients underwent a decompressive craniectomy. All data available on patients' Glasgow Coma Scale score, pupil size and reaction, and intracranial pressure were collected and analyzed. The patients' outcome was evaluated by the Glasgow Outcome Scale (GOS) and the results compared with the data available in the Traumatic Coma Data Bank. Furthermore, the results were analyzed in respect of the time of surgical intervention (early or late), age, and the preoperative Glasgow Coma Score.

Results: Decompressive craniectomy reduced the midline shift in all patients with monolateral diffuse brain edema and contusions having a median value of 7 mm; in the remaining, it ameliorated the basal cisterns effacement. At a mean follow-up of 14 months, 6 (12.5%) patients died, 7 (15%) were discharged home with a GOS of 5, 18 (40%) showed a favorable outcome after rehabilitation with a GOS of 4 and 5, 6 (12.5%) had a severe disability (GOS 3), 9 (20%) were in a vegetative state (GOS 2), and 2 were lost to follow-up. The younger age, earlier surgery, and higher preoperative Glasgow Coma Scale score were related to better outcome (P < .001, P < .05, and P < .034, respectively).

Conclusion: Our results seem to support the idea that decompressive craniectomy coupled with neurointensive care may be an effective way to reduce intractable raised intracranial pressure, and probably to improve patients outcome. However, it should be obvious that our results and those available in the literature can not be considered conclusive.