Objective: Traumatic brain injury (TBI) can result in cerebral oedema and vascular changes resulting in an increase in intracranial pressure (ICP), which can lead to further secondary damage. Decompressive craniectomy (DC) is a surgical option in the management of ICP. We aimed to investigate outcomes of DC after TBI.
Design: We performed a retrospective audit of 57 adult patients (aged > 15 years) who underwent DC after TBI, at the Royal Melbourne Hospital from 1 January 2005 to 30 June 2011. Our functional outcome measure was the Extended Glasgow Outcome Scale (GOSE).
Results: Patients had a median age of 30 years (range, 17- 73 years). The hospital mortality rate was 47% (27 patients). A higher postoperative median ICP was the most significant predictor of hospital mortality (OR, 1.1; 95% CI, 1-1.3). There was a mean decrease of 7.7mmHg in ICP between the mean preoperative and postoperative ICP values (95% CI, - 10.5 to - 5.0mmHg). There was a mean decrease of 3.5mmHg in the mean cerebral perfusion pressure (CPP) from preoperative to postoperative CPP values (95% CI, - 6.2 to - 0.8mmHg). At the 6-month follow-up, a poor outcome (GOSE score, 1-4) was seen in 39 patients (68%), while a good outcome (GOSE score, 5- 8) was noted in 15 patients (26%). A high APACHE II score on admission was the most significant predictor of a worse GOSE score at 6 months (OR, 1.3; 95% CI, 1.1-1.5). Analysis of the APACHE II and IMPACT scores as models for predicting mortality at 6 months showed an area under the curve (AUC) of 0.792 and 0.805, respectively, and for predicting poor outcome at 6 months, showed an AUC of 0.862 and 0.883, respectively.
Conclusion: DC decreased ICP postoperatively. The IMPACT and APACHE II scores are good models for prediction of death and poor outcome at 6 months.