Predictive Value in Receiving Early Unilateral Decompressive Craniectomy or Craniotomy in Traumatic Brain Injury Patients

Introduction: The outcomes in traumatic brain injury (TBI) patients who underwent unilateral decompressive craniectomy (DC) or craniotomy were evaluated.Method: A retrospective cohort with standard neurosurgical approach over a 4.5 year period was performed (46 craniectomy and 36 craniotomy). Several variables were recorded and characteristic of patients receiving craniectomy or craniotomy was compared. Using multivariate logistic regression and prognostic scoring models, Receiver-Operating Characteristic curve (ROC) was drawn to predict Glasgow Outcome Scale (GOS) 6 months after TBI.Results: Our data showed craniectomy group had a lower pre-operative Glasgow Coma Scale (GCS) (p = 0.012), unresponsive pupil reaction (p = 0.013), lower intracranial cerebral pressure (p = 0.001), cerebral perfusion pressure (p = 0.011) and cerebral circulatory-pressure index (p ＜ 0.001), but with higher post-operative cerebral infarction (p = 0.037) than craniotomy group. The favorable outcome in craniectomy group was 32.6%. Age (p = 0.06), pre-operative pupil reaction (p = 0.03), and pre-operative GCS (p = 0.04) were three independent predictors to poor outcome and ROC curve demonstrated its sensitivity 90.3% and specificity 88.2%.Conclusion: Although a small population in single medical center was enrolled, it evidenced the pre-operative neurological status was prognostic dependent. How to use aforementioned predictors before decompressive craniectomy or craniotomy to get a better outcome is still a challenge in daily practice.