Risk Factors Predicting Posttraumatic Hydrocephalus After Decompressive Craniectomy in Traumatic Brain Injury

To identify risk factors for predicting posttraumatic hydrocephalus (PTH) development after traumatic brain injury in patients who underwent decompressive craniectomy (DC).

This retrospective study included 121 patients who underwent DC performed by 6 different neurosurgeons after traumatic brain injury between January 2013 and December 2016 at Yijishan Hospital and were still alive at 6-month follow-up. Patients were divided into PTH group and non-PTH group. Logistic regression analysis was used to identify PTH potential risk factors based on results obtained from univariate analysis. Power of the regression model to discriminate PTH from non-PTH was evaluated using receiver operating characteristic curve.

With Glasgow Coma Scale (GCS) score cutoff value of 6, GCS scores <6 on admission, craniectomy site, and intraventricular hemorrhage (IVH) were significant predictors for development of PTH after DC. Receiver operating characteristic curve indicated that a final predictive model composed of these 3 factors (area under the curve [AUC] = 0.866, sensitivity = 0.78, and specificity = 0.83) was significantly better than each single model (AUC = 0.750 for GCS scores on admission, AUC = 0.650 for craniectomy site, AUC = 0.572 for IVH). A significantly positive association was found between patients' Glasgow Outcome Scale Extended scores and GCS scores on admission, whereas a significantly negative association was found between IVH, craniectomy site, and Glasgow Outcome Scale Extended.

GCS scores <6 on admission, presence of IVH on first head computed tomography scan, and need for bilateral DC might be used to predict whether patients with traumatic brain injury after DC will develop PTH. The reliability of this specific combination might be useful for clinicians to make a correct prediction.