This study aimed to address the risk factors of second decompressive craniectomy (DC) in patients with traumatic brain injury (TBI) who initially underwent mass lesion evacuation, but no primary DC. Patients were enrolled if they had had a hospital visit to Xiangya Hospital, Central South University with acute closed TBI from January 1, 2017 to December 31, 2019 and had undergone craniotomic mass lesion evacuation. Sociodemographic information, computed tomography (CT) information, clinical profiles, and surgical information were obtained from an electronic database. Twenty-four patients who had undergone a second decompressive craniectomy (SDC) and 39 patients who had not (NSO) were included in the analysis. The prevailing lesions differed between the groups (p = 0.010). The SDC group had more compressed/obliterated basal cisterns than the NSO group (p = 0.028). After closure of the dura, the SDC group also had higher intracranial pressure (ICP) than the NSO group (10.9 mm Hg vs. 6.5 mm Hg, p = 0.005). Binary logistical regression indicated that ICP after dura closure was an independent predictor of second DC (odds ratio [OR] = 1.317, p = 0.011). A model using ICP after dura closure alone had an area under the curve value of 0.757 in its receiver operating characteristic curve. An ICP >10.5 mm Hg after closure of dura for the prediction of a second DC had a sensitivity of 56.3% and a specificity of 92.6%.
Keywords: decompressive craniectomy; intracranial pressure monitor; predictive model; traumatic brain injury.