Objective:To observe the effect of early goal directed sedation (EGDS) on cerebral oxygen metabolism in patients with acute brain injury.Methods:A prospective cohort study was conducted. A total of 108 patients with acute brain injury admitted to the intensive care unit (ICU) of the Third Medical Center of the PLA General Hospital from January 2015 to December 2019 were enrolled. According to the patient's condition, dexmedetomidine contraindication and tolerance, and combined with the wishes of patients' families, they were divided into EGDS group and on-demand sedation group. Routine treatments such as surgery, mechanical ventilation, dehydration and reduction of intracranial pressure with mannitol, hemostasis or antiplatelets therapy were given according to the patient's condition. All patients were continuously given sufentanil by intravenous infusion for analgesia. Patients in the EGDS group were sedated by continuously intravenous infusion of dexmedetomidine (0.2-0.7 μg·kg -1·min -1) for 72 consecutive hours. Patients in the on-demand sedation group received intravenous bolus of propofol (0.5-1.0 mg/kg) when treatments were interfered due to agitation. Hemodynamic indexes [heart rate (HR), mean arterial pressure (MAP), cerebral perfusion pressure (CPP), intracranial pressure (ICP)], sedation indexes [bispectral index (BIS)], severity indexes [acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) score, Glasgow coma score (GCS)] and cerebral oxygen metabolism indexes [jugular venous blood lactate (Lac), jugular venous oxygen saturation (SjvO 2), cerebral arterial oxygen content (CaO 2), cerebral extraction rate of oxygen (CERO 2), cerebral arteriovenous blood oxygen content difference (a-vDO 2)] were compared between the two groups before sedation and at 24, 48 and 72 hours of sedation. Results:① Among the 108 patients, 3 patients with cerebral hemorrhage received secondary surgery or had worsening of cerebral hernia were excluded. 105 patients were enrolled in the study, including 54 patients in the EGDS group and 51 patients in the on-demand sedation group. There were no statistically significant differences in gender, age, type of craniocerebral injury, GCS score, proportion of mechanical ventilation and operation ratio between the two groups. ② Compared with before sedation, Lac, CERO 2 and a-vDO 2 of both groups gradually reduced over time of sedation while SjvO 2 and CaO 2 were gradually higher. Those changes were more quickly in the EGDS group, Lac, SjO 2, CERO 2 and a-vDO 2 significantly improved at 24 hours of sedation compared with those before sedation. Above indexes at 72 hours of sedation in the EGDS group were obviously better than those in the on-demand sedation group [Lac (mmol/L): 1.81±0.31 vs. 2.19±0.12, SjvO 2: 0.714±0.125 vs. 0.683±0.132, CaO 2 (mL/L): 201.21±15.25 vs. 179.65±14.07, CERO 2: (27.87±3.66)% vs. (33.00±2.58)%, a-vDO 2 (mL/L): 44.32±5.68 vs. 48.57±8.22, all P < 0.05]. ③ Compared with before sedation, HR, MAP and ICP decreased in the two groups over time while CPP, BIS and GCS score showed increasing trend, especially more quickly in the EGDS group, HR at 24 hours of sedation, MAP, CPP, BIS and GCS score at 48 hours significantly improved as compared with those before sedation. Hemodynamics and sedation related parameters and GCS score at 72 hours of sedation in the EGDS group were significantly better than those in the on-demand sedation group [HR (bpm): 70.69±7.80 vs. 79.85±9.77, MAP (mmHg, 1 mmHg = 0.133 kPa): 84.23±8.76 vs. 89.97±9.48, ICP (mmHg): 14.23±8.76 vs. 15.97±9.48, BIS: 60.56±24.58 vs. 56.86±33.44, GCS score: 8.06±3.63 vs. 7.86±2.98, all P < 0.05]. The APACHE Ⅱ scores were significantly reduced at 72 hours of sedation in both groups as compared with those before sedation, while there was no statistical difference between the two groups. Conclusion:Compared with the on-demand sedation, EGDS could reduce cerebral oxygen metabolism, improve the coma degree, and reduce the severity of the disease in patients with acute brain injury.