The impact of PEEP on ICP was dependent on the difference between elevated CVP levels and baseline ICP levels. ICP would increase once elevated CVP through PEEP adjustment exceeds the baseline ICP.

all patients were exposed to incremental PEEP levels of 0, 5, 10, and 15cmH2O with 100% of FiO2. The measurements were done bedside on stabilized hemodynamics and intracranial pressure. The measurement was discontinued if the following situation presented and remedies were applied accordingly: (1) CPP < 60 mmHg (norepinephrine at 0.3~1.0μg/kg.min was used); (2) ICP > 25 mmHg (PEEP was restored to 0); (3) increase of pressure plateau > 35 cmH2O (tidal volume was decreased and PetCO2 was maintained at 30~35mmHg); (4) SpO2 < 90% (PEEP was restored to 0); and (5) suspicion of pneumothorax (PEEP was restored to 0 and chest radiography was performed). An equilibration period (at least 90 seconds) was entailed to ensure a normalized baseline PetCO2 through modulating tidal volume and respiratory rate.

ICP, CVP, Pj, and MAP were measured twice or more at each level of PEEP for consecutively five days after admission. CPP was calculated according to the following equation: CPP=MAP-ICP. The difference between baseline ICP and CVP was categorized into the following three groups according to the previous findings: Group I,IVPD ≤ 3mmHg, Group II, 3 < IVPD ≤ 6 mmHg, Group III, IVPD > 6 mmHg. Relationships between PEEP and ICP, CVP and MAP, CVP and Pj were analyzed in each group respectively.

Inclusion Criteria:

• All patients diagnosed with sTBI (GCS≤8) and applied with MV were initially included.

Exclusion Criteria:

• Brain death
• Younger than 18 or older than 80 years
• Pregnancy
• Hemodynamic instability：for example heart rate >120 bpm or CPP <60 mmHg
• Bulbous lung or pneumothorax
• Myocardial infarction
• Refusal of consent