Objectives: To test the hypothesis that hyperoxia was associated with higher in-hospital case-fatality in ventilated patients with ICH admitted to the Intensive Care Unit (ICU).
Methods: Admissions of ventilated ICH patients within 24 hours of admission to the ICU at 77 United States hospitals between 2003-2008. Patients were divided into three exposure groups: hyperoxia (PaO2≥300 mmHg), hypoxia (PaO2<60 mmHg or PaO2/FiO2 ratio ≤300), and normoxia, not defined as hyperoxia or hypoxia. The primary outcome was in-hospital case-fatality.
Results: 1,388 ventilated ICH patients. Mean age 63 years (SD 15), 47% (653/1388) female, and median GCS 5 (IQR 3-8). The overall case-fatality was 59% (812/1388). Upon admission to the ICU, 36% (492/1388) were normoxic, 47% (641/1388) were hypoxic, and 17% (238/1388) were hyperoxic on ABGt1. ABGt2 was accomplished in 780 patients, of whom 46% (352/780) were normoxic, 45% (352/780) were hypoxic, and 9% (67/780) were hyperoxic. Of the initially admitted hyperoxic patients, 15% (21/138) remained hyperoxic and had a case-fatality of 82% (18/21) as compared to 49% (67/138) who became normoxic and had a case-fatality of 46% (32/67) (crude OR 6.6, 95%CI:1.8-25, χ2=9.4, p=0.002). In a multivariable analysis controlling for other predictors of poor outcome and hospital specific characteristics, and a propensity-score, failure to correct hyperoxia was associated with higher case-fatality (adjusted OR 2.5, 95%CI:1.1-6.1, p=0.04).
Conclusion: In ventilated ICH patients, failure to normalize hyperoxia was associated with higher case-fatality. These data underscore the need for studies of controlled re-oxygenation in ventilated ICH patients.
Hyperoxia is associated with higher case-fatality in ventilated patients with intra-cerebral hemorrhage