Relationship between arterial oxygen tension and mortality of patients in intensive care unit on mechanical ventilation support

Abstract

BACKGROUND: Although there are studies demonstrating hyperoxia may be an independent risk factor for increased mortality and morbidity, this issue remains unclear. Our research then aimed to examine the relationship between arterial oxygen tension, arterial carbon dioxide tension, and in-hospital mortality of critically ill patients in intensive care unit (ICU). METHODS: After obtaining ethics committee approval, we analyzed a retrospective data of patients over the age of 18 who survived at least 24 hours in the ICU on mechanical ventilatory support between year 2008 and 2012. The demographic properties, mechanical ventilation, and blood gas parameters were studied. We defined hyperoxia group as PaO2 value of ≥120 mmHg and normoxemia group as PaO2 of 60-120 mmHg. Patients with PaCO2 value <30 mmHg were determined to have hypocapnia, those with 30-50 mmHg normocapnia, and those with >50 mmHg hypercapnia. RESULTS: Between 2008 and 2012, a total of 7689 patients were admitted to the ICU. Of 450 patients meeting the inclusion criteria of the study, 263 (58.4%) were male and 187 (41.6%) were female. Normoxia was observed in 232 (51.5%) patients and hyperoxia in 218 (48.5%) patients. The mean PaO2 was 16.2 kPa (121.50 mmHg), and FiO2 was 60%. 254 (56%) of the patients had died during the five-year period. There was no statistically significant difference in mortality between PaO2 levels and PCO2 levels (p>0.05). According to the classification of PaO2 and FiO2, there was no statistically significant difference in mortality (p>0.05) among patients. In addition, no statistically significant difference was found between the survival rates according to PCO2 classification (p=0.602, p>0.05). CONCLUSION: There was no significant association between mortality and oxygen and carbon dioxide of patients in ICU on mechanical ventilatory support.