Relationship between PaO(2)/FiO(2) and delirium in intensive care: A cross-sectional study

BACKGROUND: To investigate the relationship between partial pressure of oxygen (PaO(2))/fraction of inspired oxygen (FiO(2)) and the probability of delirium in intensive care units (ICUs). METHODS: The investigation was a cross-sectional study that involved the collection of data from patients admitted to the Xiang Ya Hospital Cardiothoracic Surgical Care Unit and Comprehensive Intensive Care Unit from 01 September 2016 to 10 December 2016. Delirium was diagnosed using the simplified version of the Chinese Confusion Assessment Method (CAM) for the ICU. Demographic and medical data were obtained within 24 h of each patient admitted in the ICU. The PaO(2)/FiO(2) of each patient was recorded 24 h after admission in the ICU. The patients were divided into three groups according to PaO(2)/FiO(2) data : normal (PaO(2)/FiO(2) ≥300 mmHg), slightly low (200 ≥PaO(2)/FiO(2) <300 mmHg), and severely low (PaO(2)/FiO(2) <200 mmHg). Baseline characteristics were compared in the three groups. Results of the unadjusted model, minimally adjusted model, and fully adjusted model are presented. RESULTS: A total of 403 participants were included in the study, of which 184 (45.7%) developed delirium. Age (P <0.001), Sequential Organ Failure Assessment (SOFA) score (P <0.001), Acute Physiology and Chronic Health Evaluation (APACHE) II score (P <0.001), mechanical ventilation time (P <0.001), history of hypertension (P=0.040), heart disease (P=0.040), sedation (P=0.001), and PaO(2)/FiO(2) (P=0.006) were significantly associated with delirium in univariate analysis. Multivariate regression analysis models were used to further analyze the associations between PaO(2)/FiO(2) and delirium. In the crude model, for 1 standard deviation (SD) increase in PaO(2)/FiO(2), the odds ratio (OR) of delirium was 0.8 (95% confidence interval [CI]: 0.6–0.9), but there was no significant correlation in the fully adjusted model. There was a non-linear relationship between the PaO(2)/FiO(2) and delirium in a generalized additive model. A two-piecewise linear regression model was used to calculate a PaO(2)/FiO(2) threshold of 243 mmHg. On the left side of the threshold, the OR was 0.9 and the 95% CI was 0.9–1.0 (P=0.013) when PaO(2)/FiO(2) increased by 1 SD. CONCLUSIONS: PaO(2)/FiO(2) was negatively associated with delirium when PaO(2)/FiO(2) was below the identified threshold. As a readily available laboratory indicator, PaO(2)/FiO(2) has potential value in the clinical evaluation of risk of delirium in ICU patients.