Effect of hypothermia and hyperthermia on all-cause in-hospital mortality in emergencies: a comprehensive nationwide analysis from the Republic of Korea

Hypothermia has been shown to be a predictor of poor outcomes in various settings, but the association between elevated body temperature (BT) and patients' outcomes remains unclear. This study aimed to investigate the relationship between body temperature and mortality in general emergency departments. A nationwide cohort study was conducted to evaluate the effects of BT on all-cause in-hospital mortality in patients in the emergency departments of the Republic of Korea. Data from the National Emergency Department Information System, which stores regional and local emergency medical center data, were retrieved from 01 January 2014, to 31 December 2016. The patients were classified into a disease group (infection and cerebrovascular accident (CVA)) and an injury group (traumatic brain injury and non-traumatic brain injury), and the association between their mortality and body temperature were evaluated. The MantelHaenszel test was used to identify patterns in Odd-Ratio (OR). In all, 52.73% (837,506) of the study were male and the median age of the entire cohort was 59 (interquartile range, 44-73) years. In the Mantel-Haenszel test, adjusted ORs were negatively correlated with mortality in the disease group (chi(2) = 1087.28; p< 0.001,chi(2) = 1886.27; p< 0.001, <36.6 degrees C and >37.0 degrees C respectively). In the injured group, a negative correlation below the reference range (chi(2) = 447.21; p < 0.001) and a tendency for a positive correlation above the reference range (chi(2) = 5.62; p = 0.02) were detected. Among the disease group, BT was negatively correlated with in-hospital mortality (chi(2) = 493.90; p < 0.001, chi(2) = 1741.2; p < 0.001, <36.6 degrees C and >37.0 degrees C, respectively) in patients with infection, and negatively correlated in the lower BT range (chi(2) = 497.67; p < 0.001) but was not significant in the higher BT range (chi(2) = 5.97; p = 0.01) in patients with CVA. Lower BTs were associated with higher in-hospital mortality in patients from the disease or injury group in general emergency departments. Higher BTs were associated with lower inhospital mortality in the disease group, especially in those with an infection, but not in patients with CVA or injury.