Hypothermia association with all-cause mortality in critically ill patients with sepsis based on the MIMIC-IV database.

In sepsis, temperature fluctuations are typical. The objective of this study was to determine whether hypothermia in sepsis patients was associated with an increased risk of death. To identify septic patients with and without hypothermia, the Medical Information Mart for Intensive Care IV (MIMIC-IV) database was queried. The primary outcome was all-cause mortality. Multivariate Cox regression was utilized to determine the association between hypothermia and all-cause mortality, and propensity score matching (PSM) and an inverse probability of treatment weighting (IPTW) were used to validate our findings. Our analysis comprised 19,636 septic patients, of whom 937 (4.8%) developed hypothermia within 24 h of intensive care unit (ICU) admission. The hypothermia group had a higher 180-day and 360-day mortality than the no hypothermia group (P < 0.001). Hypothermia was associated with higher 180-day and 360-day mortality in the matched cohort and in the weighted cohort. In conclusion, hypothermia was associated with increased in-hospital and all-cause mortality in septic patients, emphasizing the significance of temperature protection in the ICU.