Temperature trajectories, active rewarming, and mortality in hypothermic sepsis: a multicenter retrospective cohort study

Hypothermia occurs in 9–35% of septic patients and is associated with increased mortality. Whether active rewarming improves outcomes remains unresolved, and the prognostic value of spontaneous rewarming has not been distinguished from externally induced warming. This study explored the effect of active rewarming and the prognostic significance of temperature trajectories in hypothermic sepsis. This multicenter retrospective cohort study included 436 septic patients with admission core temperature < 36 °C from six ICUs in China. Active rewarming was implemented at clinician discretion. The primary outcome was 28‑day mortality. Group‑based trajectory modeling was performed to identify distinct temperature patterns. Subgroup and sensitivity analyses were conducted. Active rewarming was not associated with lower 28‑day mortality (45.0% vs. 48.0%, P = 0.529) or 90‑day mortality, nor was it associated with a lower incidence of ARDS or AKI. Four temperature trajectory phenotypes were identified. Faster temperature restoration, whether occurring spontaneously or facilitated by active warming, tended to be associated with better survival, while persistent hypothermia and incomplete rewarming were linked to elevated 28-day mortality risk. In this observational cohort, active rewarming was not associated with improved overall outcomes in hypothermic septic patients. Temperature recovery trajectories appear to be more strongly linked to prognosis than receipt of the intervention itself, and faster temperature recovery tends to be associated with better survival. Prospective studies with standardized rewarming protocols are needed to further refine temperature management strategies for this patient population. Not applicable.