Purpose Sepsis is a heterogeneous clinical syndrome, and identification of its subphenotypes is essential. We aimed to investigate the correlation between the different trajectories of temperature and outcomes in septic patients. Patients and methods This was a retrospective cohort study that included sepsis patients admitted to the intensive care unit (ICU) of a tertiary care hospital. Patients’ temperature data were collected during the first 48 h after ICU admission, and patients’ temperature trajectories were identified using growth‐based trajectory modeling (GBTM). A multivariable logistic regression model was performed to assess the independent association of clinical outcomes. External validation was conducted using the MIMIC‐IV database. Results A total of 312 patients with sepsis were included. Four temperature trajectories were identified: hypothermia group ( n = 33), fever and rapid temperature drop group ( n = 45), persistently elevated temperature group ( n = 63), and normothermia group ( n = 171). Hypothermia patients exhibited the highest mortality and incidence of hospital‐acquired bloodstream infections (HA‐BSIs). Multivariable logistic regression showed that compared with the normothermia group, patients in the hypothermia group (OR 2.73; 95% CI 1.10–6.73; p = 0.030) and the persistently elevated temperature group (OR 3.40; 95% CI 1.74–6.67; p < 0.001) had significantly higher in‐hospital mortality. Similar subphenotypes were observed in the MIMIC‐IV cohort. Further analysis reached similar results in the MIMIC‐IV dataset. Conclusion This study used longitudinal body temperature data to identify the subphenotypes of sepsis, with significant variability in in‐hospital mortality and HA‐BSI. A better understanding of the temperature trajectory may be helpful in identifying deteriorating septic patients in advance.
Ding et al. (2026) studied this question.