Sepsis in the absence of fever: Determining the criteria for and feasibility of future therapeutic temperature management trials

Abstract

The purpose of this study is to examine the occurrence, characteristics, and outcomes of intensive care unit (ICU) patients with sepsis and the absence of fever. Multicentre, retrospective cohort study. Twelve ICUs in Queensland, Australia. Adults (≥18 years) admitted to the ICU with sepsis between 1 January 2015 and 31 December 2021 were eligible for inclusion. Patients admitted with seizures, traumatic brain injury, postcardiac arrest, end-of-life care, and elective surgery were excluded, as were readmissions. The primary outcome was fever deficit (defined as degree-hours under 38.3°C) during the first 72 h of ICU admission, and all-cause 30-day mortality was the key secondary outcome. Of 89,117 admissions, 15,612 were included. Admission temperatures were ≥38.3°C in 1026 (6.6%), 37.5-38.2°C in 2096 (13.4%), 36-37.4°C in 9216 (59.0%), and <36°C in 3274 (21.0%). Temperatures changed rapidly over the first 12 h and, by 24 h, approached reasonably stable levels. For the admission temperature groups of ≥38.3°C, 37.5-38.2°C, 36-37.4°C, and <36°C, fever deficits were a median of 47 (interquartile range (IQR), 24 to 72), 53 (IQR, 29 to 83), 69 (IQR, 40 to 100), and 85 (IQR, 52 to 123) degree-hours, respectively, and 147 (14%), 248 (12%), 1,104 (12%), and 549 (17%) died by day 30. After controlling for confounders, a high fever deficit, defined as a fever deficit above the median, during the first 24 h of ICU admission, was not associated with all-cause 30-day mortality (OR 1.02, 95% CI, 0.93-1.13; p = 0.7). Fever deficits were large, particularly when the initial body temperature was not febrile. Only 1 in 15 ICU patients with sepsis had an initial body temperature ≥38.3°C. Approximately 2000 adults a year with sepsis and an initial body temperature <37.5°C would potentially be eligible for a trial of therapeutic hyperthermia in our 12 ICUs.