Although critically ill patients with bloodstream infections (BSIs) who present with hypothermia are at the highest risk for death, it is not known how rewarming rates may influence the outcomes. The objective of this study was to identify the occurrence and determinants of hypothermia among patients admitted to intensive care units (ICUs) with BSI and assess how the rate of temperature correction may influence 90-day all-cause case-fatality. A cohort of 3951 ICU admissions associated with BSI was assembled. The lowest temperature measured within the first 24 hours of admission was identified, and among those who were hypothermic (<36°C), the rewarming rate [(time difference between lowest and subsequent first temperature ≥36°C) divided by hypothermia severity (difference between lowest measured and 36°C)] was determined. Within the first 24 hours of admission to the ICU, 329 (8.4%) and 897 (22.7%) subjects had the lowest temperature measurements ranging <34.9°C and 35-35.9°C, respectively. Patients with lower temperatures were more likely to be admitted to tertiary care ICUs, have more comorbid illnesses, have greater severity of illness, and have a higher need for organ-supportive therapies. The 90-day all-cause case-fatality rate was 22.9% overall and was 45.3%, 24.8%, and 19.6% for those with the lowest 24 hours temperatures of <35°C, 35-35.9°C, and ≥36°C, respectively (p < 0.001). Among 1133 hypothermic patients with documented temperatures corrected to the normal range while admitted to the ICU, the median rate of temperature increase was 0.24 (interquartile range, 0.13-0.45)oC/hour. After controlling for the severity of illness and comorbidity, a faster rewarming rate was associated with significantly lower 90-day case-fatality. Hypothermia is a significant risk factor associated with death among critically ill patients with BSI that faster rates of rewarming may modify.
Read full abstract