Matt Harmon

30 Chapter two Abstract Introduction Hypothermia is associated with high mortality in sepsis, but it is now recognized that this association may simply reflect its higher prevalence in sicker patients. Furthermore, there is evidence to suggest that hypothermia may not represent a dysfunction in sepsis. In this study, we conducted a survey to assess how this scientific evidence relates to the perceptions of health care professionals regarding septic hypothermia, and how such perceptions drive clinical conduct concerning the use of active rewarming in this population. Methods A survey with questions on opinions and management of spontaneous hypothermia in sepsis was developed and posted online at the European Society of Intensive Care Medicine (ESICM) website from March 24th, 2017 to the June 26th, 2017 and distributed by electronic email. Respondents were asked to fill in the survey from the perspective of their usual or average practice in their intensive care unit. Results In total, there were 440 survey respondents. Respondents were predominantly from Europe (66%) The majority of respondents were intensivists (78%) and worked in an academic hospital (66%). One percent of respondents were nurses. Most respondents (96%) reported that there was no protocol for the management of hypothermic sepsis. Of the respondents, 62% actively rewarmed patients with hypothermic sepsis. Hypothermia was defined as a temperature below 36°C (44%) and below 35°C (15%). Rewarming practices showed large variation in terms of the temperature, at which respondents initiate rewarming as well as the target temperature to which patients are rewarmed. The most predominant first-line rewarming method was forced-warm air followed by warm IV fluids. Rewarming decisions were mostly physician driven (58%). Most respondents thought rewarming was beneficial (43%), a small proportion thought rewarming to be harmful (9%). Conclusions In conclusion, policies, procedures, and beliefs about spontaneous hypothermia and active rewarming in patients with sepsis are variable. This must be taken into consideration in designing future trials. We propose a working group to define hypothermic sepsis to improve comparability of research.