Respective impact of lowering body temperature and heart rate on mortality in septic shock: mediation analysis of a randomized trial

Purpose We previously showed that external cooling decreases day 14 mortality in febrile septic shock. Because cooling may participate in heart rate control, we studied the respective impact of heart rate and temperature lowering on mortality. Methods Post hoc analysis of the Sepsiscool randomized controlled trial database (NCT00527007). Cooling was applied to maintain normothermia (36.5-37 A degrees C) during 48 h. We assessed the time spent below different thresholds of temperature and heart rate on day 14 mortality. The best threshold was selected by AUC-ROC and tested as a potential mediator of mortality reduction. Mediation analysis was adjusted for severity and treatments influencing temperature and heart rate evolution. Sensitivity analysis was done using only patients with appropriate antimicrobial therapy. Results A total of 197/200 patients with adequate heart rate and temperature monitoring were analyzed. The best threshold differentiating survivors and nonsurvivors was 38.4 A degrees C for temperature and 95 b/min for heart rate. During the 48 h of intervention, cooling significantly increased the time spent with a temperature below 38.4 A degrees C, p = 0.001, and with a heart rate below 95 b/min, p < 0.01. The longer was the time spent with a temperature below 38.4 A degrees C, the lower was the mortality [adjOR 0.17 (0.06-0.49), p = 0.001]. The time spent with a heart rate below 95 b/min was similar in survivors and nonsurvivors [adjOR 0.68 (0.27-1.72), p = 0.42]. Mediation analysis showed that the time spent with a temperature below 38.4 A degrees C was a significant mediator of mortality. Conclusion The time spent with a temperature below 38.4 A degrees C was independently associated with patient's outcome and explained 73 % of the effect of the randomization on the day 14 mortality. Heart rate lowering was not a mediator of mortality.