Effect of active temperature management on mortality in intensive care unit patients

Objective: To evaluate the effect of active temperature management on mortality, intensive care unit (ICU) and hospital length of stay, as well as the relative effi cacy of antipyretic medications and physical cooling devices for achieving reductions in temperature in critically ill adults. Design, setting and participants: Systematic review and meta-analysis of randomised controlled trials (RCTs) investigating treatments administered to febrile patients in order to reduce body temperature. Fifteen studies reporting results from 13 RCTs met our eligibility criteria. Interventions: Treatments administered to reduce body temperature were defi ned as physical cooling, nonsteroidal anti-infl ammatory drugs, paracetamol, or any combination of these. Main outcome measures: The primary outcome variable was all-cause mortality at the longest time point after randomisation. Secondary outcomes were ICU and hospital length of stay, and body temperature 12 hours after randomisation. Results: Active temperature control had no statistically signifi cant association with mortality (odds ratio, 1.01; 95% confi dence interval [CI], 0.81-1.28; P = 0.95, for fi xed effects). There was no statistically signifi cant association between active temperature management and ICU or hospital length of stay. Active temperature management was associated with a statistically signifi cant reduction in temperature. The fi xed effects estimate for the active minus control treatment for pharmaceutical management was −0.62°C (95% CI, −0.72°C to −0.51°C; P < 0.001) and for physical cooling was −1.59°C (95% CI, −1.82°C to −1.35°C; P < 0.001). Conclusions: Active temperature management neither increased nor decreased mortality risk in critically ill adults. When the therapeutic goal is to reduce body temperature, physical cooling approaches may be more effective than pharmacological measures in critically ill adults.