Abstract

INTRODUCTION: Admission hypothermia is associated with increased mortality in patients hospitalized with sepsis, trauma or stroke. We hypothesized that hypothermia on admission in patients with acute systolic heart failure (HF) would be associated with worse survival. METHODS: Using ESCAPE trial data, we compared patients hospitalized with acute systolic HF with admission temperature ≤35.5°C or >35.5°C. The study endpoints were 30-day and 6-month all-cause mortality. RESULTS: Among the 398 patients (mean age 56years, 74% men) who had body temperature recorded on hospital admission, 23 had a temperature ≤35.5°C (hypothermia) and 375 patients had a temperature >35.5°C. By univariate comparison, hypothermic patients had a higher blood urea nitrogen (BUN) level (46.6versus 34.6mg/dL, P=0.04) and creatinine level (1.85 versus 1.49mg/dL, P=0.035) and higher frequency of malignancy (26.1% versus 6.1%, P=0.001). There was a significant correlation between admission temperature and body mass index (r=0.201, P<0.0001). Patients with temperature≤35.5°C had a higher 30-day (17.4% versus 4.3%, univariate OR 4.724, 95% CI 1.439-15.509, P=0.01) and 6-month (43.5% versus 18.1%, univariate OR 3.473, 95% CI 1.462-8.25, P=0.005) all-cause mortality. Cox proportional hazard analysis revealed that admission temperature ≤35.5°C is an independent predictor of mortality (hazard ratio 2.222, 95% CI 1.117-4.420, P=0.023). Kaplan-Meier analysis comparing survival in patients with admission temperature ≤35.5°C or >35.5°C showed a significant inter-group difference in survival (log-rank P value=0.001). CONCLUSION: Admission body temperature ≤35.5°C is an independent predictor of short and intermediate term mortality in patients hospitalized with acute systolic HF.