Objective: To determine the incidence of body temperature (BT) alterations in critically ill patients, and their relationship with infection and outcome.
Design: Prospective, observational study. SETTING. Thirty-one bed, medico-surgical department of intensive care.
Patients: Adult patients admitted consecutively to the ICU for at least 24 h, during 6 summer months.
Interventions: None.
Results: Fever (BT > or =38.3 degrees C) occurred in 139 (28.2%) patients and hypothermia (BT< or =36 degrees C) in 45 (9.1%) patients, at some time during the ICU stay. Fever was present in 52 of 100 (52.0%) infected patients without septic shock, and in 24 of 38 (63.2%) patients with septic shock. Hypothermia occurred in 5 of 100 (5.0%) infected patients without septic shock and in 5 of 38 (13.1%) patients with septic shock. Patients with hypothermia and fever had higher Sequential Organ Failure Assessment (SOFA) scores on admission (6.3+/-3.7 and 6.4+/-3.3 vs 4.6+/-3.2; p<0.01), maximum SOFA scores during ICU stay (7.6+/-5.2 and 8.2+/-4.7 vs 5.4+/-3.8; p<0.01) and mortality rates (33.3 and 35.3% vs 10.3%; p<0.01). The length of stay (LOS) was longer in febrile patients than in hypothermic and normothermic (36 degrees C<BT<38.3 degrees C) patients [median 6 (1-57) vs 5 (2-28) and 3 (1-33) days, p=0.02 and p=0.01, respectively). Among the septic patients hypothermic patients were older than febrile patients (69+/-9 vs 54+/-7 years, p=0.01). Patients with septic shock had a higher mortality if they were hypothermic than if they were febrile (80 vs 50%, p<0.01).
Conclusions: Both hypothermia and fever are associated with increased morbidity and mortality rates. Patients with hypothermia have a worse prognosis than those with fever.