Predictors of external cooling failure after cardiac arrest

External cooling is largely employed to induce hypothermia in comatose survivors of cardiac arrest (CA), but can fail to reach the target temperature in a reasonable time. We aimed to assess the rate of failure of external cooling after CA and to determine failure predictors. The study was a retrospective review of a prospectively acquired database in the setting of a 24-bed ICU in a university hospital. All consecutive patients admitted for CA from May 2002 to April 2010 and treated by external cooling were considered. Patients who were already hypothermic on admission, patients dying within 24 h, patients cooled by an internal technique and patients in whom hypothermia had not been attempted were not studied. External cooling failure was defined as the inability to reach a temperature below 34 A degrees C during the first 12 h after CA onset. Among 1,036 patients admitted to the ICU, 594 were included in the analysis and in 191 (32 %) the target temperature could not be achieved within the 12 h following CA. Independent risk factors for external cooling failure were an early coronary angiography intervention (OR 3.75, p < 0.001), a high body weight (OR 1.02 per kilogram, p = 0.007), a high temperature on ICU admission (OR 1.47 per degree, p = 0.001) and a long delay between collapse and the start of cooling (OR 1.15, p = 0.05). Conversely, early haemodialysis (OR 0.27, p < 0.001) and male gender (OR 0.47, p = 0.02) were significantly associated with cooling success. External cooling failure occurred in nearly one-third of patients with CA and was associated with easily identified risk factors. This emphasizes the interest in early cooling and alternative techniques in these patients.