Prognosis of accidental hypothermia in the urban setting

Our goal was to study the outcome and factors affecting outcome in patients with accidental hyothermia in the urban setting. A retrospective cohort st udy was done on patients admitted to the emergency department with accident al hypothermia (core temperature less than 35 degreesC) from 1991 to 1998. Patients received general intensive care (fluid replacement, mechanical ven tilation, cardiopulmonary resuscitation) and were rewarmed actively by intr avenous application of warmed fluids, warmed air, pleural lavage, or cardio pulmonary bypass as indicated. Rewarming was successful in 76 of 80 patient s (95%) and hospital mortality was 34%. Survivors differed significantly fr om nonsurvivors with respect to demographic data (indoor finding, homelessn ess, alcohol abuse, chronic psychiatric disorder, presumed acute alcohol or drug intoxication), findings on admission (severity of hypothermia, systol ic blood pressure, heart rate, multiple organ failure (MOF) score, hypother mia outcome score (HOS), BUN, creatinine, CPK, bilirubin, AST, platelet cou nt), but not with respect to therapeutic modalities (mechanical ventilation , volume replacement, pleural lavage, mechanical ventilation, vasopressors, cardiopulmonary bypass). Rewarming time in nonsurvivors, however, was sign ificantly longer. In logistic regression analysis only indoor occurrence, B UN, and platelet count proved to be independent predictors of in-hospital m ortality. Survival of patients found outdoors was significantly higher than that of patients found indoors (p < 0.0001). Our study demonstrates that a lthough rewarming and resuscitation are highly effective, accidental urban hypothermia is a condition with a significant in-hospital mortality. There were only a few independent indicators of unfavorable outcome, of which ind oor occurrence was by far the most important. Therefore indoor occurrence o f hypothermia has to be acknowledged as a risk factor of poor outcome in ac cidental urban hypothermia.