ABSTRACT
PURPOSE: This study was conducted to derive and validate a clinical prediction rule for blood cultures obtained in the emergency department. METHODS: We conducted a retrospective chart review of patients with a fever who underwent blood cultures from August 2002 through July 2003. The student t-test, coefficient of correlation, ROC (receiver operatics characteristic) curve, and Chi-square statistics were used for the analyses. RESULTS: Included in the study were 108 patients, of whom 47 were men and 61 were women. The mean age was 45.30+/-26.21 years old. The most common diagnoses were pneumonia and acute gastroenteritis. The usefulness of blood cultures in the emergency department was statistically significant in patients with fever or with higher levels of band form. CONCLUSION: We determined that blood cultures are uncommonly useful for febrile patients and that a need rule is needed for the use of blood cutures in the emergency department