OBJECTIVE: To identify peripartum risk factors that are predictive of positive blood cultures in patients with postcesarean endometritis. STUDY DESIGN: A retrospective review of 179 patients diagnosed with postcesarean endometritis was conducted. Patients with positive and negative blood cultures obtained at the time of diagnosis were compared. Patient's charts were reviewed for intrapartum, intraoperative and postpartum factors. Chi-square and nonpaired Student's tests were used when appropriate, with P < .05 considered significant. RESULTS: During this period, 179 (20%) postcesarean patients developed endometritis. One hundred sixty-eight (94%) of those patients had blood cultures. Eleven (6.5%) were positive; however, one of these grew a skin contaminant and was disregarded. When patients with positive blood cultures were compared to those with negative blood cultures, length of labor, number of vaginal examinations, postoperative day when the diagnosis was established, estimated blood loss at the time of cesarean delivery, presence of intrapartum chorioamnionitis, number of hours of ruptured membranes, white blood cell count at the time of diagnosis, use of prophylactic antibiotics, development of wound infection or other infectious etiologies were not shown to be predictive. There were no positive blood cultures among patients with a temperature < 38.5 degrees C. At a temperature < 38.8 degrees C, 1/126 (0.79%) had a positive blood culture. At a temperature > or = 38.8 degrees C, 9/42 (21.4%) had a positive blood culture (P < .001). Approximately $5,890 was spent on obtaining positive blood cultures in patients with temperatures < 38.8 degrees C. In contrast, $218 was spent per positive blood culture obtained from patients with a temperature > or = 38.8 degrees C. CONCLUSION: The traditional practice of obtaining blood cultures at a temperature > or = 38.0 degrees C is not justified but elevating the threshold to 38.8 degrees C is equally effective and less costly.
