Septic pelvic thrombophlebitis or refractory postpartum fever of undetermined etiology.

The objective of this study was to review and characterize the presentation, diagnostic dilemmas, management, and prognosis of postpartum septic pelvic thrombophlebitis. Medical records of postpartum women with the diagnosis of septic pelvic thrombophlebitis were reviewed for the 8-year period 1986-1994. Cases of documented ovarian vein thrombosis or those with other pelvic pathology on imaging study were excluded. Thirty-one women, four following vaginal delivery and 27 following cesarean delivery, with a final diagnosis of septic pelvic thrombophlebitis were identified. All patients demonstrated refractory febrile morbidity (mean 5.5 +/- 1.9 days prior to instituting heparin therapy) despite multiagent antimicrobial therapy with ampicillin, gentamicin, and clindamycin. Imaging studies (CT and/or ultrasound) were performed in 20 women and revealed no pelvic pathology. The patients required an average of 4.7 +/- 2.1 days (median 5, range 1-9 days) of heparin therapy before defervescence. Heparin levels were therapeutic at a mean of less than 24 h (range 6-24 h). The average dose of heparin required was 16.0 +/- 3.0 U/kg/h. Nine women had 13 subsequent pregnancies without recurrent thromboembolic complications. Currently available imaging studies cannot diagnose the entity we now define as septic pelvic thrombophlebitis (once cases of ovarian vein thrombosis are excluded). Our findings do not support the time-honored rule that septic pelvic thrombophlebitis responds within 24-48 h to therapeutic anticoagulation with heparin. Therefore, criteria other than imaging studies or immediate defervescence following heparin therapy are necessary for diagnosis of septic pelvic thrombophlebitis. A more appropriate terminology for septic pelvic thrombophlebitis should be refractory postpartum fever of undetermined etiology.