OTA 2002 Posters

Deep Pelvic Infections and Osteomyelitis: Management and Outcomes

Bruce H. Ziran, MD; Ronald A. Hall, MD; Wade R. Smith, MD; N. Rao, MD; Daria L. Brooks, MD; University of Pittsburgh, Department of Orthopaedics, Pittsburgh, Pennsylvania, USA

Deep pelvic infections and osteomyelitis are serious and potentially life threatening conditions. Pelvic infections are often triaged for treatment to orthopaedic trauma surgeons because of their expertise with pelvic injuries. There are few reports on the management and outcomes of deep pelvic infections and osteomyelitis. We report our experience with the treatment of this serious condition.

Methods: We identified all patients who were treated for a deep pelvic infection over a 5-year period. Inclusion criteria were a documented deep pelvic abscess contiguous with any pelvic bone or osteomyelitis or both. Patients were stratified by whether there was pelvic osseous involvement (PO) or isolated deep pelvic infection (PI). Diagnosis, treatment, and outcomes were recorded, as were comorbidities, causative factors, and demographic data. Further stratification was based upon the Cierny-Mader host class; definitive treatment consisted of either wide debridement (host type A, B, some C) or suppression (host type C). Follow-up consisted of clinical, radiographic, and laboratory studies as well as CT and technetium/indium bone scans and MRI, when applicable. A good outcome was defined as the normalization of the laboratory parameters, clinical and radiographic absence of infection, and limb salvage. Perioperative complications, bacterial agents, number of procedures, and antibiotic protocol were recorded.

Results: Forty-one patients who met inclusion criteria (5 PI and 36 PO) included 28 men and 13 women with a mean age of 45 years (range, 16 to 88). Follow-up averaged 22 months (range, 1 to 62). Cierny-Mader host classes were B-local-7, B-systemic/local-20, and C-14. The pathogen was identified by microbiological culture and sensitivities in 33 patients; the remaining 8 patients were treated with empiric broad-spectrum antibiotics. Deep pelvic infections resulted from 17 traumatic injuries, 12 decubitus ulcers, 5 metabolic causes, 2 arthroplasties, 2 intravenous drug abuse, and 3 other postoperative related causes. Comorbidities averaged 1.9 per patient (range, 0 to 8), and the patients underwent an average of three operations each (range, 0 to 11). Ultimately, for resolution of the most severe infections, 14 hemipelvectomies (8 total and 6 internal) and 1 resection arthroplasty were required; the remainder resolved with multiple irrigations and debridements combined with antibiotics beads. There were 27 good results (66 %). Three patients required chronic antimicrobial suppression, and eight patients suffered loss of a limb. There were three deaths and all occurred within 12 days of surgery. Most wounds were polymicrobial, the most common pathogens included coagulase positive S. aureus, Enterococcus, Bacteroides, coagulase negative S. aureus, and Pseudomonas. Twenty-five antimicrobial agents were used; the most common in order of incidence were vancomycin, cefazolin, levofloxacin, oxacillin, ampicillin, ciprofloxacin, gentamicin, rifampin, and unasyn. A mean of 1.9 agents per patient was used. Duration of antibiotic therapy ranged from 4 weeks to chronic suppression. Surgical complications included one postoperative case of pneumonia, one iatrogenic bladder perforation, and postoperative wound drainage in numerous patients.

Discussion: Deep pelvic infections are difficult and complex problems associated with significant morbidity and mortality. Our results demonstrate the outcome for a standard protocol using surgical and antimicrobial therapies in the treatment of deep pelvic infections. Associated injuries and comoribund conditions further complicated the treatment paradigm. The three deaths were directly related to the severity of the initial infection in two cases and to severe life threatening injuries in association with an open pelvic fracture. Tumor principles that aim to achieve clean margins, free of infection, are applicable to deep pelvic infection surgery but are difficult to implement because patients were often metabolically or locally compromised or both. Furthermore, complete resection of the infected tissues may not be possible in the pelvis due to the complex anatomy and essential structures involved. A total cure is not always feasible. Surgical intervention often lessened the frequently life-threatening metabolic stress incurred by infection. Despite aggressive treatment by surgeons who are skilled in pelvic surgery, the complications and complexity of these conditions remain high. We believe that a multi-specialty team that includes an orthopedic surgeon experienced in pelvic surgery best treats such a severe condition.