Prosthetic joint infection (PJI) is an emergent concern given the wide usage of prosthetic joints in old and young population to assist with activities of daily living. While the public health burden of PJI appears to be relatively low compared with other potentially fatal infections such as blood stream infections, PJI is associated with excess morbidity and steep healthcare costs. Appropriate and timely diagnosis and management are crucial in preventing poor clinical outcomes and restoring adequate function in patients with PJI. There is lack of studies using robust epidemiologic methods to evaluate effectiveness of existing treatment protocols for PJI. We conducted retrospective studies using the Veterans Affairs (VA) database using data from 123 VA hospitals between 2003 and 2012. We also abstracted clinical data from VA medical records to achieve the objectives of this research. The effectiveness of an antibiotic — rifampin and a surgical management – exchange arthroplasty was assessed in separate studies. These treatments were adjudicated based on their effectiveness in prevention of PJI recurrence and two-year postoperative mortality, respectively.
Seven hundred thirty-one of the 2838 patients with first episode of PJI were treated with surgery and medical management for Staphylococcus aureus PJI and were retained in the thesis dataset. In the first study, we compared 300 patients treated with rifampin for the first 42 days of treatment period to 364 patients treated with antibiotics other than rifampin during the 42 days (unexposed). Overall, 255 patients (38.4%) were observed to have a treatment failure defined as recurrence of S. aureus PJI with or without a repeat surgery or death in the 90 days after surgery for the PJI. The rifampin-treated and unexposed groups did not differ on time to treatment failure (p=0.92). It was noted that patients treated with a less invasive surgery that resulted in retention of the infection prosthesis with removal of infected tissue and rifampin treatment had lower risk for treatment failure compared to patients with similar surgery but not treated with rifampin (HR=0.79, 95%CI 0.52-1.20). We also observed that patients who had surgery for removal of the infected prosthesis and were treated with rifampin had significantly greater risk for treatment failure compared to those treated with the prosthesis removal surgery but no rifampin. In our second study we further analyzed this research question in depth using advanced epidemiologic methods to attenuate any bias in our previous findings. We noted that the groups did not differ in their risk for treatment failure after matching patients on their probability of receiving rifampin treatment (HR=1.08, 95%CI 0.71-1.65). In addition, we observed that patients whose treatment decision with rifampin was determined by the rifampin prescription rate in the treating facility substantially benefited with use of rifampin antibiotic for S. aureus PJI.
In the third study we assessed the effectiveness of a surgery — exchange arthroplasty (involves removal of infected prosthesis and insertion of new prosthesis) to reduce the risk of death two years after surgery in a sample of 566 S. aureus PJI patients. Patients with this procedure were compared to patients treated with retention of their infected prosthesis and removal of infected tissue (DAIR). Overall, 90 patients (15.9%) died in the 2 years after surgery. Of these, 14 (9.9%) died after exchange surgery while 76 (17.9%) died after the DAIR procedure. We observed that the exchange surgery reduced the risk for death by almost 60% compared to the DAIR (OR=0.42, 95%CI 0.19-0.89).
In conclusion, epidemiological studies conducted as part of this thesis identified considerable burden of treatment failure (38.4%) and all-cause mortality (15.9%) among veterans treated for S. aureus PJI. Rifampin combination antibiotic regimen was not observed to be effective in reducing the burden of the infection, in comparison to other antibiotics. Exchange arthroplasty reduced the risk for death in old veteran patients compared to a less invasive but potentially life-threatening procedure such as DAIR. Choice of treatment should be made on a case-by-case basis for patients with S. aureus PJI after thorough consideration of patient characteristics.