Authors: Parvizi J, Huang R, Zmistowski B, Baron D.
Rothman Institute at Thomas Jefferson University Hospital Philadelphia, PA
Title: Surgical Treatment of Periprosthetic Joint Infection: Prognostic Classification
Background: It is our belief that the outcome of surgical treatment for periprosthetic joint infection (PJI) is under the influence of numerous host, organism and surgical factors. Based on current literature, there is no way of quantifying these risks. This study was designed to develop a prognostic classification for surgical treatment of PJI, which will allow a more informed discussion with the patients and may also influence the type of surgical treatment selected for patients with PJI.
Hypothesis/Purpose: What are the risk factors that can be used to determine the prognosis of patients undergoing treatment of PJI?
Methods: We retrospectively reviewed 187 consecutive patients with PJI of the hip treated at our institution with two stage revision between January 2000 and March 2010. Six patients died before reimplantation and were therefore excluded from our analyses. Of the remaining 181 patients, 71 patients required reoperation for infection control. Medical records were reviewed to identify patient demographics, comorbidities, microorganism factors, reoperations, and complications. Treatment failure was defined as any reoperation for infection. Logistic regression analyses were utilized to compare suspected risk factors between the treatment failure and controlled infection groups. A nomogram was created to create a prognostic classification criteria that allows us to determine outcome based on presence of risk factors.
Results: There was a significant difference in outcome of surgical treatment of PJI between patients. Elevated BMI, smoking, anemia, and connective tissue disease were the comorbid factors identified as predictors of treatment failure. Of organism factors, patients with culture negative, polymicrobial, and MRSA infections were more likely to fail 2 stage treatment for PJI.
Discussion: A prognostic model was created using these variables to define the risk of treatment failure based on presence of risk factors.
Conclusion: Although individualization of care is still needed, this prognostic classification can be used as guidelines to decide on the best treatment option for patients with PJI and allow better counseling of our patients with regard to outcome of treatment and ensuing expectations. In addition, this study highlights the importance of some reversible variables such as malnutrition, obesity, uncontrolled diabetes, and poor soft tissues that could be addressed before delivery of definitive care to these patients.