Background: Propionibacterium acnes has been grown on culture in half of the reported cases of chronic infection associated with shoulder arthroplasty. The presence of this organism can be overlooked because its subtle presentation may not suggest the need for culture or because, in contrast to many orthopaedic infections, multiple tissue samples and weeks of culture incubation are often necessary to recover this organism. Surgical decisions regarding implant revision and antibiotic therapy must be made before the results of intraoperative cultures are known. In the present study, we sought clinically relevant prognostic evidence that could help to guide treatment decisions.
Methods: We statistically correlated preoperative and intraoperative observations on 193 shoulder arthroplasty revisions that were performed because of pain, loosening, or stiffness with the results of a Propionibacterium acnes-specific culture protocol. Regression models were used to identify factors predictive of a positive culture for Propionibacterium acnes.
Results: One hundred and eight of the 193 revision arthroplasties were associated with positive cultures; 70% of the positive cultures demonstrated growth of Propionibacterium acnes. The rate of positive cultures per shoulder increased with the number of culture specimens obtained from each shoulder. Fifty-five percent of the positive cultures required observation for more than one week. Male sex, humeral osteolysis, and cloudy fluid were each associated with significant increases of ≥600% in the likelihood of obtaining a positive Propionibacterium acnes culture. Humeral loosening, glenoid wear, and membrane formation were associated with significant increases of >300% in the likelihood of obtaining a positive Propionibacterium acnes culture.
Conclusions: Preoperative and intraoperative factors can be used to help to predict the risk of a positive culture for Propionibacterium acnes. This evidence is clinically relevant to decisions regarding prosthesis removal or retention and the need for immediate antibiotic therapy at the time of revision shoulder arthroplasty before the culture results become available.
Level of Evidence: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.
1Department of Medicine, University of Washington Medical Center, Box 356420, 1959 NE Pacific Street, Seattle, WA 98195
2Department of Laboratory Medicine, University of Washington Medical Center, Box 357110, 1959 NE Pacific Street, NW120, Seattle, WA 98195
3The Mountain-Whisper-Light Statistics, 1827 23rd Avenue East, Seattle, WA 98112-2913
4Department of Orthopedics and Sports Medicine, University of Washington Medical Center, Box 356500, 1959 NE Pacific Street, Seattle, WA 98195. E-mail address for F.A. Matsen: firstname.lastname@example.org