Applying the new shoulder periprosthetic joint infection consensus definition to a case series of revision shoulder arthroplasty procedures to assess concordance between consensus definitions and diagnoses

Document Type


Publication Date


Publication Title

Seminars in Arthroplasty: JSES


Purpose As the number of shoulder arthroplasty procedures performed rises yearly, so does the number of periprosthetic joint infections (PJIs). In this study, PJI consensus definitions were compared and contrasted in a series of revision shoulder arthroplasty cases preoperatively diagnosed as PJI. Understanding the variations in these definitions may guide PJI diagnoses, thereby improving treatment strategies and patient outcomes in the setting of infected shoulder arthroplasty.

Methods All revision shoulder arthroplasty cases with preoperatively-diagnosed or suspected PJI (determined by procedure code) performed from 2008 – 2017 at a single institution by a single surgeon (fellowship-trained in shoulder and elbow surgery) were retrospectively evaluated. Following Institutional Review Board approval, patient demographic, treatment, and laboratory data were collected. Musculoskeletal Infection Society (MSIS; 2011) and International Consensus Meeting on Orthopaedic Infections (ICM; 2013, 2018 Revision, 2018 Shoulder) definitions of PJI were applied to the data. Statistical analysis assessed significant associations between culture status and PJI classification algorithm criteria.

Results Thirty-seven patients with suspected PJI were identified; 24 culture-positive (CP) and 13 culture-negative (CN). In this series, the 2018 ICM Shoulder definition for definite infection was met at lower rates than all other definitions (CP; 71% vs. 96%; CN; 62% vs. 69%). 2018 ICM Shoulder major criteria showed stronger correlations to 2011 MSIS, 2013 ICM, and 2018 ICM Revision major criteria when “gross intra-articular pus” was excluded than when pus was included as a major criterion. 2018 ICM Revision cases determined to be infected were very strongly, positively, correlated with the 2018 ICM Shoulder cases determined to have definite or probable infections (ρ = 1.000, P < .0001). Additionally, cases classified as “definite” or “probable” infections with the 2018 ICM Shoulder definition were more likely to require reoperation for suspected recurrent infection after completion of antibiotic therapy.

Conclusions In this series, the 2018 ICM Shoulder definition and previous PJI definitions classified cases as PJI at similar rates. However, the inclusion of a third major criterion of “gross intra-articular pus” weakened the correlation with prior definitions.

Level of evidence Level IV; Case Series.





First Page


Last Page