What is the problem?
We know that periprosthetic infections (PJIs) can be devastating for the patient and costly for the health care system. Yet in spite of years of great research by many thoughtful individuals, we’re still pretty much in the dark about the treatment of patients that may have an infected arthroplasty.
A recent JBJS review, The Persistent Challenges of Diagnosing Orthopaedic Implant-Related Infections, concluded that no single test can confirm or exclude implant infection and that the field is moving toward estimating the probability of infection rather than declaring it present or absent. [1] The review is about diagnosis.
We ask, should the surgeon’s approach to a patient with a 75% probability of a periprosthetic infection (PJI) be different than that for a patient with a 25% probability of PJI? Many patients seem to lie in this intermediate zone.
The article does not address the question the surgeon and patient face at revision: should this patient’s shoulder be treated as if it were infected, and how?
Consider, for example, a patient presents with a failed shoulder arthroplasty: pain, stiffness, a loose component, a worn glenoid. Before we revise it, we'd like to know if we should treat the shoulder as a periprosthetic joint infection, recognizing that the results of cultures obtained at surgery will not be finalized until weeks after the wound is closed.
The size of the problem
Shoulder PJI may be uncommon in any one practice but is costly in total.
• Rate. In the US Nationwide Inpatient Sample, the PJI rate after primary shoulder arthroplasty was 0.98% from 2002 to 2011 and did not change significantly over that period. [2] From 2011 to 2018, the rate was 1.1%, rising from 0.8% in 2011 to 1.4% in 2018. [3] Registry data with longer follow-up give a different picture by implant type: in 17,730 Nordic primary arthroplasties, the ten-year cumulative rate of revision for infection was 1.4% overall, 3.1% for reverse arthroplasties and 8.0% for reverse arthroplasties in men. [4] In the US inpatient data, by contrast, anatomic total shoulders had the highest proportion of infections (2.0%), followed by hemiarthroplasty (1.0%) and reverse (0.3%). [3] We cannot yet say with confidence which implant type carries the higher risk.
• Cost per patient. At a single high-volume center, the mean hospital cost per shoulder for two-stage reimplantation was $35,824, compared with $16,068 for a primary arthroplasty, in 2013 dollars. [5] These figures leave out outpatient costs, antibiotics between stages, therapy, extended-care stays, reoperations after the second stage and lost work. [5] The cost of a PJI is incurred on top of the cost of the original arthroplasty. [2]
• National cost. Hospital charges for shoulder PJI in the US rose 324%, from $44.8 million in 2011 to $190.3 million in 2018, and are projected to approach $500 million annually by 2030. [3] Charges are not costs, and administrative codes can miss the second stage of a two-stage exchange, so these totals are estimates. [5]. The cost of physician time and patient hardship are not included in these numbers.
These numbers do not tell us what we most need to know to manage it: for a patient with a given presentation, which treatment leads to the best comfort and function over the long haul. Registries record a revision for infection as an endpoint; they do not record how the shoulder works afterward. The studies that report treatment results use inconsistent definitions of infection [6, 7], few report patient-reported outcomes, and none compares treatments in matched patients.
For a complication of this frequency and expense, we have remarkably little operational intelligence on how to manage patients at risk for it.
What is an infection?
Bacterial infection is not the presence of bacteria; it is bacteria doing harm. The 2025 International Consensus Meeting on Musculoskeletal Infection (ICM) takes a similar position. It notes that joints carry their own microbiome, calls the distinction between pathogens and normal flora “an important but unresolved question,” and suggests that the host’s inflammatory status, not the mere detection of an organism, be an important test of whether an infection is present. [8]
We carry about 38 trillion bacteria, roughly as many as our own cells. [9] Most of them keep us healthy, and Cutibacterium is one of them. It is the most abundant bacterium on the healthy, oily skin of the face, chest and back, where it helps keep the surface acidic, competes with potential pathogens for space and nutrients, and limits oxidative stress. [10] In laboratory cultures it modifies the tight junctions between skin cells that form the skin's barrier. [11] It is the loss of Cutibacterium, or of the diversity of its strains, that is associated with skin disorders; people with acne carry about as much Cutibacterium as people without. [10] Removal of Cutibacterium can make the skin less healthy.
Around the shoulder, Cutibacterium lives in the sebaceous glands, below the reach of skin preparation. The incision for shoulder arthroplasty cuts through these glands: after chlorhexidine preparation, 40% of dermal biopsies at open shoulder surgery grew Cutibacterium. [12] Every shoulder arthroplasty wound is inoculated with Cutibacterium.
When a failed arthroplasty is revised, three questions are often blended into one:
1. Are bacteria present?
2. Are those bacteria contributing to the pain, stiffness or loosening?
3. Will treating this shoulder as a PJI change the patient's outcome?
Attempting to “define” PJI does not settle the matter. A systematic review of 20 studies of revision shoulder arthroplasty found that 6 (27%) gave no explicit definition of infection. [6] A 2025 review of 231 shoulder studies found the diagnostic workup still inconsistent. [7] A 2026 review in JBJS notes that there is no true diagnostic gold standard and that up to 15% of hip and knee cases remain inconclusive under the 2018 ICM criteria, which most diagnostic studies still use. [1] The 2025 ICM changed its recommendations on 37 (22%) of the 172 questions it shared with the 2018 meeting; 23% of those shared questions concerned the shoulder. [13]
How often failed shoulders grow bacteria
About half of failed shoulder arthroplasties will grow bacteria at revision, most often Cutibacterium, and more often in men.
• Of 193 revisions for pain, loosening or stiffness, 108 (56%) were culture positive. Of the positive cultures, 70% grew Cutibacterium and 55% took more than a week to grow. Male sex, humeral osteolysis and cloudy fluid each increased the likelihood of a positive Cutibacterium culture dramatically. [14]
• Of 221 revisions without obvious infection, at least one positive culture was found in 54% of loose glenoid components and 51% of secure ones (P = .77). [15]
• Deep cultures at revision were positive in 12 of 20 men and 1 of 20 women. [16]
• In a multicenter series of 490 revisions, 14% met ICM criteria for definite infection and 57% were classed as unlikely. Cutibacterium was the most common organism in both groups: 25% of definite and 27% of non-definite cases. The definite infections that grew Cutibacterium presented with gross purulence 72% of the time. [17]
Two points follow. The chance of bacteria in a failed shoulder is high before we know anything else about it. And the organism found in "definite" infections is the same one found in shoulders classified as "probable," "possible" or "unlikely" infections. Authors suggest that labeling organisms "virulent" or "low virulence" may need reconsideration, and that classification might rest better on the clinical presentation than on the species. [17]
Why the tests cannot settle it
An “obvious” infection may be signaled by pus, draining sinus, abnormal inflammatory labs, fever, erythema, and/or positive aspirate culture for virulent organisms. In a failed shoulder without obvious infection, no test available before or during surgery is accurate enough to rule bacteria in or out at the moment the surgical and postoperative antibiotic decisions need to be made.
Serum tests. In 579 revision shoulder arthroplasties, CRP discriminated poorly even in definite infections: at its best cutoff of 5.9 mg/L, the area under the curve was 0.66, with 82% sensitivity and 54% specificity. In the remaining revisions, ESR performed worse than chance. [18] The 2025 ICM recommends CRP and ESR before every hip and knee revision, while noting that serology is often normal in infections with Cutibacterium and coagulase-negative staphylococci. [19]
Synovial fluid. Aspiration of the shoulder often yields little or no fluid. Where fluid was obtained in 425 revisions, the synovial neutrophil percentage discriminated best, with cutoffs between 72% and 78%. [20] In 105 aspirated shoulders, alpha-defensin had 75% sensitivity and 96% specificity; leukocyte esterase had 50% and 87%. [21] Aspiration is likely to miss Cutibacterium. [22] In a four-center series, aspiration detected none of the infections found at open biopsy; percutaneous synovial biopsy detected 37%. [23] For hip and knee revisions, the 2025 ICM named leukocyte esterase the best intraoperative test (pooled sensitivity 82.5%), with the caveat that its sensitivity is limited in low-grade infection; [24] in the shoulder its sensitivity was 50%. [21]
Frozen section. In the same 425 revisions, frozen section detected only 20% to 28% of the shoulders with bacteria among the cases other than those with definite infection, which are the cases in which we need help. [20]
Nuclear imaging. Labeled white-cell SPECT/CT failed to diagnose chronic shoulder infection. [25]
What this means in numbers. Start with a hypothetical patient with a 50% chance of bacterial presence. Suppose the CRP is below 5.9 mg/L. With 82% sensitivity and 54% specificity, a normal CRP has a likelihood ratio of about 0.34, and the chance of bacteria falls from 50% to about 25%. One shoulder in four has not been ruled out, and those figures come from the definite infections, where CRP performs best and where we need it the least.
How useful is this test in determining treatment?
Even the cultures are ambiguous
A positive culture does not prove infection, and a negative set does not exclude it.
Positive cultures without infection. In primary open shoulder surgery, 7 of 54 sterile control specimens (13%) grew bacteria, and 17% of shoulders without prior surgery had at least one positive Cutibacterium culture. [26] Air swabs exposed for 5 seconds in laminar-flow operating rooms grew Cutibacterium in 6 of 40 cases (15%); none of those patients had an infection at 90 days. [27] Positive cultures are common even at primary shoulder arthroplasty. [28]
Negative cultures with bacteria present. In 137 Cutibacterium-positive revisions with at least four specimens, a mean of 43% of the specimens from each positive shoulder showed no growth. [29] The organism is unevenly distributed, so the number and source of specimens change the answer; the rate of positive shoulders rises with the number of specimens taken. [14] In a hip and knee series, ultrasonic disruption of biofilm during revision raised the rate of positive cultures from the surgical-field fluid from 40% to 76%, revealing bacteria that standard sampling had missed. [30]
What helps sort them. Load and timing. Every probable true-positive Cutibacterium culture turned positive by day 11, while 4 of 9 probable contaminants turned positive later; true positives also had more positive specimens. [31]
This is why we take five deep tissue or explant specimens, culture them for 14 days on aerobic and anaerobic media, and grade each plate from 0 to 4+. The result is a measure of bacterial load and where the bugs are – better information than characterizing a shoulder as “culture positive” or “culture negative”. The 2025 ICM recommendations are similar: four to six specimens, at least three of them tissue; anaerobic incubation for up to 14 days; and both solid and liquid media for Cutibacterium. [8]
Do positive cultures change the outcome?
When failed shoulders are revised in a single stage and given antibiotics, those that turn out to be culture positive do as well as those that do not.
• In 55 single-stage revisions without obvious infection, 27 shoulders with at least two Cutibacterium-positive cultures (89% male) were compared with 28 with zero or one. All received antibiotics adjusted to the cultures. In the culture-positive group, the Simple Shoulder Test score rose from 3.2 to 7.8 at a mean of 46 months, at least as good as in controls. Three patients (11%) in each group needed further surgery for pain or stiffness; none of those re-revisions was culture positive. Fourteen patients reported antibiotic side effects. [32]
• In 92 revisions, patients judged at high risk received intravenous antibiotics and those at low risk received oral antibiotics, both adjusted to the cultures. At a mean of 4.1 years, infection-free survival was 91%; infection-free survival, revision-free survival and patient-reported outcomes were similar between groups, but the intravenous group had more antibiotic adverse events. [33]
• In 124 revisions for reasons other than suspected infection, 27 (22%) had unexpected positive cultures, mostly Cutibacterium. Gains in shoulder score and motion did not differ from those of culture-negative shoulders. [34]
• In 65 conversions of anatomic to reverse arthroplasty with the humeral stem retained, all classed as unlikely infection by ICM criteria, 9 (14%) had positive cultures, 8 of them Cutibacterium. Only 6 patients received postoperative antibiotics, so at least 3 culture-positive shoulders went untreated. Two patients (3%) needed reoperation for infection; neither had a positive culture at the index revision. [35]
• In an early series of 28 revisions presumed aseptic, 8 (29%) were culture positive, and 2 of those 8 later developed infection. [36]
What these studies demonstrate: Culture-positive failed shoulders revised in one stage and given antibiotics reach outcomes similar to those of culture-negative shoulders. No study has randomized culture-positive shoulders to antibiotics or none, so the benefit of the antibiotics themselves is inferred, not shown. The stem-retention series suggests that some low-load positives do well without treatment; the early series reminds us that some unexpected positives go on to declare themselves. Both series are small. Here our practice departs from the 2025 ICM, which advises that an organism found without an inflammatory response not be interpreted as a pathogen. [8] The reason for the difference is in the threshold argument below.
Every treatment carries its own risk
The treatments for infection range from a few weeks of oral pills to months with removal of a shoulder implant; the more aggressive treatments endanger bone and soft tissue, imposing their harms on every patient who receives them, infected or not.
• Two-stage exchange. Of 49 shoulders scheduled for two-stage exchange, 14 (29%) were never reimplanted. Mortality was 10% within 90 days of implant removal and 25% at latest follow-up. [37]
• One stage versus two in the shoulder. Across 13 single-stage and 30 two-stage series, reinfection was 6.3% versus 10.1%, a difference that was not significant; two-stage cases had more MRSA (9.7% versus 2.5%), so the comparison is confounded. [38] A 2024 meta-analysis found reinfection of 10.9% versus 12.9% and an average of 1.16 versus 2.25 further operations, with selection bias as a major limitation. [39]
• One stage versus two, randomized, in the hip and knee. In 323 chronic hip and knee infections randomized to one or two stages, success at 2 years was 97% (131 of 135) versus 91% (112 of 123), meeting the criterion for noninferiority. Prior revisions, immunosuppressed patients, and culture-negative and fungal infections were excluded. There is no such trial in the shoulder. [40] A single-arm hip and knee series of single-stage revision with ultrasonic debridement reported 96% apparent freedom from infection at a mean of 29 months. [30] Of course there is no test for shown the absence of bacteria.
• Consensus. The 2025 ICM found no absolute contraindications to one-stage exchange for chronic hip and knee infection (90% delegate agreement), considered it viable for culture-negative infections and, when the soft tissues can be closed, for joints with a sinus tract, and recommended at least 6 weeks of antibiotics after a one-stage exchange. [41] It found too little evidence to set the duration of antibiotics between the stages of a two-stage exchange or to support an antibiotic holiday before reimplantation. [42]
• Oral versus intravenous antibiotics, randomized. In 1,054 bone and joint infections, treatment failure at 1 year was 14.6% with intravenous and 13.2% with oral antibiotics; oral therapy was noninferior. [43]
• Antibiotic harms after shoulder revision. Among 175 revisions, 19% of patients had adverse effects from postoperative antibiotics, more often with the intravenous protocol. [44]
The shoulder comparisons come from observational series in which sicker patients with more virulent organisms were chosen for two stages, which makes two-stage results look worse than they would in matched patients. Even so, nothing in the shoulder literature shows that two stages produce better comfort or function.
Tipping the balance: the host and the operating room
Infection is a balance between bacteria and host, and part of that balance is in our hands before and during surgery.
• The 2025 ICM found preoperative anemia to be a risk factor for infection (in shoulder arthroplasty, odds ratio 2.04) and allogeneic transfusion to be associated with infection (risk ratio 1.94). It found tranexamic acid associated with less periprosthetic infection (odds ratio 0.69), although delegate agreement on that point was only 63%; it favored aspirin as the thromboprophylaxis with the lowest infection risk. [45]
• In 7,713 shoulder arthroplasties, cefazolin prophylaxis, compared with vancomycin or clindamycin, was associated with 69% less periprosthetic infection and 78% less Cutibacterium infection. [46]
• The 2025 ICM concluded that a corticosteroid injection within 3 months of shoulder arthroplasty is associated with a higher risk of PJI. [13]
• In 33,470 total shoulder arthroplasties, surgical-site infection rose once operative time passed 180 minutes. [47]
• Neither chlorhexidine wash nor benzoyl peroxide soap used at home removed Cutibacterium from the shoulder skin, [48] and after chlorhexidine preparation, Cutibacterium repopulated the skin surface within an hour. [49] Hydrogen peroxide added to the skin preparation reduced but did not eliminate Cutibacterium in a controlled trial. [50]
• Dilute povidone-iodine lavage before closure was associated with fewer acute infections after hip and knee arthroplasty (0.97% versus 0.15%). [51]
• Vancomycin prevented Cutibacterium growth in a laboratory model of shoulder arthroplasty, [52] and a cost analysis concluded that topical vancomycin pays for itself if it prevents only a small number of infections. [53]
Most of these are associations from observational data, and few are specific to shoulder revision.
When to treat: a threshold, not a diagnosis
Consider treating a failed shoulder arthroplasty as if it were infected, and scale the intensity of treatment to the likelihood of infection, the virulence of the organism and what the patient, their bone and soft tissues can tolerate.
The idea comes from clinical decision analysis. Any treatment does some good in the patients who need it and some harm in the patients who do not. Whether it is worth giving depends on how likely it is that this patient needs it.
Here is a hypothetical illustration. Imagine 100 failed shoulders. Suppose that treating an infected shoulder with oral amoxicillin prevents 10 units of harm, such as a recurrent failure and another operation. Suppose that treating an uninfected shoulder causes 1 unit of harm, from side effects, cost and resistance. If 9 of the 100 shoulders are infected, treating all 100 prevents 90 units of harm and causes 91: roughly a draw. If more than 9 in 100 are infected, treating does more good than harm; if fewer, it does more harm than good. That break-even point, about 9%, is the treatment threshold.
Neither number has been measured for the shoulder, but the logic explains why the answer depends on the treatment. When the treatment carries little harm, the threshold is low. A failed shoulder starts near a 50% chance of having bacteria in it, far above that threshold, and no test we have can bring it below. That supports a single-stage exchange with oral antibiotics for every revised shoulder as long as it can be accomplished without significant bone or soft tissue damage. A two-stage exchange is different. For an uninfected patient, months without a shoulder arthroplasty, a failed reimplantation or death is a harm nearly as large as the benefit to an infected one. When the harm equals the benefit, the threshold is 50%; when it is larger, the threshold is higher still. Thus, two-stage exchange may be appropriate for shoulders in which infection is obvious: purulence, a sinus tract or sepsis.
In practice:
1. Before surgery, estimate the starting probability. Male sex, humeral osteolysis, onset of pain or stiffness after a good early result, and prior surgery raise it.
2. Decide where the shoulder lies on the spectrum. Fever, a sinus tract, purulence or sepsis, whatever the organism, suggests aggressive debridement matched to what the patient can tolerate.
3. For most cases, consider the single stage. Take five deep specimens and a frozen section, remove the components if this can be accomplished without significant bone and soft tissue damage, debride thoroughly, irrigate copiously, lavage with povidone-iodine, apply topical vancomycin powder and reimplant without cement.
4. Start oral antibiotics at once. We use amoxicillin 1 g three times daily for 3 weeks while the cultures incubate, extended to 6 months if they are positive.
5. Support the host. Correct anemia and malnutrition; we recommend 2 g of protein per kg per day and 2,000 IU of vitamin D daily until healed.
The cultures arrive after the important decision has been made. Their job is assure that the antibiotics initiated are appropriate the culture findings and to set the duration of antibiotic treatment. Infectious disease consultation is often helpful in making these decisions.
What we still do not know
Several links in this reasoning have never been tested directly.
• No trial has randomized culture-positive shoulders, once the cultures are known, to stop or continue antibiotics, so we do not know how much the antibiotics contribute, or whether low-load positives need treatment at all.
• No matched trial has compared single-stage and two-stage revision in the shoulder.
• The optimal duration of oral antibiotics after a culture-positive revision is unknown.
• The literature says almost nothing about the failed shoulder that is not being revised: the patient managed without surgery or with surgery that does not include implant exchange.
• Most importantly, data are lacking on the clinical outcomes (i.e., patient comfort and function) for patients with different sets of findings (e.g., abnormal lab tests, purulence, draining sinus vs none of these; bacteria recovered; bacterial load) having different treatments (e.g., non-operative, debridement only, single-stage revision, two-stage revision).
Conclusion
A lot of good research has gone into the definition of shoulder periprosthetic infection.
What is lacking is evidence on clinical outcomes–– patient comfort and function––, for the common clinical presentations managed with the common treatment strategies. The 2025 ICM reached a similar view: with 87% delegate agreement, it recommended that patient-reported outcomes take priority over infection eradication, reoperation and continued antibiotics for reporting the success of PJI treatment. It found that only one randomized trial of PJI treatment had used a patient-reported outcome as its primary endpoint. [55]
The path forward may lie in (1) developing a classification system for the most common clinical presentations, for example by history, physical findings, laboratory findings, the presence of purulence or a draining sinus, and bacterial load; (2) developing a classification system for the most common treatment strategies, from non-operative care to debridement, single-stage and two-stage revision; and (3) mapping the outcomes of each strategy onto each presentation.
Are there bugs in there? How much does it matter?
Follow on twitter/X: https://x.com/RickMatsen
Follow on facebook: https://www.facebook.com/shoulder.arthritis
Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133/
References
1. Lum ZC, Cohen-Rosenblum A, Yao JJ, et al. The persistent challenges of diagnosing orthopaedic implant-related infections. J Bone Joint Surg Am. 2026;108(15):1142-8. doi:10.2106/JBJS.25.01516
2. Padegimas EM, Maltenfort M, Ramsey ML, Williams GR, Parvizi J, Namdari S. Periprosthetic shoulder infection in the United States: incidence and economic burden. J Shoulder Elbow Surg. 2015;24(5):741-6. doi:10.1016/j.jse.2014.11.044
3. Schick S, Elphingstone J, Murali S, et al. The incidence of shoulder arthroplasty infection presents a substantial economic burden in the United States: a predictive model. JSES Int. 2023;7(4):636-41. doi:10.1016/j.jseint.2023.03.013
4. Moeini S, Rasmussen JV, Salomonsson B, et al. Reverse shoulder arthroplasty has a higher risk of revision due to infection than anatomical shoulder arthroplasty: 17 730 primary shoulder arthroplasties from the Nordic Arthroplasty Register Association. Bone Joint J. 2019;101-B(6):702-7. doi:10.1302/0301-620X.101B6.BJJ-2018-1348.R1
5. Baghdadi YMK, Maradit-Kremers H, Dennison T, et al. The hospital cost of two-stage reimplantation for deep infection after shoulder arthroplasty. JSES Open Access. 2017;1(1):15-8. doi:10.1016/j.jses.2017.02.001
6. Hsu JE, Somerson JS, Vo KV, et al. What is a "periprosthetic shoulder infection"? A systematic review of two decades of publications. Int Orthop. 2017;41(4):813-22. doi:10.1007/s00264-017-3421-6
7. Clifford AL, Hurley E, Rowe D, et al. How is infection diagnostic criteria for shoulder periprosthetic joint infection reported in literature: systematic review. JSES Int. 2025;9(1):219-25. doi:10.1016/j.jseint.2024.09.022
8. Rottman M, Gordon AM, Kang KS, et al. 2025 ICM: sample collection and agent identification. J Arthroplasty. 2025;41(1 Suppl 1):S157-63. doi:10.1016/j.arth.2025.08.040
9. Sender R, Fuchs S, Milo R. Revised estimates for the number of human and bacteria cells in the body. PLoS Biol. 2016;14(8):e1002533. doi:10.1371/journal.pbio.1002533
10. Rozas M, Hart de Ruijter A, Fabrega MJ, et al. From dysbiosis to healthy skin: major contributions of Cutibacterium acnes to skin homeostasis. Microorganisms. 2021;9(3):628. doi:10.3390/microorganisms9030628
11. Bolla BS, Erdei L, Urbán E, et al. Cutibacterium acnes regulates the epidermal barrier properties of HPV-KER human immortalized keratinocyte cultures. Sci Rep. 2020;10(1):12815. doi:10.1038/s41598-020-69677-6
12. Phadnis J, Gordon D, Krishnan J, et al. Frequent isolation of Propionibacterium acnes from the shoulder dermis despite skin preparation and prophylactic antibiotics. J Shoulder Elbow Surg. 2016;25(2):304-10. doi:10.1016/j.jse.2015.08.002
13. Ferrini A, He M, Bernaus M, Veloso M, Mont MA, Parvizi J. International Consensus Meeting on Orthopaedic Infection: differences between ICM 2018 and ICM 2025. J Arthroplasty. 2026;41(2):304-7. doi:10.1016/j.arth.2025.10.113
14. Pottinger P, Butler-Wu S, Neradilek MB, et al. Prognostic factors for bacterial cultures positive for Propionibacterium acnes and other organisms in a large series of revision shoulder arthroplasties performed for stiffness, pain, or loosening. J Bone Joint Surg Am. 2012;94(22):2075-83. doi:10.2106/JBJS.K.00861
15. Lucas RM, Hsu JE, Whitney IJ, et al. Loose glenoid components in revision shoulder arthroplasty: is there an association with positive cultures? J Shoulder Elbow Surg. 2016;25(8):1371-5. doi:10.1016/j.jse.2015.12.026
16. Matsen FA 3rd, Butler-Wu S, Carofino BC, et al. Origin of Propionibacterium in surgical wounds and evidence-based approach for culturing Propionibacterium from surgical sites. J Bone Joint Surg Am. 2013;95(23):e181. doi:10.2106/JBJS.L.01733
17. Ricchetti ET, Collins AP, Khazzam MS, et al. Evaluation of the International Consensus Meeting (ICM) diagnostic categories for shoulder periprosthetic joint infection: Cutibacterium acnes predominates even in definite infections. J Shoulder Elbow Surg. 2026;35(10):2362-71. doi:10.1016/j.jse.2026.04.053
18. Collins AP, Whitson AJ, Budge MD, et al. Can standard preoperative serum laboratory tests predict bacterial presence at the time of revision shoulder arthroplasty? J Shoulder Elbow Surg. 2026;35(10):2372-81. doi:10.1016/j.jse.2026.04.036
19. Pupaibool J, Tarabichi S, Shahi A, et al. 2025 ICM: serological diagnosis of surgical site infection (SSI)/periprosthetic joint infection (PJI). J Arthroplasty. 2025;41(1 Suppl 1):S248-60. doi:10.1016/j.arth.2025.10.102
20. Collins AP, Whitson AJ, Duquin TR, et al. Defining optimal synovial aspirate thresholds and utility of frozen sections in predicting bacterial presence at the time of revision shoulder arthroplasty. J Shoulder Elbow Surg. 2026;35(11):2607-17. doi:10.1016/j.jse.2026.05.005
21. Unter Ecker N, Koniker A, Gehrke T, et al. What is the diagnostic accuracy of alpha-defensin and leukocyte esterase test in periprosthetic shoulder infection? Clin Orthop Relat Res. 2019;477(7):1712-8. doi:10.1097/CORR.0000000000000762
22. Streck LE, Forster J, von Hertzberg-Boelch SP, et al. The role of synovial fluid aspiration in shoulder joint infections. BMC Musculoskelet Disord. 2022;23(1):390. doi:10.1186/s12891-022-05285-x
23. Lapner P, Nam D, Cheema A, et al. Diagnostic accuracy of preoperative percutaneous synovial biopsy and aspirate compared with open biopsy for prosthetic shoulder infections. J Shoulder Elbow Surg. 2025;34(2):441-8. doi:10.1016/j.jse.2024.08.016
24. Mortazavi SMJ, Abdelazeem A, Abdelnasser MK, et al. 2025 ICM: diagnosis of surgical site infection (SSI)/periprosthetic joint infection (PJI) with samples from the wound and clinical signs. J Arthroplasty. 2025;41(1 Suppl 1):S243-7. doi:10.1016/j.arth.2025.10.107
25. Falstie-Jensen T, Daugaard H, Søballe K, et al. Labeled white blood cell/bone marrow single-photon emission computed tomography with computed tomography fails in diagnosing chronic periprosthetic shoulder joint infection. J Shoulder Elbow Surg. 2019;28(6):1040-8. doi:10.1016/j.jse.2018.10.024
26. Mook WR, Klement MR, Green CL, et al. The incidence of Propionibacterium acnes in open shoulder surgery: a controlled diagnostic study. J Bone Joint Surg Am. 2015;97(12):957-63. doi:10.2106/JBJS.N.00784
27. Namdari S, Nicholson T, Parvizi J. Cutibacterium acnes is isolated from air swabs: time to doubt the value of traditional cultures in shoulder surgery? Arch Bone Jt Surg. 2020;8(4):506-10. doi:10.22038/abjs.2020.40642.2095
28. Wong JC, Schoch BS, Lee BK, et al. Culture positivity in primary total shoulder arthroplasty. J Shoulder Elbow Surg. 2018;27(8):1422-8. doi:10.1016/j.jse.2018.05.024
29. Ahsan ZS, Somerson JS, Matsen FA 3rd. Characterizing the Propionibacterium load in revision shoulder arthroplasty: a study of 137 culture-positive cases. J Bone Joint Surg Am. 2017;99(2):150-4. doi:10.2106/JBJS.16.00422
30. Ji B, Yang C, Aimaiti A, et al. Effective treatment using a single-stage revision with non-contact low frequency ultrasonic debridement in the treatment of periprosthetic joint infection: a prospective single-arm study. Bone Joint J. 2025;107-B(6 Supple B):9-14. doi:10.1302/0301-620X.107B6.BJJ-2024-0848.R2
31. Frangiamore SJ, Saleh A, Grosso MJ, et al. Early versus late culture growth of Propionibacterium acnes in revision shoulder arthroplasty. J Bone Joint Surg Am. 2015;97(14):1149-58. doi:10.2106/JBJS.N.00881
32. Hsu JE, Gorbaty JD, Whitney IJ, Matsen FA 3rd. Single-stage revision is effective for failed shoulder arthroplasty with positive cultures for Propionibacterium. J Bone Joint Surg Am. 2016;98(24):2047-51. doi:10.2106/JBJS.16.00149
33. Yao JJ, Jurgensmeier K, Whitson AJ, et al. Oral and IV antibiotic administration after single-stage revision shoulder arthroplasty. J Bone Joint Surg Am. 2022;104(5):421-9. doi:10.2106/JBJS.20.02263
34. Falstie-Jensen T, Lange J, Daugaard H, et al. Unexpected positive cultures after revision shoulder arthroplasty: does it affect outcome? J Shoulder Elbow Surg. 2021;30(6):1299-308. doi:10.1016/j.jse.2020.12.014
35. Shenoy DA, Parsons KE, Henriquez AR, et al. Occult infection risk with stem retention in revision anatomic to reverse shoulder arthroplasty. J Shoulder Elbow Surg. 2026. Epub ahead of print. doi:10.1016/j.jse.2026.10.007
36. Kelly JD 2nd, Hobgood ER. Positive culture rate in revision shoulder arthroplasty. Clin Orthop Relat Res. 2009;467(9):2343-8. doi:10.1007/s11999-009-0875-x
37. Akgün D, Wiethölter M, Maziak N, et al. Two-stage exchange arthroplasty for periprosthetic shoulder infection is associated with high rate of failure to reimplant and mortality. J Clin Med. 2021;10(21):5186. doi:10.3390/jcm10215186
38. Belay ES, Danilkowicz RM, Bullock GS, et al. Single-stage versus two-stage revision for shoulder periprosthetic joint infection: a systematic review and meta-analysis. J Shoulder Elbow Surg. 2020;29(12):2476-86. doi:10.1016/j.jse.2020.05.034
39. Bdeir M, Lerchl A, Hetjens S, et al. One- vs. two-stage revision for periprosthetic shoulder infections: a systematic review and meta-analysis. Antibiotics (Basel). 2024;13(5):440. doi:10.3390/antibiotics13050440
40. Fehring TK, Otero JE, Fehring KA, et al. One-stage versus two-stage exchange arthroplasty for periprosthetic joint infection. J Bone Joint Surg Am. 2026;108(14):1070-82. doi:10.2106/JBJS.25.00713
41. Hansen E, Ji B, Dietz MJ, et al. 2025 ICM: one-stage exchange. J Arthroplasty. 2025;41(1 Suppl 1):S326-40. doi:10.1016/j.arth.2025.10.099
42. Elhence A, Böhler C, Kolhoff F, et al. 2025 ICM: two-stage. J Arthroplasty. 2025;41(1 Suppl 1):S385-98. doi:10.1016/j.arth.2025.10.095
43. Li HK, Rombach I, Zambellas R, et al. Oral versus intravenous antibiotics for bone and joint infection. N Engl J Med. 2019;380(5):425-36. doi:10.1056/NEJMoa1710926
44. Yao JJ, Jurgensmeier K, Woodhead BM, et al. The use and adverse effects of oral and intravenous antibiotic administration for suspected infection after revision shoulder arthroplasty. J Bone Joint Surg Am. 2020;102(11):961-70. doi:10.2106/JBJS.19.00846
45. Mortazavi SMJ, Palmer A, Yildiz F, et al. 2025 ICM: anemia, blood-saving, and thromboprophylaxis. J Arthroplasty. 2025;41(1 Suppl 1):S19-29. doi:10.1016/j.arth.2025.10.082
46. Marigi EM, Bartels DW, Yoon JH, et al. Antibiotic prophylaxis with cefazolin is associated with lower shoulder periprosthetic joint infection rates than non-cefazolin alternatives. J Bone Joint Surg Am. 2022;104(10):872-80. doi:10.2106/JBJS.21.00445
47. Schmitt MW, Chenault PK, Samuel LT, et al. The effect of operative time on surgical-site infection following total shoulder arthroplasty. J Shoulder Elbow Surg. 2023;32(11):2371-5. doi:10.1016/j.jse.2023.05.010
48. Hsu JE, Whitson AJ, Woodhead BM, et al. Randomized controlled trial of chlorhexidine wash versus benzoyl peroxide soap for home surgical preparation: neither is effective in removing Cutibacterium from the skin of shoulder arthroplasty patients. Int Orthop. 2020;44(7):1325-9. doi:10.1007/s00264-020-04594-x
49. Hsu JE, Whitson AJ, Van Dyke R, et al. Dynamics of Cutibacterium repopulation onto the skin surface of the shoulder after chlorhexidine application. Int Orthop. 2023;47(6):1511-5. doi:10.1007/s00264-023-05786-x
50. Chalmers PN, Beck L, Stertz I, Tashjian RZ. Hydrogen peroxide skin preparation reduces Cutibacterium acnes in shoulder arthroplasty: a prospective, blinded, controlled trial. J Shoulder Elbow Surg. 2019;28(8):1554-61. doi:10.1016/j.jse.2019.03.038
51. Brown NM, Cipriano CA, Moric M, et al. Dilute betadine lavage before closure for the prevention of acute postoperative deep periprosthetic joint infection. J Arthroplasty. 2012;27(1):27-30. doi:10.1016/j.arth.2011.03.034
52. Miquel J, Huang TB, Athwal GS, et al. Vancomycin is effective in preventing Cutibacterium acnes growth in a mimetic shoulder arthroplasty. J Shoulder Elbow Surg. 2022;31(1):159-64. doi:10.1016/j.jse.2021.06.015
53. Hatch MD, Daniels SD, Glerum KM, Higgins LD. The cost effectiveness of vancomycin for preventing infections after shoulder arthroplasty: a break-even analysis. J Shoulder Elbow Surg. 2017;26(3):472-7. doi:10.1016/j.jse.2016.07.071
54. Matsen FA 3rd, Whitson A, Hsu JE. Preoperative skin cultures predict periprosthetic infections in revised shoulder arthroplasties. JBJS Open Access. 2020;5(4):e20.00095. doi:10.2106/JBJS.OA.20.00095
55. Manning L, Slullitel PA, Buttaro MA, et al. 2025 ICM: monitoring and declaring success. J Arthroplasty. 2025;41(1 Suppl 1):S437-42. doi:10.1016/j.arth.2025.10.104
