Management of periprosthetic joint infection of the shoulder: A narrative review

Abstract

Evidence for management of shoulder periprosthetic joint infection (PJI) has traditionally originated from the hip and knee literature. The differing microbiome, anatomy and implants used in the shoulder mean this evidence is not always directly transferrable. The 2018 Philadelphia International Consensus Meeting for the first-time produced evidence-based guidelines and diagnostic criteria relating specifically to PJI of the shoulder. These guidelines and criteria recognize the pathogenicity of lower virulence organisms in the shoulder which often means clinical presentation is less obvious than other joints.

The role of Cutibacterium acnes in shoulder PJI is the subject of increasing basic science and clinical research and advances in microbiological research may help to understand the pathology behind shoulder infections. There is new evidence that outcomes after revision shoulder arthroplasty are dependent on the virulence of the causative organism. An individualised approach to treatment considering host factors, organism, soft tissues and bone stock is recommended.

Debate continues in the literature regarding the indications of one- or two-stage revision and the latest evidence is discussed and synthesized in this review article. We advocate careful multidisciplinary team decision making for cases of shoulder PJI and recognize a limited role for debridement and implant retention in acute shoulder PJI (<6 weeks). There appears to be a role for one-stage revision in lower risk cases with low virulence organisms but caution against its’ universal adoption. In higher risk or complex cases, there remains a clear role for two-stage revision arthroplasty, and we detail the specifics of this protocol and procedure from our tertiary shoulder and elbow unit.

1. Introduction

Shoulder arthroplasty is an effective treatment for degenerative shoulder pathology ¹ and has an increasing role in managing displaced 3-or- 4 part proximal humerus fractures² with high-level evidence demonstrating greater than 90 % ten-year implant survival rates.³ There is a rising incidence of all types of shoulder replacements with one study predicting up to 20,000 procedures per year by 2050(4), a rise of more than 230 % of current volume. This increase in volume is therefore likely to be accompanied by a larger revision workload.^4,5 The third commonest cause for revision,⁶ periprosthetic joint infection (PJI) complicates up to 3 % of reverse total shoulder replacements (rTSR) compared with 0.5–1% of anatomical total shoulder replacements (aTSR).6, 7, 8, 9 The present PJI revision burden can be estimated at around 200 cases per year in the UK, forecast to more than double this figure within 10 years.⁴

This article discusses management of PJI of the shoulder with reference to modern diagnostic and treatment guidance and discussion of the rationale and specifics of surgical management in our tertiary shoulder and elbow unit. 3 main PJI diagnostic criteria are discussed; the 2011 Musculoskeletal Infection Society (MSIS) criteria¹⁰ and The 2019 European Bone and Joint Infection Society (EBJIS) criteria¹¹ relate to all joints, and the 2018 Philadelphia International Consensus Meeting (2018 Philadelphia ICM) shoulder diagnostic criteria and evidence summary^12,13 relate to the shoulder specifically. In 2025, the ICM will meet again to discuss and debate the current criteria and update guidance based on the latest evidence.

1.1. Diagnosis and microbiology

Patients presenting with PJI of the shoulder can present with overt signs of infection, or suppuration, that include some of the major criteria for diagnosis e.g., discharging sinus,^6,8,10, 11, 12 unexpected wound discharge^11,12 or wound healing problems.¹⁰ Systemic inflammatory response syndrome (SIRS), fever, gross erythema and swelling are less common than in the knee or hip but are more likely to represent infection with a virulent organism.⁹ Often, shoulder PJI commonly presents with a less obvious picture with increasing pain or reduced range of movement in a previously well-functioning shoulder replacement.14, 15, 16 A detailed surgical and medical history followed by clinical examination of the joint is proceeded by request of multiple view plain film radiographs of the affected shoulder. Signs such as periosteal scalloping, progressive lucency around the bone/cement interface or implant/bone interface (see Fig. 1) and widened joint space should be elicited by comparison with historical radiographs.^9,12,15,16

Fig. 1.

(a) Pre-revision surgery radiographs of an infected rTSR demonstrating periosteal scalloping, scapular notching and lucency around the humeral implant. (b) Post-operative first-stage radiographs showing antibiotic laden bone cement (ALBC) spacer.

Serological investigations to aid diagnosis should include WBC, neutrophil count, CRP, ESR, crucially, all of these can be normal in confirmed cases of shoulder PJI due to infection with low virulence organisms. WBC is raised in only 6.8 % of confirmed shoulder PJI(17) and is therefore absent from the MSIS, EBJIS and 2018 Philadelphia criteria.10, 11, 12 Raised neutrophil count or >80 % polymorphic neutrophils (PMN) is more sensitive than WBC count and features in all 3 criteria but can be normal.10, 11, 12 Various diagnostic criteria feature elevated CRP(10) (MSIS), or CRP >10 mg/L(11,12) (EBJIS/2018 Philadelphia ICM) but it is often normal in confirmed cases.

Bone scans are not currently recommended in the management of shoulder PJI(12) and the role of SPECT computed tomography (CT) is under review with a sensitivity of 31 % and specificity of 80 % in the hip and knee.¹⁸ CT imaging should be requested to assess bone stock of the proximal humerus and glenoid. MRI or ultrasound scans are not routinely required unless there is need to assess the rotator cuff status when considering the revision of an anatomical prosthesis or there are concerns about malignancy.

As PJI in the shoulder presents differently to the hip and knee and with a differing causative organism profile, a joint-specific approach is favoured. The 2018 Philadelphia International Consensus Meeting definition for PJI of the shoulder stratifies patients into definite, probable, possible and unlikely PJI based on the criteria listed in Table 1 below.^12,13

Table 1.

2018 Philadelphia International Consensus Meeting diagnostic criteria for PJI of the shoulder.¹².

2018 Philadelphia ICM Diagnostic Criteria for Shoulder PJI
Major Criteria	Weighting
Sinus tract OR presence of intra-articular pus OR Positive culture of phenotypically identical virulent organism	Definite PJI
Minor Criteria
Unexpected wound drainage	4
Single positive culture VIRULENCE organism	3
Single positive culture LOW VIRULENCE organism	1
Second positive culture of SAME LOW VIRULENCE organism	3
Humeral loosening (radiological or clinical)	3
Positive frozen section (5 PMN in at least 5 high-power fields)	3
Positive preoperative aspirate culture (ANY)	3
Elevated synovial neutrophil percentage >80 %	2
Elevated synovial WBC (>3000 cells/uL)a	2
ESR >30 mm/h	2
CRP >10 mg/L	2
Elevated synovial alpha-defensin	2
Cloudy appearance of synovial fluid	2
6 or >6 from minor criteria WITH identified organism	Probable PJI
6 or >6 from minor criteria WITHOUT identified organism	Possible PJI
<6 from minor criteria	Unlikely PJI

PMN, polymorphonuclear leukocyte; WBC, white blood cell; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein.

^aBeyond six weeks from recent surgery.

Much of the variance in PJI management in the shoulder compared with the hip or knee is due to frequent isolation of Cutibacterium acnes, Staphylococcus epidermidis and coagulase negative staphylococci (CONS),19, 20, 21 these organisms are commonly referred to as low virulence. Notwithstanding in one series CONS was deemed high virulence, although this likely due to grouping of multiple species.²² This makes them difficult to culture from fluid or tissue and are common skin commensals resulting in high rates of sample contamination and often do not present clinically with suppuration.^20,23, 24, 25, 26 These challenges have been addressed in the 2018 ICM Philadelphia criteria and are therefore at present our favoured diagnostic criteria for shoulder PJI.¹²

1.1.1. Cutibacterium acnes

Cutibacterium acnes is a gram-positive skin commensal bacillus that is responsible for up to 50 % of PJI of the shoulder,⁶ it expresses virulence factors that make it problematic for management of infected shoulder arthoplasty.²⁶ It is often described as low virulence, or indolent,27, 28, 29 referring to the fact that the clinical manifestations are often not as severe as other organisms, such as Staphylococcus aureus.²⁹ C. acnes has a strong and rapid propensity to form biofilms on implants (via adhesin and extracellular polysaccharide production) and has a powerful ability to evade macrophages (via resistance to enzymatic degradation).^6,26, 27, 28, 29 This is also why serum markers (WCC, CRP, neutrophil count and ESR) are often within normal limits and cannot be relied upon for diagnosis. Interleukin-6 serum levels for detecting C. acnes infection are not recommended because of a low sensitivity of 12 % and specificity of 7 %.³⁰

Its' indolence means that culture negative synovial fluid and tissue sampling are common,^24,30,31 other techniques such as 16 S ribosomal RNA PCR and sonication of retrieved implants can increase detection rates but these investigations are not yet widespread.32, 33, 34 A single sample positive for C. acnes is often thought to be skin contaminant (much like S. epidermidis),^27,34 which is why a single culture of a ‘low virulence’ organism is not diagnostic in modern PJI diagnostic criteria.^10,11 Next generation genomic sequencing techniques may lead to increased PJI detection rates in the shoulder but these do not constitute diagnostic criteria at present.³² Antimicrobial therapy to cover c. Acnes often includes a penicillin or a cephalosporin and rifampicin, which is effective at treating biofilms. Further research to understand the host-organism response and propensity to cause PJI could improve diagnosis and treatment.

1.2. Principles of management

As recommended by several authors and guidelines, a truly multi-disciplinary team (MDT) approach is needed for successful treatment of shoulder PJI, this should include a clinical microbiologist, radiologist and physiotherapist in addition to surgeons.^35,36 Also crucial to all treatment decisions are consideration of the patient's comorbidities, host factors pertinent to PJI and fitness for major (often multi-stage) surgery. Host factors, soft tissue status as well as bone loss are all considered in the BACH classification, a prognostic tool to help predict outcome of PJI.³⁷

Regardless of procedure chosen, five deep tissue samples should be taken with individual forceps and scalpels to reduce the chance of contamination and ensure maximum chance of detecting a causative organism on 2 or more samples whilst balancing the risk of culturing a contaminant.³⁸ All samples must be cultured for a minimum of 14 days to ensure low causative virulence organisms are detected.^19,24,35,38 Synovial fluid for culture and deep tissue should also be sent for histological analysis during any procedure. Patients should stop antibiotics for 2-weeks prior to any sampling or aspiration as long as they are not clinically septic or displaying a SIRS response. With regards to antibiotic therapy, decisions for choice of antibiotic should be made by the MDT microbiologist based on culture and sensitivity results considering local resistance patterns, see Table 2.

Table 2.

Evidence for antimicrobial therapy in shoulder PJI.

Evidence for antimicrobial therapy in shoulder PJI
Preoperative	Topical antibiotics e.g., Mupirocin, clindamycin.9,41
	Skin decolonisation e.g., Benzyl peroxide or chlorhexidine gluconate.12,41
	Prepare intra-operative antibiotic plan with MDT.35
Intraoperative	Intra-operative antibiotics – broad spectrum or targeted if known organism, given only after tissue and fluid sampling completed.6,35,38
	Up to 2 antibiotics to be mixed into bone cement spacer (ALBC) (if two-stage).12,35,42,43
	Consider defensive antimicrobial coating (DAC) to reduce chance of implant colonisation.44 ±
	Local antibiotics e.g., local eluting beads (e.g., Stimulan) or vancomycin powder prior to closure of wound.12 ±
Postoperative	Tissue and fluid cultured for a minimum of 14 days post-operatively.9,12,14,45
	2 weeks of intravenous therapy with switch to oral antibiotic therapy if sensitivities permit.46
	Re-discussion in MDT prior to second stage, if planned, to decide intra-operative and bone cement antibiotics.35

+-No strong evidence to support these practices in shoulder PJI.

Revision of an infected aTSR is often to a rTSR via a single or two-stage revision,³⁹ as infection quickly spreads to the rotator cuff tissues and often debridement results in little functional cuff tissue. Another factor to consider is the poor condition of the subscapularis after multiple takedowns and repairs.⁴⁰ Retention of an aTSR would be indicated in an acute infection managed with DAIR with well-fixed implants but evidence of aTSR exchange via one- or -two-stage revision is lacking likely due to factors mentioned above, our preference would be revision to a rTSR.

1.2.1. Use of arthroscopic biopsy for diagnosis

Arthroscopic tissue biopsy has emerged as a technique to evaluate the arthroplasty and increase diagnostic yield of synovial culture. A meta-analysis of 112 patients undergoing arthroscopic biopsy showed a positive predictive value of 73.6 % (62.3–81.8 %) and a negative predictive value of 89.8 % (80.6–95.0 %), vastly better than any serum or synovial investigation.⁴⁷ All arthroscopic biopsy results were consistent with tissue culture results from the revision surgery.⁴⁷ This conclusion was also corroborated by another group who also demonstrated the ability of arthroscopic biopsy to consistently identify c. Acnes.⁴⁸ This procedure is described in aspirate negative clinically suspicious cases where it is advantageous to visually inspect the joint, take tissue samples in the hope to prevent a 2-stage revision, if a low virulence organism is cultured.^47,48

1.3. Surgical options for shoulder PJI management

No randomised controlled trials (RCT) have assessed the rates of reinfection, reoperation and patient reported outcome measures (PROMS) when comparing one-versus two-stage revision shoulder arthroplasty. In absence of RCT evidence, the literature predominantly contains single surgeon or institution series which have then been pooled on several occasions into systematic reviews or meta-analyses.

Debridement and implant retention (DAIR) is generally reserved for acute infections within 6 weeks of the primary shoulder replacement⁴⁹ as supported by The British Elbow & Shoulder Society.³⁵ If clinical symptoms evolve quickly in a well-functioning prosthesis with history of a clear infection elsewhere that has spread via the haematogenous route then a DAIR can be considered. It is necessary to carefully counsel the patient about the risk of needing a two-stage revision if eradication with DAIR is not successful. There is some evidence from Europe and the USA to support this practice outside of the <6 week window.^36,49

DAIR should always be performed via an open approach, sampling described above, with thorough systematic debridement of devitalised or obviously infected tissue, pulse lavage and exchange of removable components (e.g., glenosphere, polyethylene tray or insert). Patients should be counselled that if pus or gross loosening of one or more components is found at DAIR then decision to proceed to two-stage revision is required.

The role of single-stage revision for shoulder PJI is debated extensively in the literature, the exact indications are still unclear but include factors such as low virulence organism,^12,25,27,28 favourable host factors, patient choice and soft tissue condition. The technique is important and the procedure should be a two-phase procedure with implant removal, thorough debridement and sampling phase followed by reimplantation after re-draping and opening new sterile instruments. An adequate shoulder PJI debridement was described by Frankle et al. as removal of all diseased bone and soft tissue including complete resection of capsule and bursa and thorough posterior humeral and subdeltoid space debridement.²²

In a study from the same group, reinfection rate for one-stage revision (as measured by clinical signs, serological markers and reoperation) is 4.3 % (minimum 1 year follow up)⁵⁰ rising to 5.0 % at subsequent review (minimum 2 years),²² lower than most papers reporting reinfection for two-stage revision. Similar European studies report reinfection rates of 5.9 %⁵¹ and 8.8 %⁵² following one-stage revision. Staphylococcus aureus or CONS was associated with multiple surgical procedures,⁵⁰ this may suggest that if these organisms are isolated pre-operatively then the patient may be better served with a definitive two-stage revision. Risk of failure of eradication of infection must be carefully explained to the patient. There may be a role for a single-stage approach in patients who are unfit for multiple procedures when balancing anaesthetic risk as part of a shared decision-making process.

Studies examining success of two-stage revision are more numerous, these are often limited by attrition bias, a proportion of patients who receive a first stage of debridement and an antibiotic-laden bone cement spacer (ALBC) do not wish to, or are not fit to undergo definitive implantation. Several well-recognised US institutions report reinfection rates of 18 %⁴² and 21 %.⁵³ Within our own tertiary shoulder and elbow unit we observed a similar 20 % reinfection rate of those who went onto reimplantation in a cohort of complex chronic PJI patients.⁵⁴ The literature in this area is prone to selection bias as most studies are retrospective reviews and do not capture diagnostic and surgical decision making.

Systematic reviews and meta-analyses examining revision shoulder arthroplasty are prone to heterogeneity, selection and reporting bias which makes grouping effect sizes unreliable. Many of the pooled studies have different methods to assess reinfection or ‘success’ following single stage revision, especially as detection rates of P. acnes from synovial aspirate samples are low compared with tissue biopsy ^45,55 and there are an overall greater number of two-stage studies reported.⁵⁶ The conclusion is often that single stage is as effective, if not better at infection eradication, than two-stage revision for PJI of the shoulder. These reviews are summarised in Table 3.

Table 3.

Summary of systematic reviews and meta-analyses of outcomes of one-stage and two-stage revision for shoulder PJI. *N.B. All outcomes reported or converted to reinfection rates rather than infection clearance/eradication rates for consistency.

	Institution	No. Shoulders	Findings*	Notes
Nelson, G. N. et al.17 (2016)	Rothman Institute for Orthopaedic Surgery, PA, USA	543	6.2 % reinfection rate for two-stage, 8.3 % for one-stage	38.9 % Cutibacterium acnes infections. No significant difference between one stage (with or without unexpected positive cultures) or two-stage revision protocols
George, D. et al.57 (2016)	University College London Hospitals, London, UK	217	9.2 % reinfection rate for two stage, 5.3 % for one-stage	Almost double the number of studies examining two-stage vs. one stage. No significant difference between groups.
Kunutsor, S.K. et al.58 (2019)	Musculoskeletal Research Unit, University of Bristol, UK	498	11.5 % reinfection rate for two-stage 5.3 % for one-stage.	147 one-stage, 351 two-stage. Higher infection clearance rate with one-stage vs. two-stage but significant. Robust methodology, protocol registered prospectively.
Fiore, M. et al.59 (2023)	IRCCS Istituto Ortopedico Rizzoli/University of Bologna, Italy	486	14.3 % reinfection rate for two-stage, 4.4 % for one-stage.	Significant difference reported, majority infections with P. acnes or S. epidermidis. Conclude single stage is at least as effective as two-stage revision.
Rodrigues-Lopes, R. et al.60 (2024)	University Hospital Center of Sao Joao, Porto, Portugal	711	8.8 % reinfection rate for two-stage, 1.2 % for one-stage.	185 shoulders for one-stage and 526 for two-stage revision. Mentions that the ‘right conditions’ are needed for one-stage; low virulence organism, favourable host and minimal bone loss.
Bdeir, M. et al.56 (2024)	University Medical Center Mannheim/University of Heidelberg, Germany	1048	12.9 % two-stage reinfection rate, 10.9 % one-stage reinfection rate	Significant difference between two groups. 7x number of Staph. Aureus infections in one-stage group. Predominantly C. acnes. Double number of two-stage vs one-stage procedures.

As several reviews confirm, the right conditions for a one-stage revision include having a pre-operative confirmed organism and no significant bone loss⁶⁰ but may also include absence of high-risk host factors,^11,37 acuity of infection and confidence of diagnosis based on agreed shoulder-specific diagnostic criteria.¹²

Review protocols are often not registered with PROSPERO(56,59,60) and there is variation in risk of bias tools used. Diagnostic criteria have changed over time during the period of these reviews, with 2018 ICM Philadelphia now preferred.^15,61 For reasons outlined above we recommend caution in interpretation and change in practice based on these studies. Our preferred technique for two-stage revision is detailed in the next section of this article.

1.4. RNOH 2-stage revision for shoulder PJI

We describe our unit's surgical protocol for two-stage revision for shoulder PJI with supporting literature below (see Fig. 2), we appreciate this approach is tailored to tertiary revision practice but outlines clinical decision making and important aspects of surgical practice that are important for single or two-stage revision surgery. All patients are discussed in the upper limb MDT, if there is diagnostic uncertainty then further investigations are completed (e.g., radiologically guided biopsy or aspiration) and the patient re-discussed with these results prior to surgery.

Fig. 2.

Our treatment pathway for management of shoulder PJI at The Royal National Orthopaedic Hospital Shoulder and Elbow Unit. After all revision cases, microbiology results and antimicrobial therapy are re-discussed in the MDT. *Patient carefully counselled of potential need to proceed to on table two-stage in presence of gross signs of infection. Updated and adapted from BESS PJI guidelines.³⁵.

All surgery is carried out via a deltopectoral approach with the patient in the beach chair position. At the 1st stage a thorough debridement is carried out, including removal of all infected soft tissue and bone, and excision of the sinus tract(s). An infection tray with five instrument sets is used to obtain 5 separate tissue samples from around the joint³⁸ and 1 fluid sample for microbiology. A separate tissue sample is sent to histology. Empirical antibiotic treatment is started with intravenous ceftriaxone, teicoplanin and amikacin. All prosthetic components and cement are removed using osteotomes, curettes, bone nibblers and a Midas Rex Legend High-Speed Burr (Medtronic, Minneapolis, MN, USA) and Ultradrive-3 Ultrasonic Revision System (Biomet, Warsaw, IN, USA) if required. Image intensifier is used to ensure complete cement removal. The removed implants are sent for sonification^33,62 and all tissue and fluid is cultured for a minimum of 14 days as reported in the literature.^9,12,14,45 Copious irrigation with 0.9 % saline is performed. A spacer moulded from ALBC containing targeted antibiotics (Heraeus Medical, Hanau, Germany) is inserted.^42,63 All wounds are closed over a suction drain (removed after 24 h).

Post-operative antimicrobial treatment is continued with ceftriaxone, teicoplanin 10 mg/kg OD IV and amikacin 15 mg/kg OD IV. Amikacin is discontinued after two doses, and treatment adjusted after microbiology results and sensitivities are available. Where sensitivities allow, patients are managed on oral antimicrobials after two weeks of intravenous therapy as per the OVIVA trial.⁴⁶ Patients are reviewed in the surgical clinic at two and six weeks post-operatively, with monitoring of inflammatory markers (WCC, ESR and CRP). In addition, patients are reviewed by the Out Patient Antimicrobial Therapy (OPAT) team, consisting of consultant microbiologists, clinical nurse specialists and pharmacists.³⁵ If there are no clinical signs of infection, and inflammatory markers return to normal, antibiotics are usually discontinued after six weeks.

The decision to proceed with a 2nd stage is dependent on the following criteria being met¹: no clinical signs of infection,² antibiotic treatment having been completed and³ normal inflammatory markers following cessation of antibiotics for at least two weeks and re-discussion in the MDT. Second stage was performed at a median of 6.7 months after first stage in a previous evaluation of our protocol.⁵⁴ If at the six-week post-operative review there are clinical signs of infection, or inflammatory markers have not normalised, a fluoroscopically-guided aspiration of the affected joint under aseptic conditions is undertaken. If the aspirate is positive, then patients undergo a repeat 1st stage procedure (debridement and pulse lavage).^35,54 If the aspirate is negative then a 2nd stage procedure is planned, but with a low threshold for making an intra-operative decision to repeat the 1st stage if there is concern regarding the possibility of ongoing infection owing to variable sensitivity of aspiration to detect infection.^20,31 At the 2nd stage soft tissue and bone samples are sent for microbiological analysis. If purulent fluid is identified during the 2nd stage an intra-operative decision to repeat the 1st stage is made. In the absence of concerns about ongoing infection, the antibiotic spacers are removed and the humerus and glenoid prepared for reimplantation. The choice of implant is dependent on the residual bone stock and the integrity of the rotator cuff tendons. Following the second stage, microbiology results are rediscussed in the MDT and antimicrobial therapy adjusted accordingly.

1.5. Reconstructive options and salvage procedures

Bone loss is a crucial consideration in PJI management, minor humeral bone loss <5 cm in the revision setting can be treated with cementoplasty³⁶ and contained glenoid bone loss can be managed with bone grafting from the iliac crest.⁶⁴ In cases of major humeral (>5 cm) or uncontained glenoid bone loss there are 2 main options: use of allograft-prosthetic reconstruction (APC), or massive endoprosthesis.^36,64,65 With either of these options, our preference is always for a two-stage procedure with second stage only occurring following confirmation of eradication.

In cases of failed two-stage arthroplasty with continued infection long-term antibiotic suppression for patients unfit for further surgery is sometimes recommended. In those fit for further major surgery, joint sacrificing surgery with excision arthoplasty is first line.⁶⁶ Very rarely, forequarter amputation or glenohumeral arthrodesis may be considered, the latter often limited by poor bone stock.⁶⁶

2. Conclusion

There is good evidence to support use of a thorough DAIR and removable component exchange in acute infections (<6 weeks) as first line treatment. There is an increasing body of evidence to suggest that under the right conditions a single stage revision is a safe option and results in low reinfection rates between 5 and 10 %.^22,51,52 These conditions may include: favourable host factors,³⁷ minimal bone loss and a low virulence organism such as P. acnes or S. epidermidis, mixed evidence exists as to whether CONS falls into this group.^22,50,67 Quality of debridement and thorough removal of all remaining cement are key, which are also key principles needed in two-stage revisions.^22,54

We advocate careful decision making for shoulder PJI, these decisions should be made and documented in MDT meetings. In ‘unlikely’ PJI cases (Table 1)¹² it is appropriate to consider one-stage revision performed in 2-phases. For most other instances, we prefer a two-stage approach at present, especially considering the increasing evidence for higher virulence organisms requiring multiple procedures.^22,50

There is a pressing need for further high-quality research in shoulder PJI, studies investigating bacterial genomic techniques for diagnosis and role of prognostic factors are needed. RCTs would need very careful design and focus on defined clinical phenotypes of PJI and must avoid treating all cases of shoulder PJI as a homogenous group.

Funding

No dedicated funding secured for this article.

Credit statement

Mr Thomas Stringfellow: conceptualization, investigation, data curation, writing - original draft, writing – review and editing. Mr Addie Majed: conceptualization, supervision, writing – review and editing. Miss Deborah Higgs: conceptualization, supervision, writing – original draft, writing – review and editing, supervision, project administration.

Conflicts of interest

No conflicts of interest to declare.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.