Scoping review: Diagnosis and management of periprosthetic joint infection in elbow arthroplasty

Abstract

Background

Total elbow arthroplasty is an effective treatment for patients with painful elbow arthritis. Infection can be a serious complication. The aim of this scoping review was to document the available evidence on periprosthetic elbow infection.

Methods

A search of Medline, Embase and PubMed was performed; two authors screened results independently. Systematic reviews, randomised controlled trials, cohort studies, case–control studies and case series including periprosthetic elbow infection were eligible.

Results

A total of 46 studies were included. The median rate of periprosthetic elbow infection reported from recent published studies is 3.3%. The most commonly identified causative organisms are Staphylococcus aureus and Staphylococcus epidermidis. Risk factors include younger age, rheumatoid arthritis, obesity, previous surgery or infection to the elbow, and postoperative wound complications. Debridement, antibiotics and implant retention results in implant survival rates of 50–90%. Two-stage revision results in improved functional outcome scores, but with recurrent infection rates of 12–28%.

Conclusions

Total elbow arthroplasty carries a higher risk of infection when compared to other major joint replacements. The current body of literature is limited and is almost exclusively low volume retrospective case series. The best management of periprosthetic elbow infection is difficult to determine, but two-stage revision appears to be the gold standard.

Introduction

Total elbow arthroplasty (TEA) carries an increased risk of infection compared to other major joint replacements, with rates as high as 12% reported in the literature.^1,2 The reasons for this are unknown, but poor soft tissue cover has been postulated. No single definition for an infected TEA is used, although most authors employ a combination of clinical parameters (erythema, pain, swelling, sinus), haematological abnormalities (raised WCC, ESR and CRP), radiological findings (lucent lines around implants) and microbiology results (aspirate, biopsy, ultrasound sonication or intra-operative histology).

Infected elbow arthroplasty can be classified using the system of Yamaguchi et al.³ that aids when planning and determining treatment. This classification system includes: (i) infection with stable implants, (ii) infection with unstable implants and adequate bone stock and (iii) infection with poor bone stock that prevents reimplantation.

At present, there is no consensus regarding the best practice for the investigation and management of patients with periprosthetic elbow infection (PEI). The aim of this scoping review was to review the currently available literature regarding incidence, risk factors, diagnosis and management of PEI to inform development of evidence and consensus-based guidelines by the British Elbow and Shoulder Society.

Methods

Search strategy

A single search of Medline, Embase and PubMed was performed on 1 February 2018 (online Appendix 1). A search of unpublished literature or of trial databases was not carried out. No date restrictions were applied to the search, but studies were restricted to the English language and human studies. All age groups and genders were included.

Study selection

Studies were eligible for inclusion if they included data on incidence, risk factors, diagnosis or management of PEIs. Study designs eligible for inclusion were systematic reviews, randomised controlled trials, cohort studies, case–control studies and case series of elbow periprosthetic joint infections. Case reports, expert opinions, letters to the editor and articles relating to periprosthetic infections of joints other than the elbow and literature reviews with no indication of systematic process were excluded.

Studies identified from the literature search were downloaded and were de-duplicated and screened. Titles and abstracts identified the search were independently reviewed by two authors and those not meeting the inclusion criteria were excluded before full text review. The same two authors completed the full text search, which were evaluated against the inclusion and exclusion criteria. A search of the references of the selected studies was performed to ensure no other relevant studies were missed. Any disagreements on inclusion were resolved through discussion and if no agreement was reached the senior author arbitrated (online Appendix 2 and 3).

Quality assessment

As this is a scoping review the quality of papers was not formally evaluated for risk of bias using validated assessment tools.

Data extraction

Data extraction was performed by one author (ACW) and this was checked by a second author (ADD). Data extracted from each study included the year of publication, the study design, the number of participants and the results.

Data synthesis

A narrative synthesis was undertaken of the epidemiology, microbiological profile, risk factors, prophylactic measures against infection, investigations and management. The aim of the review was not to assess the effectiveness of diagnostic or treatment methods so pooling of data for meta-analysis was not performed.

Results

Study characteristics

The literature search of Embase, Medline and PubMed returned 891 unique studies. A total of 63 studies were selected at initial screening and following full text screening 46 studies were eligible for inclusion in the scoping review. This included two systematic reviews, two current concept reviews, 34 retrospective case series or cohort studies, two diagnostic studies, one clinical practice guideline and five large database epidemiological studies.

Prevalence of PEI (Table 1)

The search identified 29 studies reporting on the prevalence of PEI.^1–29 This included two systematic reviews, one current concepts review, 22 retrospective case series or cohort studies, and four large database studies. There is a clear discrepancy and inconsistency in the reporting of infection rates and infecting organisms, which is most likely due to the methods of analysis used, the case heterogeneity of available series as well as the varying diagnostic criteria used to define deep infection and how it is reported. This results in a wide range of reported deep infection rates from 0 to 11.7%.^1,2 The infection rates at the higher end of this range are when the authors report overall deep infection rates, with the lower end of the scale from series where deep infection is defined by infection requiring revision. Voloshin et al.¹⁵ performed a systematic review of 2938 cases of elbow arthroplasty published between 1993 and 2009 and calculated a rate of deep infection of 3.3 ± 2.9%. This rate was noted to have decreased from the 4.6% reported by Gschwend et al.²⁸ on a review of the literature published between 1986 and 1992.

Conclusions

Based on the available literature, the median rate of PEI from published studies ranges from 0 to 11.7%. The limited literature indicates the rate of deep infection is declining but remains higher when compared to other major joint replacement surgery. Reporting standards are required to accurately determine the overall rate of PEI and the rate requiring revision.

Microbiological profile (Table 1)

Our search identified 13 studies reporting on the microbiological profile of PEI. This included one current concepts review and 12 retrospective case series or cohort studies.^{3–5,8,9,11,16,27,29–33} Where reported, the organism most commonly isolated was Staphylococcus aureus (>50%) and Staphylococcus epidermidis (approximately 25%).^3,4,16,31 Other reported organisms include corynebacterium, Pseudomonas aeruginosa, propionibacterium acnes, beta-haemolytic streptococcus, Enterobacter cloacae and Klebsiella.²⁹

Conclusions

The most common organisms identified following deep PEI are S. aureus and S. epidermidis.

Risk factors for PEI

The search identified seven studies reporting on the microbiological profile of PEI. This included one systematic review, four retrospective case series or cohort studies and two large database studies.^{4,5,23,24,26,34,35} Documented risk factors for PEI are younger age at the time of surgery, duration of surgery and previous surgeries to the same joint.^4,5,24 Other contributory factors have been noted in case series.

Jeon et al.³⁴ reported a large series of 1749 TEA from the Mayo Clinic and documented wound problems in 5.5% (n = 97). The definition of wound problems in this series was delayed healing or continuing drainage of the wound including secondary to haematoma. Of these cases, 9% developed a deep prosthetic infection, with the authors noting an increased incidence in patients with rheumatoid arthritis.

Obesity was considered by Griffin et al.²³ in large database review of 7580 patients who underwent TEA. They found that patients with a BMI of less than 30 had an infection rate which was 2.1%, for those with a BMI of 30–40 it was 4.7% and for those with a BMI 40 or above it was 7%. This study also reported a higher rate of overall medical complications, including venous thromboembolism in obese patients.

Pope et al.³⁵ utilised the US Nationwide Inpatient Sample database to identify 13,698 patients that underwent TEA, of which 2270 (16.5%) were diabetic, with the aim to determine the rate of perioperative complications associated with diabetes. The authors reported that diabetic patients had significantly increased length of stay, higher blood transfusion rates and an overall increase in the number of perioperative complications. Despite the limitations with large databases studies including the length of follow-up and the accuracy of recording infection, wound infection rates were found to be twice as common in diabetic patients (1.6% versus 0.83%).

Prkic et al.²⁶ from the Netherlands reported a revision rate for primary TEA of 13.5% in a systematic review of 1253 cases. Of the patients undergoing revision, 19% had a deep infection. The authors noted a higher revision rate in patients with rheumatoid arthritis rather than post-traumatic osteoarthritis and trauma, but with infection rates not statistically different.

Conclusions

Risk factors for PEI include a younger age at the time of TEA, rheumatoid arthritis, obesity, a history of multiple previous surgeries prior to the index TEA, length of surgery for TEA, previous infection to the elbow and postoperative wound complications. The current literature in this area is limited due to small samples sizes and the limitations associated with current large database studies.

Prophylaxis for PEI

There were no studies identified specifically investigating the role of prophylaxis and PEI.

Diagnosis

Our search identified 13 studies reporting on the diagnosis of PEI.^{3–5,8,29,31,36-42} This included two current concepts reviews, eight retrospective case series or cohort studies, two diagnostic studies and one clinical practice guideline.

Clinical assessment

The clinical presentation of PEI is variable and systemic symptoms are not commonly seen.^4,5 The time of onset from the index procedure is important and can be used to classify the type of infection into acute (less than three months following surgery), subacute (3–12 months following surgery) and late (more than 12 months following surgery).^3,5

Serological markers

Serum markers (raised white cell count, ESR and CRP) are not reliable for diagnosis of infected elbow arthroplasty when used in isolation.^4,5 Somerson et al.²⁹ in a current concept review reported that the usual markers of raised white cell count, CRP and ESR were not particularly specific for infection in TEA. This is further complicated by the high numbers of patient undergoing TEA that have a background of inflammatory arthritis (Table 1), where elevated CRP and ESR are commonly seen.^4,41

Table 1.

Characteristics of included studies for the prevalence of PEI.

Authors	Cases (n)	Indication for TEA	Mean length of follow-up (years)	Infection rate (%)	Most common organism
Morrey and Bryan4	156	Rheumatoid arthritis (n = 112) Post-traumatic (n = 40) Other (n = 4)	Not reported	9.0	S. aureus (57%)
Wolfe et al.5	184	Rheumatoid arthritis (n = 148) Post-traumatic (n = 36)	Overall not available	7.0	S. aureus (82%)
Ewald et al.6	202	Rheumatoid arthritis (n = 170) Post-traumatic (n = 2) Bilateral (n = 30)	5.8	1.5	Not reported
Kasten and Skinner1	34	Rheumatoid arthritis (n = 22) Post-traumatic (n = 10) Other (n = 2)	7.6	11.7	Not reported
Yamaguchi et al.3	757	Primary (n = 591) Revision (n = 166)	5.9 (infection group)	3.3	S. aureus (56%)
Hildebrand et al.7	39	Inflammatory arthritis (n = 21) Post-traumatic (n = 8) Fracture (n = 7) Other (n = 3)	4.2	7.7	Not reported
Gille et al.8	305	Rheumatoid arthritis (n = 235) Post-traumatic (n = 64) Primary OA (n = 6)	6.8 (infection group)	1.9	S. aureus (100%)
Aldridge et al.2	41	Inflammatory arthritis (n = 12) Post-traumatic (n = 24) Primary OA (n = 4) Other (n = 1)	Minimum 10	0	Not reported
Schneeberger et al.9	23	Post-traumatic (n = 15) Revision (n = 4) Inflammatory arthritis (n = 3) Fracture (n = 1)	4	4.3	P. aeruginosa (100%)
Shi et al.10	37	Rheumatoid arthritis (n = 16) Post-traumatic (n = 10) Fracture (n = 9) Other (n = 2)	7.2	8.1	Not reported
Tachihara et al.11	72	Rheumatoid arthritis (all)	3.5	4.8	S. aureus (33%) P. aeruginosa (33%) E. cloacae (33%)
Prasad and Dent12	99	Rheumatoid arthritis (all)	5 and 9	2.0	Not reported
Qureshi et al.13	22	Rheumatoid arthritis (all)	11.9	9.1	Not reported
Krenek et al.14	1625	Rheumatoid arthritis (n = 364) Osteoarthritis (n = 189) Unknown (n = 1072)	4	5.4	Not reported
Voloshin et al.15	2938	Not reported	Not reported	3.3	Not reported
Spormann et al.16	262	Not reported	5 (infection group)	7.6	S. aureus (57%)
Gay et al.17	1155	Post-traumatic (n = 719) Rheumatoid arthritis (n = 283) Primary OA (n = 76) Oncology or other (n = 1)	Not reported	3.3	Not reported
Jenkins et al.18	1146	Inflammatory arthritis (n = 908) OA (n = 108) Trauma (n = 74) Unknown (n = 56)	Not reported	8.2	Not reported
Park et al.19	84	Rheumatoid arthritis (n = 55) Post-traumatic (n = 24) Primary OA (n = 5)	13	1.2	Not reported
Plaschke et al.20	324	Rheumatoid arthritis (n = 237) Post-traumatic (n = 61) Primary OA (n = 18) Other (n = 8)	8.7	1.5	Not reported
Hastings et al.21	46	Rheumatoid arthritis (n = 23) Post-traumatic (n = 6) Fracture (n = 5) Primary OA (n = 5) Revision (n = 2) Other (n = 5)	4	2.2	Not reported
Alizadehkhaiyat et al.22	100	Inflammatory arthritis (n = 58) Primary OA (n = 17) Post-traumatic (n = 14) Fracture (n = 7) Other (n = 4)	4	2.0	Not reported
Griffin et al.23	7580	Not reported	0.25	2.9	Not reported
Toulemonde et al.24		Rheumatoid arthritis (n = 45) Fracture/Nonunion (n = 28) Revision (n = 16) Post-traumatic (n = 5) Other (n = 6)	5	4	Not reported
Sanchez-Sotelo et al.25	435	Rheumatoid arthritis (all)	10 (median)	7.8	Not reported
Prkic et al.26	9308	Not reported	6.8	2.6	Not reported
Minami et al.27	421	Rheumatoid arthritis (all)	12.3	1.9	S. aureus (63%)

Source: Table adapted from Somerson et al.²⁹

OA: osteoarthrosis; TEA: total elbow arthroplasty.

Imaging

Serial plain radiographs of the elbow can show signs of progressive lucency at the cement–bone interface, although this can be present in the absence of chronic deep infection.^4,29,36 Some authors have suggested that early progressive lucency and endosteal scalloping are characteristic of PEI.^5,29,42 Gille et al.⁸ reported that four of the six elbows in their series that underwent single-stage revision for infection had evidence of radiolucency on imaging. Further imaging in the form of bone scanning, leucocyte scans, CT, MR or PET is not currently recommended by the Infectious Diseases Society of America guidelines.^29,39

Joint aspiration

As with serological tests, while a positive aspiration result may identify the organism and aid preoperative planning, a negative result does not exclude PEI. Achermann et al.³⁸ found joint aspiration to be an unreliable tool to exclude infection, with only 22% of 27 infected elbow arthroplasties having a positive aspirate culture in their study. However, the number of cases with prior antibiotic treatment prior to aspiration was not clear. Gille et al.⁸ reported their findings of single-stage revision for deep infection in six (1.9%) of 305 primary TEAs and found five out of six aspiration cultures were positive with the sixth elbow having had previous antibiotic therapy.

Intra-operative histology

Ahmadi et al.⁴⁰ undertook intraoperative histology and microbiological samples from 227 revision TEAs and reported a high negative predictive value, although whether this had any advantage over subsequent cultures was unclear. The study reported diagnostic performance characteristics of sensitivity 51.3%, specificity 93.1%, positive predictive value 60.6% and negative predictive value 90.2%.

Open biopsy

Tissue biopsy, either open or arthroscopic, can be performed at the time of revision or as a separate pre-revision investigation and is considered the gold standard investigation.³⁹ Atkins et al.³⁷ indicated that overreliance on biopsy cultures may lead to overtreatment and recommended that at least five specimens are obtained and that at least two should identify indistinguishable organisms. Wee et al.⁴¹ reported on the outcome of TEAs with unexpected positive intraoperative cultures. They found little or no correlation, with 7.5% (16/213 revisions) of cases unexpectedly having positive cultures of S. epidermidis or Propionibacterium acnes. Of these 16 cases, only two cases had culture-proven infection and of the 12 cases with a minimum of two years follow-up only one patient was a confirmed deep infection in association with early implant loosening. It should be noted that in this study only three patients had three intra-operative specimens sent, 10 had two and three had one.

Biofilm sonification

Biofilm sonification was looked at by Vergidis et al. in their analysis of 27 patients with aseptic TEA failure and nine patients with PEI. All patients had ≥2 tissue cultures and one implant sonicate culture taken.³¹ The authors reported a sensitivity of 89% and specificity of 100% for the technique, which although were superior to the tissue culture, this was not significant. The additional value of this is not clear.

PCR/IL-6/alpha-defensin

There were no studies identified specifically reporting on these investigations in PEI. There are studies that have combined cases of periprosthetic joint infection, with the number of PEIs included too small to make any meaningful conclusions.

Conclusions

The time of onset following index TEA is important. Blood tests including white cell count and inflammatory markers (CRP and ESR) are sensitive but not specific for PEI. Imaging has a very limited role in the diagnosis of PEI, although early progressive radiological changes can be indicative for infection. Joint aspiration is not sensitive or specific for the diagnosis of PEI with a low negative predictive value apparent. In the presence of clear clinical infection, aspiration should not delay definitive surgery. Intra-operative histology is specific but has limited sensitivity and is best to confirm suspected PEI. Open biopsy has a documented false positive rate, with positive culture not comprehensively indicating deep infection. In the absence of strong evidence, a minimum of five specimens should be obtained and to confirm deep infection at least two should identify indistinguishable organisms. Biofilm sonification does not appear to improve our ability to diagnose PEI and literature is needed analysing the role of more advanced techniques to diagnose PEI.

Surgical management of PEI

Our search identified 12 studies reporting on the surgical management of PEI.^{3,8,16,29,30,32,33,38,43–46} This included one current concepts review and 11 retrospective case series or cohort studies.

Debridement, antibiotics and implant retention (DAIR)

Achermann et al.³⁸ identified four criteria that should be met to permit DAIR:

1. Short duration of infection (less than three months)

2. Short duration of clinical signs (up to 21 days)

3. No severe soft tissue damage

4. Availability of antimicrobial agents active against biofilms.

If these above criteria were not met either a single- or two-stage revision were undertaken. In this series, 21 of 27 patients (78%) were treated with DAIR and a prolonged three-month course of antibiotics including rifampicin for staphylococcal infection. For those having two-stage revision, antibiotics were continued for 6–8 weeks and a two-week antibiotic-free period was required before reimplantation. The relapse-free survival was 79% after one year and 65% after two years, but if the correct treatment algorithm was followed throughout, infection-free survival was 100%.

Spormann et al.¹⁶ from the same clinic reported DAIR in 20 patients who developed infection out of 262 primary total elbow arthroplasties. They divided patients into three groups according to the timing of presentation: early infection within three months of implantation (nine cases), delayed infection between three months and two years (one case) and late infection more than two years after implantation (10 cases). If there were no radiographic signs of loosening, the components were disarticulated to allow thorough debridement and washing with 6 l of irrigation. The bushings were then exchanged but the components retained. If the components were loose, a single-stage revision was performed. Intravenous antibiotics were administered, and further debridement undertaken if the inflammatory markers (CRP or WCC) did not decrease following one week. DAIR was performed in eight of nine early infections with no recurrences, and in all 10 late infections, with six cases having recurrence of infection. Three of these went on to a two-stage revision and three elected to have antibiotic suppression.

Streubel et al.³³ reported on a regimen of implant disengagement and debridement followed by a later debridement and re-linkage. The medical records of 20 patients were retrospectively reviewed, with the mean time from the onset of symptoms to debridement of 59 days (range 0–565) and 62% were operated on within 30 days. In 90% of cases the implant was retained but further surgical debridement was required for recurrent infection in four (20%) cases. There was no relationship found between the duration of infection and the likelihood of recurrence.

In patients with PEI associated with S. epidermidis, some authors recommend against the use of DAIR if the clinical scenario allows due to potential problems associated with biofilm formation and antibiotic resistance.^{3,16,29,47,48} There has been a single case report of successful management of PEI with a single arthroscopic washout followed by three months of intravenous antibiotics, with no evidence of infection at 10 months following the procedure.⁴⁹

Single-stage revision

Single-stage revision has been reported in six infected elbows out of 305 primary TEA with successful reimplantation as a single-stage procedure in five out of six cases.⁸ All six cases of PEI involved on a background of rheumatoid arthritis and steroid therapy. The outcome was graded as good in three cases, fair in two (one resection arthroplasty) and poor in one.

Two-stage revision

Peach et al.³² have reported on the outcome of two-stage revision in a retrospective series of 34 patients with proven infection, of which 26 (76%) went onto have second-stage reimplantation. Their protocol involved meticulous debridement of the soft tissues, removal of the implants and all the cement using osteotomes, power burrs, curettes, a nylon brush and an ultrasound device if needed. Windowing of the bones was used for well-fixed implants and the canals were irrigated with chlorhexidine solution diluted with 0.9% sodium chloride. Antibiotic beads were inserted. Further surgical debridement was undertaken if signs of infection persisted, as was the case in 15% of elbows. Recurrent infection occurred in three of 26 elbows (11.5%) that went on to have a second-stage surgery, with the mean Mayo Elbow Score in those without recurrent infection 81 at a minimum of two years following surgery. Similar positive results have been reported from the Mayo clinic in four out of five elbows from a subgroup of patients presented by Yamaguchi et al.³ Cheung et al.³⁰ subsequently reported on the outcome of 29 patients from the same centre that underwent two-stage revision with a mean time between stages 72.5 weeks. Following resection patients received six weeks of antibiotics, with infection eradication confirmed using a combination of serological markers, joint aspiration and tissue samples. At a mean of 7.4 years following reimplantation the mean Mayo Elbow Score significantly improved from 36 to 66, although the infection was not eradicated in eight elbows (28%) that proceeded to resection. Of the 21 patients with infection eradication the outcome was graded as good to excellent in 15, fair in three and poor in three.

Resection arthroplasty

This option is often considered if patients with either catastrophic bone loss, patient choice or where there is little or no chance of infection eradication.²⁹ Zarkadas et al.⁴⁴ reported the outcome in 51 elbows that underwent resection for a deep infection of a TEA, of which 30 elbows were reviewed at a mean of 11 years post-surgery. The mean DASH score was 71 and the mean Mayo Elbow Score improve significantly from 37 to 60 points post-resection, with the increase in the pain component of the score noted. Eight cases were graded as good, 11 fair and 11 poor. The authors noted that post-resection stability correlated with a superior long-term Mayo Elbow Score.

In another study, Rhee et al.⁴⁶ reported the outcome of resection arthroplasty in 10 elbows from nine patients with an average follow-up of 52 months. Surgery was performed with the aim of removing all infected tissue, the implant and cement, while preserving the distal humeral columns to achieve relative stability after resection. The elbow was placed in a splint at 90° flexion for six weeks. Intravenous antibiotics were given for an average of 22 days, with oral for a further 25 days on average. The mean DASH score was 53 and according to the Mayo Elbow Score the final outcome was good in six elbows, fair in three and poor in one. One elbow required additional surgery to treat persistent infection.

Fusion

The results of fusion as salvage for infected elbow arthroplasty that had undergone an initial debridement and placement of antibiotic space have been reported by Otto et al.⁴⁵ in four patients, all of whom were smokers. None of the patients had a confirmed fusion at follow-up ranging from 12 to 73 months, with two patients developing an asymptomatic fibrous non-union and two patients subsequently proceeding with resection arthroplasty. Complications noted by other authors following elbow arthrodesis include fractures of the humerus.^29,43

Antibiotic regime for PEI

Peach et al.³² in their series of two-stage revisions used local antibiotics after first-stage debridement with cement sphere spacers mixed with 1 g of vancomycin and 1 g gentamicin unless microbiology indicated otherwise. Three doses of a broad-spectrum antibiotic were given, the first after specimens had been harvested for microbiology. No prolonged antibiotic therapy was given systemically.

Achermann et al.³⁸ utilised two weeks of intravenous antibiotic therapy followed by three months of oral antibiotics following DAIR, with concomitant rifampicin used for staphylococcal infections. Streubel et al.³³ reported using a regime of antibiotic-impregnated cement (tobramycin or vancomycin) in a two-stage procedure with retained implants. Broad-spectrum intravenous antibiotics were given until microbiology culture results were available. Intravenous antibiotics continued for at least four weeks with oral lifelong antibiotic prophylaxis administered after re-linkage.

Conclusions

DAIR for PEI has good implant survival rates in the absence of loosening but requires good soft tissue cover and targeted antibiotic therapy. Single-stage revision is not adequately reported in the literature to draw meaningful conclusions. Two-stage revision appears to provide improvement in functional outcome scores but with recurrent infection rates ranging from 12 to 28%. Resection arthroplasty reports recurrent infection rates of 10%. Elbow joint fusion has 0% union rate and 50% conversion to resection arthroplasty in the one case series available.

Conclusions

TEA carries an increased risk of infection compared to other major joint replacements. The diagnosis of PEI is based on a combination of clinical history and diagnostic tests. However, there is no gold standard for diagnosing infection due to the poor diagnostic performance characteristics reported in the literature. Positive cultures from biopsy do not always indicate deep infection and the methods by which these are collected are very important. Future diagnostic tests such as PCR, implant sonication or cytokine levels have the potential to be useful diagnostic adjuncts, although the current literature in this area is very limited.

There are a range of surgical regimes reported for the management of PEI, which may be based on the Yamaguchi stage and on the infecting organism if known. The current body of literature is very limited and is almost exclusively based on small retrospective case series, thus the optimal management of PEI is difficult to determine. Large prospective cohort database studies and multi-centre randomised trials are clearly needed. However, based on the current literature the accepted gold standard treatment for PEI appears to be two-stage revision, particularly when the implants are loose. DAIR can be used for early cases of PEI where it presents shortly after implantation (less than three months) with a short duration of clinical signs (less than 21 days) and where there is good soft tissue coverage and targeted antibiotic therapy against biofilms is achievable.

Supplementary Material

Supplementary material is available at: journals.sagepub.com/doi/suppl/10.1177/1758573218789341.

Declaration of Conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Wrightington Upper Limb Unit received funding from Wright Medical, Lima and Zimmer Biomet for research.

Ethical Review and Patient Consent

Ethics approval was not required for this literature scoping review. Patient consent was not required.