Diagnosis and management of prosthetic joint infectionBMJ 2009; 338 doi: https://doi.org/10.1136/bmj.b1773 (Published 29 May 2009) Cite this as: BMJ 2009;338:b1773
- Philippa C Matthews, specialist registrar infectious diseases/microbiology, clinical research fellow123,
- Anthony R Berendt, consultant in infectious diseases12,
- Martin A McNally, consultant orthopaedic surgeon in limb reconstruction1,
- Ivor Byren, consultant in infectious diseases12
- 1Bone Infection Unit, Nuffield Orthopaedic Centre NHS Trust, Headington, Oxford OX3 7LD
- 2Department of Infectious Diseases, Oxford Radcliffe Hospitals NHS Trust, John Radcliffe Hospital, Headington, Oxford OX3 9DU
- 3University of Oxford, Peter Medawar Building for Pathogen Research, Oxford OX1 3SY
- Correspondence to: P C Matthews
Joint replacement is safe, cost effective,1 and widely undertaken. Most prosthetic joint replacements are hips and knees; more than 130 000 people underwent such procedures in England and Wales in the 12 months from April 2006.w1 Subsequent prosthetic joint infection is uncommon—the incidence varies between 0.6% and 2% per joint per year.2 3 4 5 However, this complication is associated with substantial morbidity and economic cost ($30 000 (£20 500; €22 800) to $50 000 per patient).3 4 6 w2 The diagnosis of prosthetic joint infection is difficult,w2 because symptoms, signs, and investigations may all be non-specific.7 w3 Defining diagnostic criteria and optimum management is complicated by patient heterogeneity and the small numbers in many published studies.w4 However, prompt recognition and diagnosis of prosthetic joint infection facilitates timely intervention to salvage infected joints, preserve joint function, prevent morbidity, and reduce costs.
Early diagnosis of prosthetic joint infection reduces morbidity and improves outcomes
Infection is eradicated and joint function is preserved in most patients who receive appropriate combined surgical and medical treatment
A multidisciplinary team is often needed for optimal diagnosis, treatment, and rehabilitation; specialist referral may be required.
Well fixed implants can be salvaged by aggressive debridement, antibiotics, and implant retention
Antibiotics are usually needed for three to six months if the prosthesis is retained, or for up to six weeks after revision arthroplasty
Sources and selection criteria
We performed Medline searches between November 2008 and January 2009 using the search terms “prosthetic joint” and “arthroplasty” combined with “infection”, “guidelines”, “septic arthritis”, “infection diagnosis”, “infection epidemiology”, and “infection revision arthroplasty”. Where possible, we focused on articles published in the past five years and restricted our search to literature published in English. We also drew from the experience of, and articles and documents published by, our multidisciplinary bone infection unit in the United Kingdom (www.noc.nhs.uk), which manages around 40 secondary and tertiary referral cases of prosthetic joint infection a year.
What are the clinical features of prosthetic joint infection?
Early prosthetic joint infection (box 1) characteristically presents with wound inflammation, joint effusion, loss of function, and pain,8 9 w5 w6 with or without wound dehiscence and discharge.8 9 Later disease is more likely to present with pain or mechanical dysfunction.w2 Systemic features, such as pyrexia, nausea, and malaise, are neither sensitive nor specific.w7 The onset of symptoms may be acute or insidious, with progressive pain or loss of function.7 9
Box 1 Definitions of prosthetic joint infection
Purulence around a prosthesis at arthrotomy or arthroscopy
Presence of one or more sinus tract communicating with the joint
Histological features of infection
Isolation of an indistinguishable organism from at least two deep culture samples (“indistinguishable” refers to widely performed laboratory characterisation of an organism; in most cases this will be identification of the genus and species, plus antibiotic susceptibilities13). Isolation of a virulent organism, such as Staphylococcus aureus, Escherichia coli, or Candida spp, in one deep tissue sample is regarded by some as sufficient to confirm the diagnosis.7
Who is at risk of prosthetic joint infection and when does it present?
Infection rates vary according to the joint replaced, indication for arthroplasty, comorbid conditions, and prophylactic strategies (box 2).5 15 Obesity and diabetes have been associated with early infection.9 15 w8 No increase in prosthetic joint infection has been reported in small cohorts of patients with kidney transplantsw9 or HIV infection,w10 w11 but as arthroplasty is increasingly performed the number at risk may increase.
Box 2 Common risk factors for prosthetic joint infection
Patients with the first three operative factors are at greatest risk of developing infection at the surgical site
In publications from the Mayo Clinic, 25% of wound infections were associated with joint infection,6 and the presence of any wound complication was significantly associated with deep infection.12 In a prospective case-control study, persistence of wound drainage for longer than five days after surgery and wound haematoma were associated with prosthetic joint infection (odds ratios 1.3 and 11.8).w12 These studies highlight the need for careful scrutiny of all wound complications overlying an arthroplasty.
The proportion of prosthetic joints that become infected within three months of arthroplasty varies widely—from 29% to 69% in cohorts from the United Kingdom.2 10 Early infection is usually acquired at the time of surgery or as a consequence of wound infection, whereas later infections arise because of haematogenous spread8 or because the pathogens are of low virulence and slow to cause symptoms.
Which organisms commonly cause prosthetic joint infection?
Gram positive organisms, especially staphylococci (commensal skin organisms), are most common (table 1)⇓.2 9 10 15 17 In early infection, pathogens are usually more virulent (for example, Staphylococcus aureus),18 whereas more indolent organisms predominate later on (for example, coagulase negative staphylococci, Propionibacterium acnes).10 In a recent UK study, however, most early cases were caused by coagulase negative staphylococci.8 In this cohort, most antibiotic resistant and polymicrobial infections presented within three months of arthroplasty.10 Infection with antibiotic resistant organisms poses an increasing challenge, with meticillin resistant S aureus (MRSA) being more common than meticillin sensitive S aureus (MSSA) in some series.9 15
Which laboratory investigations are needed?
Figure 1⇓ summarises key steps in the diagnostic process. UK guidelines recommend baseline blood tests for inflammatory markers (C reactive protein, erythrocyte sedimentation rate, leucocyte count) for any case of septic arthritis.w5 However, these parameters are raised for up to two weeks after orthopaedic surgery so are non-specific for early infection.8 Serial measurements can be useful,22 because persistently or progressively high inflammatory markers may help to distinguish joint infection from an uncomplicated postoperative course. Normal results do not exclude joint infection,7 14 particularly infection with indolent organisms.
Blood for culture should be taken before patients with suspected acute infection are started on antibiotics, although cultures are usually negative.3 9 Early empirical treatment reduces the diagnostic yield from cultures: more than half of culture negative cases from the Mayo Clinic had received previous antibiotics.21 The resulting failure to identify the causative pathogen(s) and the antimicrobial sensitivity profile may jeopardise outcomes. However, it may be impossible to take deep cultures before starting antibiotics in patients with systemic sepsis or rapidly evolving local infection.
Superficial swabs reflect colonising flora only, and results must therefore be interpreted with caution. Occasionally, swab results may help to inform empirical antibiotic choicesw13—for example, by identifying carriage of resistant organisms. However, deep samples of synovial fluid and tissue taken at arthrotomy, arthroscopy, or by aspiration are needed for definitive diagnosis. Such cultures were positive in 85% of 297 patients undergoing revision arthroplasty who had prosthetic joint infection.13 In this study, the identification of indistinguishable organisms from at least three culture samples was highly predictive of infection; to optimise diagnostic sensitivity and specificity, the authors therefore recommended that five or six samples are sent for culture.13
Histopathology can be a crucial adjunct to microbiology in the diagnosis of infection.7 13 This test has 94-98% sensitivity and adds significantly to the number of cases diagnosed by cultures alone.7 w14 Multicentre analysis suggests that a white cell count and neutrophil differential of the synovial fluid are also sensitive markers.23
Which imaging modalities are useful?
Plain radiography is often non-specific in early infection but may be useful to monitor serial changes. Loosening or bone loss around a previously well fixed implant is associated with chronic prosthetic joint infection (fig 2)⇓.3 7 8 w14 Ultrasound is useful to confirm effusion and to facilitate aseptic aspiration. One series reported abnormalities on 99Tc labelled scintigraphy in patients with early prosthetic infections,9 but the study did not discuss the low specificity of this test.8 w2 Computed tomography and magnetic resonance imaging may be useful in the evaluation of complex cases, but metal inserts interfere with these tests,w15 and abnormalities may be non-specific.14 w15
What management challenges does prosthetic joint infection pose?
Management of infection in arthroplasty poses the dual challenge of eradicating infection while preserving mechanical joint function. Biofilm is almost universally associated with infection of prosthetic material (table 2)⇓, particularly in chronic cases, and this underpins the need for thorough debridement and prolonged treatment with a combination of antibiotics.3 11 w16 w17
As with all iatrogenic complications after an elective intervention, a diagnosis of prosthetic joint infection can be difficult to accept, for both patients and clinical teams. Management often necessitates prolonged inpatient stays, repeated surgical procedures, and long term antibiotics, all of which may be associated with further morbidity, pain, and anxiety.w6 Although these aspects of prosthetic joint infection have not been formally studied, it is important to provide a multidisciplinary environment in which patients are supported and informed throughout their management.
What are the principles of management?
Long term outcomes are significantly better when management adheres to guidelines.10 24 Major considerations are whether the implant should be retained, what surgical strategy should be used, and which antimicrobial treatment to choose. The decision is influenced by disease duration, causative pathogen(s), extent of bone and soft tissue involvement, comorbid conditions, technical abilities of the surgical team, and the wishes of the patient (table 3)⇓.
What surgical options are available?
Initial surgical aims are to drain abscesses and remove dead tissue, to confirm the diagnosis with multiple tissue samples,13 and to achieve primary soft tissue closure over drains. Management options can thereafter be divided into three broad categories (table 3)5 8 w2:
Palliative strategies are needed in a minority of cases, usually in patients with severe comorbid disease. The prosthesis may be removed and not replaced, or long term antibiotic suppression may be attempted.9 w21 If all other strategies fail, arthrodesis or amputation may be undertaken.
Debridement, antibiotics, and implant retention
This approach incorporates thorough debridement of infected tissue around the joint, exchange of replaceable parts of the prosthesis, soft tissue closure, and treatment with antibiotics.18 19 20 w18 w21 In a recent study, open arthrotomy was superior to arthroscopic washout.w18 Certain pathogens, such as penicillin susceptible streptococci, are associated with good outcomes.w19 Infections with S aureus carry an increased risk of relapse,w18 but this management strategy can still be successful, even for infections with MRSA.20
This management approach can eradicate infection while preserving joint function, and it is a particularly attractive option in early infection.8 18 24 w19 An increasing body of evidence also points to its success in managing late infection,10 20 w18 and for frail or elderly people,w21 but patients must be carefully selected to avoid the risk of treatment failure.17
Removal of an infected prosthesis together with the affected bone, soft tissues, and all cement is a demanding procedure that needs considerable surgical expertise. Preoperatively, patients need counselling and optimisation of their general health.
Two stage revision is the most common approach to chronic prosthetic joint infection in many centres,7 17 21 w2 and it offers the option of local delivery of antibiotics via antibiotic eluting cement spacers.5 w20 However, this approach is expensive, time consuming, and may result in further tissue damage. In our experience, many patients also find it difficult to cope with restricted mobility before reimplantation.
Although one stage revision is less common,17 it is standard in some centres and can be successful.14 Outcome is probably related to the thoroughness of surgical resection, regardless of whether one or two stage reconstruction is used.w22 w23
Serious soft tissue defects can be caused by aggressive local infection or can be secondary to surgical debridement. Around the knee in particular, plastic surgery may be needed to provide soft tissue cover for exposed bone or prosthesis.w24
What is the best antimicrobial strategy?
The choice of antibiotics should be based on culture results; antibiotics should achieve high concentrations in the tissue and be active against slow growing organisms and biofilms.12 w4 w16 Liaison with microbiology services is advisable.17 w4 We present an overview of commonly selected agents (table 1), but the choice of agent, route of administration, and duration of treatment varies greatly.
Empirical antibiotics may be needed while culture results are awaited and for the duration of treatment for culture negative infection. Local policies based on the prevalence of resistant organisms guide the choice of agent. If prescribing parenteral treatment, a glycopeptide is often included to cover MRSA.3 10 22 Gram negative cover may be added by using an intravenous cephalosporin or carbapenem,10 especially if polymicrobial infection is likely.w25 An aminoglycoside is an alternative choice if concerns exist about Clostridium difficile diarrhoea.22
Prolonged intravenous therapy is usually delivered by a percutaneous indwelling central catheter or central venous access device.25 Selected patients can continue intravenous antibiotics in the community under the supervision of an outpatient parenteral antibiotic therapy programme.25 w26 A once daily antibiotic regimen using agents with a long half life, such as teicoplanin or ceftriaxone, is optimum.9 w27
Oral antibiotics can prolong safe affordable treatment. Rifampicin has excellent biofilm activity but selects rapidly for resistance if used as monotherapy. The most robust evidence comes from a randomised controlled trial of rifampicin plus a fluoroquinolone versus fluoroquinolone alone12; the success of this regimen is confirmed by other studies of staphylococcal infection.9 12 15 18 Rifampicin and fusidic acid is an alternative combination for treating fluoroquinolone resistant organisms.20 In a study of 40 episodes of prosthetic knee infection, any rifampicin containing regimen was more than 95% successful.24
Duration of antibiotic treatment
The duration of antibiotic treatment varies according to surgical management. If an infected prosthesis is retained, many clinicians favour up to six weeks’ parenteral treatment.9 17 If joint revision is undertaken some centres reduce the duration of intravenous antibiotics to two to four weeks.17 18 Robust data are lacking and clinical practice varies. Oral antibiotics are generally prescribed for a minimum of three to six months for a retained prosthesis and six weeks for revision arthroplasty, without prolonged parenteral treatment.11
In the minority of patients in whom surgery is precluded or declined, indefinite long term oral antibiotics (≥1 year) may be given, with the aim of suppressing but not curing the infection.17 w21 Despite these general recommendations, antibiotic prescribing varies considerably.3 9 17 18 20 22 w2 w28 An American study found no relation between the duration of parenteral treatment and the risk of relapse,22 and recent UK data suggest that oral treatment beyond six months does not increases the cure rate.w18
What preventive strategies exist?
The incidence of prosthetic joint infection has been progressively reduced by improvements in ultraclean air, preoperative preparation, surgical technique, theatre design, prophylactic antibiotics, wound care, and hand hygiene.w2 Ring fencing of beds for elective orthopaedic patients reduced the rate of MRSA infection by 70% in one UK centre.w29 Decolonising patients who are colonised with MRSA before elective surgery reduced the risk of implant infection in some centres,w30 w31 but local guidelines vary. Many orthopaedic surgeons advocate the use of antibiotic prophylaxis for patients with arthoplasties who are undergoing invasive dental procedures,w32 although this practice is not supported by clear evidence. Rigorous protocols (including surgical site infection care bundles)w33 are essential if infection rates are to be reduced.
What are the current developments in this field?
Diagnostic sensitivity has been improved by amplification of bacterial ribosomal DNA using the polymerase chain reaction,w3 although false positive results are a problem. Treatment options for antibiotic resistant pathogens are also expanding, but data on the safety and efficacy of new antibiotics are limited. Parenteral daptomycin and dalbavancin have extended spectrum activity against resistant Gram positive pathogens; tigecycline also provides Gram negative cover.w34 w35 Newer (unlicensed) oral options include linezolid, used with success in Gram positive bone infection,w36 w37 and moxifloxacin, which has broad spectrum and anaerobic activity, but there have been safety concerns with both these agents. In the longer term, colonisation of prostheses may be reduced by biofilm inhibiting furanones.w16
Further multicentre collaborations and randomised trials are needed to answer key questions about optimum diagnosis and management of these challenging infections. For now, attention to sound surgical and antimicrobial principles in the context of a coordinated multidisciplinary team offer the best treatment options.
Areas for future research
Further development and wider availability of molecular diagnostic techniques may help improve microbiological diagnosis in culture negative cases
Large multicentre randomised trials are needed to establish optimum management strategies (in particular, to identify who would benefit from a “debride and retain” approach); to evaluate antibiotic regimens; and to stratify the duration of antibiotic treatment. This will allow delivery of the most cost effective care
Additional educational resources
For healthcare professionals
UpToDate (www.uptodate.com)— This evidence based, peer reviewed resource for clinicians includes information on prevention, diagnosis, and management of prosthetic joint infection. Follow the link to infectious diseases then contents, then see the section on skin, soft tissue, and bone infection
Infectious Diseases Society of America (www.idsociety.org)—Guidelines on the diagnosis and management of prosthetic joint infection will be available online later this year
Coakley G, Mathews C, Field M, Jones A, Kingsley G, Walker D, et al. BSR & BHPR, BOA, RCGP and BSAC guidelines for management of the hot swollen joint in adults. Rheumatology (Oxford) 2006;45:1039-41
UK Guidelines are currently being drafted for the British Orthopaedic Association, British Infection Society, and Association of Medical Microbiologists (projected date of publication is late 2009)
Oxford Bone Infection Unit (www.noc.nhs.uk/ourservices/bone_infection.aspx)—Patient information leaflets on prosthetic joint infection, use of antibiotics in bone and joint infection, and central venous access devices for parenteral treatment at home
Uptodate for Patients (www.uptodate.com/patients)—Additional information can be found using the search term “prosthetic joint infection”
Tips for non-specialists
Patients with arthroplasty who present with wound discharge or erythema over the joint, or who report swelling, pain, or restriction of joint movement, should be promptly investigated for prosthetic joint infection8 11
Refer such patients immediately to an orthopaedic team. In hospital, take baseline bloods (for inflammatory markers and cultures)w5 and have the joint imaged and aspirated. Plain radiographs may be normal, but loosening of the prosthesis raises suspicion of infection
Collaborate with a microbiologist or infectious diseases doctor early in the course of management17
Avoid giving empirical antibiotics until samples have been taken for culture unless the patient is systemically unwell or has local signs of rapidly advancing infection21
If an attempt is made to retain the prosthesis, surgical debridement of infected soft tissues around the joint plus 3-6 months of antibiotics, using rifampicin based regimens for staphylococcal infection is recommended9 12 15 18 20
Cite this as: BMJ 2009;338:b1773
Contributors: PCM searched the literature and wrote the original article. ARB, MAMcN, and IB added references, contributed data for the algorithms and tables, and provided editorial input and advice. IB is the guarantor.
Competing interests: IB has received consultancy fees from Pfizer. ARB has received consultancy or speaker fees from Merck and Pfizer. PM is a Medical Research Council (UK) clinical research training fellow.
Provenance and peer review: Commissioned; externally peer reviewed.