• Summary
  Click here to print a Quick Reference Guide
• Background
• Clinical Diagnosis
• Investigation
• Treatment
• Non-Antimicrobial Treatment
• Empirical Antimicrobial Treatment
• Directed Antimicrobial Treatment (when microbiology results are known)
• Duration of Treatment
• Switch to oral agent(s)
• Treatment Failure

Criteria for use:
Patients with suspected or confirmed septic arthritis of native joints with a short history of a hot, swollen, tender joint [or joints] with restriction of movement

Investigations required:

• Synovial fluid aspirate for Gram stain, polarising light microscopy and culture.
• Full blood count, C-reactive protein, Urea and electrolytes, liver function tests
• Serum urate
• Blood cultures in patients with systemic symptoms or signs of infection
• Plain radiography of affected joint(s)
• MRI if osteomyelitis is suspected

Treatment:
When initial microscopy of a joint aspirate is negative or when joint aspiration is contraindicated or in a patient with severe sepsis/septic shock [when joint aspiration is likely to be delayed], empirical antimicrobial therapy is indicated as below until results of culture are available:

• High risk for Gram negative or MRSA infection: Teicoplanin IV (see dosing guideline) plus iv Ceftazidime 1g 8-hourly.
• Penetrating joint trauma commence: iv Flucloxacillin 2g 6-hourly plus iv Ceftazidime 1g 8-hourly plus iv Metronidazole 500mg 8-hourly.
• Immunocompromised. Discuss with microbiology.
• Suspected gonococcal or meningococcal infection: iv Cefotaxime 1g 8-hourly iv.
• No risk factors as above [including intravenous drug users] iv Flucloxacillin 2g 6-hourly.

See full guideline for treatment of specific organisms

Back to top

“Septic arthritis” is a term commonly used to describe infection of a joint. Infection can occur in a native or natural joint as well as a joint containing, or replaced with, prosthetic material. This guideline relates to the management of an infected native joint. Inadequate or delayed treatment of septic arthritis is associated with irreversible joint damage and its associated morbidity. Mortality has remains high at about 10% especially in the elderly (Cooper & Cawley, 1986; Weston & Coakley, 2006; Weston et al., 1999). Septic arthritis may be difficult to diagnose and may present to a wide variety of healthcare professionals in both primary and secondary care, hence the need for a guideline to standardise the approach to diagnosis and treatment.
It is apparent that evidence-based recommendations [based on randomized controlled trials] are scarce in the field of septic arthritis management. The British Society of Rheumatology convened a multidisciplinary group of experts and produced guidelines based on available evidence (Coakley et al., 2006). These guidelines are based on the National guidelines but adapted to local systems and epidemiology.

If a patient has had a prosthetic implant at the site of suspected infection these guidelines do not apply and specialist advice is recommended.

If the patient has had gastroenteritis or symptoms of genital infection in the preceding six weeks consider a diagnosis of reactive arthritis.

Back to top

Pain is almost universally present in native joint infections. Local warmth, erythema, tenderness, joint effusions, systemic upset and fever all support a diagnosis of possible infection, but are not necessarily all present. Referred pain can also be present and may cause diagnostic difficulties, it should not be overlooked.

Patients with a short history of a hot, swollen, tender joint [or joints] with restriction of movement should be regarded as having septic arthritis until proven otherwise. Evidence level B.

In the case of preexisting joint disease [e.g. rheumatoid arthritis] the symptoms and signs in affected joints are out of proportion to previous background disease (Coakley et al., 2006; Kherani & Shojania, 2007).
Evidence level B.

Back to top

A synovial fluid aspirate for Gram stain, polarising light microscopy and culture should be sent from all cases of suspected native joint infection and all clinically affected joints. Warfarinisation is not a contraindication to aspiration (Coakley et al., 2006). Evidence level B.

The synovial fluid aspirate should not be routinely inoculated into blood culture bottles (Coakley et al., 2006). Evidence level C.

Synovial fluid should be inoculated into blood culture bottles if infection with Brucella spp. is suspected and such cases must be discussed with microbiology to ensure cultures are prolonged and because of the laboratory infection risk posed by these organisms (Sorlin et al., 2000; von Essen & Holtta, 1986; Yagupsky et al., 2001). Evidence level B

Inoculation of synovial fluid into blood culture bottles should be considered if antimicrobial therapy has been started prior to aspiration. False negative synovial fluid cultures are common [30-70%] due to prior administration of antimicrobials

A full blood count [FBC] is indicated at the time of diagnosis. A raised peripheral blood polymorphonuclear leucocye [neutrophil] count is insensitive for the diagnosis of native joint infection, but because of the likely need for aspiration and potential effects of various antimicrobials on bone marrow or blood cells a baseline measurement is needed. Evidence level B

C-reactive protein [CRP] should be measured at baseline in all cases. A raised CRP is a sensitive but non-specific result. A positive result gives support to a clinical diagnosis of native joint infection in the absence of other potential causes of a raised CRP. It may also be raised if infection is present at a site distant from the joint. Serial measurements are desirable to monitor the clinical course and the efficacy of the antibiotic treatment (Cooper & Cawley, 1986) Evidence level B.

Urea and electrolytes [U&E], liver function tests [LFT’s] are required at baseline to assess renal and hepatic function; the antimicrobials commonly used in this situation may either affect these organs or dosing may be affected by renal or hepatic dysfunction. Evidence level B.

Serum urate is of limited diagnostic value in septic arthritis or gout (Coakley et al., 2006) but a high level may point towards a diagnosis of gout and is recommended locally. Evidence level C.

Two sets of blood cultures, should be sent from patients with suspected septic arthritis [http://www.emedicine.com/med/TOPIC3394.HTM last accessed 17/03/2008] and a fever, any symptoms of systemic upset [e.g. chills, myalgia, general malaise] or with a systemic inflammatory response syndrome, more severe inflammatory response and from immuno-compromised patients. A 50% positivity rate has been reported in non-gonococcal septic arthritis compared with less than 20% in gonococcal arthritis (Goldenberg & Reed, 1985). Blood cultures may be positive when joint aspirates are culture negative (Weston et al., 1999). Evidence level B.

If there are symptoms or signs of other infections, such as urinary tract infection or wound infection, appropriate samples should be sent to microbiology. Evidence level C.

Plain radiography of the affected joint[s] is recommended as a baseline; it is of no diagnostic value but to assess future joint damage and because chondrocalcinosis suggestive of pyrophosphate arthropathy may be demonstrated (Coakley et al., 2006). Evidence level C.

Magnetic resonance imaging [MRI] is not recommended routinely in the investigation of a hot swollen joint but is the investigation of choice when osteomyelitis is suspected (Coakley et al., 2006). Evidence level B.

Back to top

A patient with suspected native joint infection should be referred urgently to an orthopaedic surgeon or rheumatologist (Coakley et al., 2006). Evidence level C.

Removal of infected material from the joint is key to management and infected joints should be surgically debrided and irrigated by an orthopaedic surgeon or aspirated to dryness by a competent person. Evidence level C.

Thorough surgical toilet and lavage is required for penetrating injuries involving joints

Back to top

Native joint infections are most commonly caused by staphylococci, streptococci [including beta haemolytic streptococci groups A, B, C and G, Streptococcus pneumoniae] and Gram-negative bacilli [including Haemophilus influenzae, Escherichia coli, Pseudomonas aeruginosa] (Ryan et al., 1997). Neisseria spp. Remain a less common but important cause (Ryan et al., 1997). Anaerobes are more common when there is penetrating joint trauma. MRSA is a recognised cause of native joint infection in Leeds [7% of culture positive septic arthritis cases over an 18 month period].

It is important not to start antimicrobials until a synovial fluid sample has been taken (Coakley et al., 2006). Evidence level C.

If resuscitation of a patient with severe sepsis or septic shock are likely to delay joint aspiration, do not delay starting empirical antimicrobial therapy but ensure that Blood cultures have been taken first. Evidence level C.

Empirical therapy
When initial microscopy of a joint aspirate is negative or when joint aspiration is contraindicated or in a patient with severe sepsis/septic shock [when joint aspiration is likely to be delayed], empirical antimicrobial therapy is indicated as below until results of culture are available:

• High risk for Gram negative or MRSA infection [there is significant overlap between the risk factors both these pathogens e.g. elderly >80, frail, recent hospitalisation, nursing home resident, recurrent UTI, previous MRSA colonisation or infection Teicoplanin IV (see dosing guideline) plus iv Ceftazidime 1g 8-hourly.
• Penetrating joint trauma commence: iv Flucloxacillin 2g 6-hourly plus iv Ceftazidime 1g 8-hourly plus iv Metronidazole 500mg 8-hourly.
• Immunocompromised. Discuss with microbiology on a case-by-case basis.
• Suspected gonococcal or meningococcal infection: iv Cefotaxime 1g 8-hourly iv.
• No risk factors as above [including intravenous drug users] iv Flucloxacillin 2g 6-hourly.

Negative joint fluid microscopy.
A negative Gram stain should not delay empirical antimicrobial therapy if the clinical suspicion of infection is high (Coakley et al., 2006). The Gram stain is less than 50% sensitive for detection of bacteria in synovial fluid (Faraj et al., 2002). Evidence level B

Positive joint fluid microscopy.
Gram positive cocci: treat according to empirical regimens but withhold Ceftazidime pending culture results [Leeds data from >1000 fluid samples examined indicate that Gram-negative infections were not identified in any synovial fluid with positive microscopy for a Gram-positive organism – thus the scenario of mixed infection must be considered very rare and does not justify the added risk of the broader-spectrum antimicrobial regimen].
Gram negative bacilli: treat according to empirical regimens.
Other Gram result: discuss with microbiology on a cases-by-case basis.

Back to top

When a joint aspirate is culture-positive, directed therapy can be commenced according to the culture results. In each case, susceptibility of the causative organisms will determine therapy. For the more common causative organisms specific recommendations can be made:

Antimicrobial therapy should be stopped promptly once a diagnosis of infection seems unlikely.

Staphylococcus species
meticillin susceptible:
Initial therapy: iv Flucloxacillin 2g 6-hourly.
Oral switch: oral Flucloxacillin 1g 6-hourly.

Penicillin allergic patient: use regimen for meticillin-resistant Staphylococcus.

meticillin resistant:
Initial therapy: intravenous Teicoplanin (see dosing guidelines) plus oral Rifampicin 300-600mg 12-hourly.

Oral switch: oral Clindamycin 450mg 6-hourly [If isolate is Clindamycin resistant or patient is >65, a care home resident or has other risk factors for Clostridium difficile infection discuss with microbiology].

Penicillin-susceptible Streptococcus species
Initial therapy: iv Benzyl penicillin 1.2g 4-hourly.
Oral switch: oral Amoxicillin 1g 8-hourly.

Penicillin allergic patient: Teicoplanin IV (see dosing guideline)  

Oral switch: oral Clindamycin 450mg 6-hourly [If isolate is Clindamycin resistant or patient is >65, a care home resident or has other risk factors for Clostridium difficile infection discuss with microbiology].

Other organisms should be discussed with Microbiology on a case-by case basis.

Back to top

4 weeks in total is recommended assuming PV and CRP are normal. Gonnococcal septic arthritis can be treated for 2 weeks (Ohl, 2005). Evidence level C.

Back to top

Evidence to guide appropriate timing from intravenous to oral antimicrobials is lacking. In practice normally a minimum of two weeks intravenous therapy is used but staphylococcal infection may require longer. Infection affecting joints affected by chronic joint pathologies [such as rheumatoid arthritis] may require more prolonged intravenous therapy (Goldenberg, 1989).

If regimens are not given above, cases should be discussed on a case-by case basis. Evidence level C.

Back to top

Back to top