Skip to content

Septic arthritis

Septic arthritis of native and prosthetic joints: synovial fluid diagnosis, surgical drainage, and antibiotic treatment.

  • 16 min read
  • Last verified
  • Checked against sources

In one minute

  • Septic arthritis is infection of a joint space. It is an orthopaedic emergency: cartilage can be damaged within days, and delay means permanent loss of joint function or death.
  • Most adult disease is native joint septic arthritis (NJSA): sudden monoarthritis of one large joint, most often the knee, caused by bacteria that reach the joint through the blood.
  • Staphylococcus aureus is the commonest organism, responsible for roughly 58% of isolates in large cohorts. About 83% of isolates are Gram-positive, and roughly a quarter of episodes are culture-negative. [1]
  • There are no perfect tests. Synovial fluid examination supports the diagnosis but cannot confirm it on its own; the presence of a pathogen, or compatible histology or operative findings, is required.
  • Treatment is drainage plus antibiotics: drain the joint promptly, cover S. aureus while awaiting cultures, then switch to a pathogen-directed regimen.
  • Prosthetic joint infection (PJI) is a related but distinct problem: it needs the EBJIS three-level diagnostic approach, tissue sampling, and usually DAIR or staged revision surgery.

Learning objectives

By the end of this page you should be able to:

  1. Explain how bacteria reach a joint and why cartilage is destroyed so quickly.
  2. Take an appropriate history and examination for an acutely hot, swollen joint.
  3. Plan synovial fluid analysis and interpret its limits in native joint septic arthritis.
  4. Empirically manage a suspected native joint infection, including when to operate and what to cover.
  5. Apply the EBJIS approach and sample a prosthetic joint correctly.
  6. Recognise high-risk patients and low-resource adaptations.

Definition

DefinitionSeptic arthritis

Septic arthritis is inflammation of a joint space caused by infection, most often bacterial, but occasionally fungal, mycobacterial, viral or another pathogen. Native joint septic arthritis (NJSA) affects a joint with no prosthesis. Prosthetic joint infection (PJI) is infection of a periprosthetic joint space and is considered separately because diagnosis, sampling and surgery differ.

NJSA is the main focus of this page, with PJI addressed in its own section. The two overlap in microbiology but diverge in management.

Epidemiology

Population data describe NJSA as uncommon: about 2 to 15 cases per 100,000 person-years.[1] The incidence rises sharply in people aged 65 or older, and in those with comorbidity or adverse social circumstances.[1] In some regions, injection drug use linked to the opioid epidemic has driven rising rates and changed demographics and microbiology.[1]

The affected joint is usually a single large joint: the knee most often, then the hip or shoulder.[1] About one in five patients with NJSA has no identifiable risk factor, so a high index of suspicion is needed in any adult with an acute swollen joint.[1]

The healthcare burden is substantial: between 2009 and 2012 in the United States there was a 26% increase in total charges and a 24% increase in inpatient charges for septic arthritis, with no change in length of stay or outcome trends.[4]

Aetiology and risk factors

Three routes bring organisms into a joint.[1]

  • Hematogenous seeding. The commonest route. Organisms arrive during bacteraemia. Risk is highest with damaged synovium (osteoarthritis, rheumatoid arthritis, crystal arthritis, prior trauma or surgery) and with conditions that cause bacteraemia.
  • Direct inoculation. After trauma, arthroscopy, intra-articular injection or aspiration, or other joint-invasive procedures.
  • Contiguous spread. From adjacent skin or soft-tissue infection, bursitis or deeper musculoskeletal infection, particularly where skin integrity is poor.

A review groups risk factors by which joint is likely to be involved and which organisms to expect.[1]

Risk factors for native joint septic arthritis and their microbiological clues
Risk factor or exposureTypical mechanismMicrobiological associationsTypical joint pattern
Structural joint damage (osteoarthritis, rheumatoid arthritis, crystalline arthritis, prior trauma or surgery)HematogenousS. aureus; beta-haemolytic streptococciLarge joints, especially knee
Injection drug useHematogenous, sometimes embolicS. aureus (MSSA/MRSA), streptococci, water-borne Gram-negative organisms such as Pseudomonas aeruginosa and Serratia speciesGirdle joints (sternoclavicular, acromioclavicular, sacroiliac), axial facet joints; often oligo- or polyarticular
HaemodialysisHematogenousS. aureus (including MRSA), coagulase-negative staphylococci, Gram-negative organismsLarge joints; often oligo- or polyarticular
Unpasteurised dairy productsHematogenousBrucella speciesSacroiliac joint, often chronic
Sexually active adults, suspected gonococcal diseaseHematogenousNeisseria gonorrhoeaeMigratory arthritis, flexor tenosynovitis, arthritis-dermatitis syndrome, small or large joints
Ixodes tick exposure in endemic regionsHematogenousBorrelia burgdorferiKnee monoarthritis more than other large joints
Diabetes mellitusHematogenous plus contiguousS. aureus, streptococci, Gram-negative organisms including Pseudomonas aeruginosa, polymicrobial infectionLarge and small joints; often oligo- or polyarticular
Immunocompromise (chronic kidney disease, cirrhosis, HIV, malignancy, glucocorticoids, biologics)Hematogenous plus contiguousWider Gram-negative spectrum, atypical and opportunistic pathogensLarge joints; often oligo- or polyarticular
Recent orthopaedic surgery or intra-articular procedureDirect inoculationStaphylococci (including coagulase-negative), streptococci, occasional polymicrobial infectionThe joint that was accessed
Traumatic arthrotomy, human or animal bitesDirect inoculation plus contiguousPolymicrobial: Pasteurella, Capnocytophaga, Eikenella, staphylococci, streptococci, anaerobesHand (metacarpophalangeal and interphalangeal) and the injured joint
Skin barrier disruption (psoriasis, eczema, ulcers)ContiguousPolymicrobial, staphylococci, streptococci, anaerobesJoint adjacent to the skin disease

Rheumatoid arthritis deserves special mention because chronic joint damage and immunosuppressive therapy confer a particularly high risk, and a flare can mimic infection.[1]

Pathophysiology

Normal synovial fluid is a small volume of clear, viscous plasma filtrate produced by the synovial membrane; it lubricates and nourishes cartilage. In infection, the joint becomes a closed space with a rich blood supply, rapid bacterial multiplication and an intense neutrophil response. Proteolytic enzymes and reactive oxygen species degrade cartilage and synovium, and accumulating exudate raises intra-articular pressure until perfusion falls. For this reason, damage becomes irreversible within about three days if the joint is not drained.[1]

Rheumatoid-arthritis-like synovitis and crystal arthritis may look similar early on, which is why fluid examination and cultures matter.

Clinical features

The classic picture is acute onset over hours to a day of a single hot, swollen, very painful joint, with pain on any movement. The patient often holds the joint still in the position that maximises joint volume.

  • Pain is the dominant symptom and is severe. The joint is exquisitely tender.
  • Swelling and warmth are marked and usually confirmable by inspection and palpation.
  • Fever and systemic upset are common but may be absent, particularly in older or immunocompromised patients.
  • Loss of movement; severe pain on passive motion is the hallmark.
  • Polyarticular disease can occur and is more likely with bacteraemia, immunocompromise, or injection drug use, especially in girdle and axial joints.

Common adult presentations include a middle-aged person with osteoarthritis and diabetes who develops acute knee pain overnight, or an injection-drug user with sternoclavicular pain and fever.

Examination should document effusion and synovial tenderness, active and passive range, fever, nearby skin or bursal infection, and whether aspiration is obtainable.

Differential diagnosis

The main alternatives for an acutely swollen joint are crystal arthritis (gout or pseudogout), rheumatoid arthritis flare, reactive arthritis, osteoarthritis with haemarthrosis, trauma, avascular necrosis, tumour, and Lyme or gonococcal arthritis. The distinguishing table below is a teaching scaffold; the diagnosis turns on aspiration and culture.

Conditions that can resemble septic arthritis
ConditionClues that separate it from septic arthritis
GoutVery painful single joint, often first metatarsophalangeal; negatively birefringent needle-shaped crystals; urgent Gram stain and culture still needed because the two can coexist
Pseudogout (calcium pyrophosphate)Positively birefringent rhomboid crystals; more common in older adults and wrists and knees
Rheumatoid arthritis flarePolyarticular synovitis, known RA, elevated inflammatory markers but sterile aspiration
Reactive arthritisRecent infection, sterile joint, asymmetric oligoarthritis, often with enthesitis or mucocutaneous findings
HaemarthrosisTrauma or bleeding tendency, bloody fluid, no organisms
Gonococcal arthritisYounger, sexually active adult, migratory arthritis and tenosynovitis, positive NAAT from exposed sites

Investigations

Imaging

Plain radiographs exclude fracture and show pre-existing arthritis but are usually normal early. Ultrasound localises effusion and guides aspiration of difficult joints such as the hip, shoulder or small joints. Magnetic resonance imaging is useful for defining adjacent osteomyelitis or soft-tissue abscess.

Blood tests

White cell count, C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) support the picture and provide a baseline for follow-up. Blood cultures should be taken before antibiotics and are positive in a substantial minority of episodes.

Synovial fluid

Aspiration is the single most important investigation. It should be performed, with aseptic technique, for any joint that might be infected, and repeated if the first attempt is unsuccessful or the patient is already on antibiotics.

Appearance: normal synovial fluid is clear and pale yellow; septic fluid is often turbid, yellow-green or frankly purulent, with reduced viscosity. Total leucocyte count is raised, and neutrophils predominate. Protein is raised and glucose reduced relative to blood.

There is no validated diagnostic cut-off for native joint septic arthritis in adults: counts can be very high in infection, but they also overlap with crystal and other inflammatory arthritis. A systematic 2026 review of paediatric data confirms that 50,000 cells per mm3 has long been adopted as a threshold in children, yet microbiologically confirmed infection occurs below this level, with overlap against Lyme arthritis and juvenile idiopathic arthritis.[2] Because of this, synovial fluid leucocyte count is best used as a marker of intra-articular inflammation rather than a definitive diagnostic test.[2]

Gram stain has modest sensitivity and should never replace culture. Culture is the gold standard for organism identification. Nucleic acid amplification tests (NAATs) directed at exposed urogenital or extragenital sites help identify Neisseria gonorrhoeae, since joint cultures are often negative. Where Borrelia burgdorferi is suspected, two-tier serology is used.

Empirically collected fluid may be cultured, and crystal examination should always be requested because crystal arthritis and septic arthritis can coexist.

Diagnostic criteria and staging

There is no universal diagnostic criterion for native joint septic arthritis beyond isolation of a pathogen. Newman's criteria and similar scores are clinical tools, not definitions. The important practical steps are aspiration, Gram stain and culture, blood cultures, and baseline inflammatory markers.

For septic arthritis in children, a separate full discussion is deferred to paediatric resources, but the same principles apply.

Prosthetic joint infection: EBJIS approach

PJI is diagnosed with the European Bone and Joint Infection Society (EBJIS) three-level system, endorsed by MSIS and ESGIAI, which allocates tests to three groups: infection unlikely, infection likely, and infection confirmed.[3] This reflects the fact that available tests have false positives and false negatives, and there is no single definitive test.

Useful cut-offs and sampling rules from that guideline include:

  • Sinus tract or exposed prosthesis. The only fully specific clinical findings. Fever and erythema are specific but insensitive.[3]
  • Serum CRP. A level above 10 mg/L has sufficient specificity to be associated with PJI in the absence of another cause; a normal CRP does not exclude infection.[3]
  • Synovial fluid. Term cut-offs in the literature vary between 1,500 and 4,000 cells per microlitre, with PMN 65 to 80%. For confirming PJI in the hip and knee there is strong consensus for 3,000 cells per microlitre and PMN above 80%.[3]
  • Histology. The presence of five or more neutrophils in each of five high-power fields confirms infection; one or more high-power fields with five or more neutrophils makes infection likely.[3]
  • Microbiology. At least five separate tissue samples should be taken with separate instruments during revision surgery, and the patient should have been off antibiotics for at least two weeks beforehand; antibiotic prophylaxis need not be withheld. Cultures are incubated for at least 14 days. A positive culture of a common contaminant from a single sample does not confirm infection, but an uncommon contaminant or virulent organism suggests infection. Sonication fluid growth of over 50 colony-forming units per mL confirms infection, rising to 200 when the concentration technique is used.[3]
  • Bone scintigraphy. Three-phase bone scintigraphy is only reliable two years after hip arthroplasty and five years after knee arthroplasty, because of post-surgical remodelling.[3]

A middle group, infection likely, is important: one positive test does not by itself imply infection, but a raised CRP without other cause together with another positive test does.

Management

Immediate measures (all septic arthritis)

  • Analgesia, splint the joint in a comfortable position, and assess hydration, sepsis and comorbidity.
  • Obtain blood cultures and baseline CRP/ESR before starting antibiotics if the patient is not septic. Do not delay analgesia or drainage for systemic sepsis.
  • Arrange orthopaedic or infectious-disease referral early. The surgical and medical teams should work together.

Drainage and source control

Drainage is the defining step. Without it, irreversible cartilage damage may occur within about three days.[1] The review of native joint disease emphasises timely surgical source control alongside antibiotics.[1]

Options, chosen with orthopaedics:

  • Aspiration and washout through the needle for small effusions or isolated joints under ultrasound guidance.
  • Arthroscopic lavage, favoured for the knee, shoulder and ankle: best visualisation and least tissue damage.
  • Open arthrotomy, needed for the hip, for loculated pus, or when arthroscopy fails.
  • Repeated drainage or a further washout if the joint re-accumulates purulent fluid or the patient fails to improve.

The joint is usually splinted initially, then mobilised as pain allows to prevent stiffness.

Antibiotics for native joint disease

No universally accepted guideline specifies exactly which drug, dose and duration for every adult; recent reviews identify optimal duration and the oral transition as gaps;[1] follow your local formulary and susceptibility results. General evidence-based principles:

  • Empirical therapy must cover S. aureus, the predominant pathogen (about 58% of isolates), and should include MRSA where local prevalence is high (up to nearly 20% of cases in United States series during the opioid era).[1]
  • Typical adult feeling: an anti-staphylococcal beta-lactam such as flucloxacillin, dicloxacillin, cefazolin or oxacillin for methicillin-susceptible disease, and vancomycin or daptomycin for severe illness or where MRSA is likely; add Gram-negative cover (for example a third-generation cephalosporin) when Enterobacterales or Pseudomonas are risk (older adults, immunocompromise, urological source, injection drug use).
  • Gram-negative organisms account for about 12% of isolates and polymicrobial infection for about 10%, so cover on risk rather than routinely.[1]
  • Duration and oral transition. Begin intravenously and step down to oral agents with good bioavailability and joint penetration once the organism is known and the patient is improving. Recent reviews identify exact duration and oral transition as unresolved, so follow local protocol. Baseline and follow-up inflammatory markers help judge response.[1]
  • Adjust for renal and hepatic impairment and for allergies. Beta-lactams are preferred where possible; vancomycin levels need monitoring. Check local guidance if the patient has received recent intra-articular injections.
  • Culture-negative episodes. About 25% of episodes are culture-negative.[1] Treat as septic arthritis if clinical and fluid findings are consistent, often with broad-spectrum cover; re-check the diagnosis if recovery is slow.

Antibiotics for prosthetic joint infection

  • A confirmed or likely PJI requires organism-directed therapy chosen with the microbiology team. Debridement and implant retention (DAIR) is used in acute postoperative or acute haematogenous infection with a stable implant and short duration, using exchange of modular components and prolonged biofilm-active therapy; staged revision is used in chronic infection. Suppressive therapy is considered when surgery is impossible.
  • Antibiotic-loaded cement spacers are standard in staged revision. Exact durations follow local protocols and the organism; evidence reviews identify duration and oral transition as unresolved questions.[1]

Non-pharmacological care

Elevation, splinting in functional position, graduated mobilisation, physiotherapy and analgesia. Avoid weight-bearing on a septic hip until the orthopaedic team agrees.

Follow-up

Repeat the CRP weekly until it falls. Re-aspirate if the patient is not improving. Watch for secondary osteomyelitis on imaging. A failure to improve raises the chance that the process is crystal arthritis, avascular necrosis, tuberculous or fungal infection, or an unrecognised foreign body.

Complications

  • Irreversible destruction of articular cartilage and loss of joint space.
  • Osteomyelitis of adjacent bone.
  • Ankylosis or severe stiffness requiring manipulation or arthrolysis.
  • Prosthetic joint failure and revision surgery, sometimes with staged reconstruction.
  • Bacteraemia and sepsis, especially in the elderly.
  • Chronic infection, relapse or reinfection.
  • Death in frail patients with comorbidity.

Prognosis

Outcome depends on the time to drainage: same-day drainage gives the best chance of preserving function. Large joints and high-virulence organisms do worse, as does delay. In older patients, septic arthritis contributes to longer-term mortality through comorbidity rather than through the joint itself.[4]

Prevention and screening

There is no screening programme for septic arthritis. Prevention rests on treating skin and soft-tissue infections promptly, strict sterile technique for joint injections and surgery, prophylactic antibiotics for arthroplasty and high-risk procedures, and optimising diabetes, renal disease and immunosuppression. In patients with crystal arthritis, treating gout reduces recurrence of acute painful joints but does not prevent septic infection.

Special groups

  • Children. Presentation is often in the hip, with reluctance to weight-bear, irritability and refusal to move the limb. Kingella kingae is a characteristic paediatric pathogen. Management follows paediatric orthopaedic guidance. The 50,000 cells per mm3 synovial fluid threshold was originally proposed in children.[2]
  • Older adults. Symptoms may be muted and fever absent. Comorbidity impairs outcome and raises mortality.
  • Immunocompromised patients. A broader microbiological spectrum, an attenuated inflammatory response, and more difficult aspiration; consider atypical organisms.
  • Pregnancy. No specific change to drug choice beyond the usual pregnancy cautions; joint drainage is safe. Obstetric and paediatric input is useful.
  • Prosthetic joint recipients. Treat any early postoperative fever or wound problem as possible PJI until excluded; late infections may be haematogenous after a distant source.

Low-resource settings

  • A sterile needle, a basic light microscope and culture media are enough to start. Gram stain and crystal examination are cheap and high-yield.
  • Ultrasound performed by a trained clinician can localise joint fluid and guide aspiration.
  • Where arthroscopy is unavailable, open arthrotomy is safe and effective if performed promptly.
  • Empirical anti-staphylococcal therapy followed by oral step-down can be used where intravenous access and monitoring are limited, with a low threshold to re-aspirate.
  • If surgical expertise or theatre time is scarce, refer early, because stiffness and long-term disability last for life.

Key points

  • Septic arthritis is an orthopaedic emergency; delay beyond a few days risks irreparable cartilage loss.[1]
  • Typical adult presentation is acute monoarthritis of a large joint, often the knee, with systemic upset.
  • Roughly 58% of adult isolates are S. aureus (about 44% MSSA and 14% MRSA across cohorts), 83% Gram-positive and 12% Gram-negative; about 10% are polymicrobial and about 25% culture-negative.[1]
  • Risk factors include structural joint damage, bacteraemia-prone conditions, direct inoculation and immunocompromise; about one in five patients has no identifiable risk factor.[1]
  • Synovial fluid supports but does not confirm the diagnosis; leucocyte count is a marker of inflammation.[2]
  • PJI is diagnosed by the EBJIS three-level approach, with synovial fluid cut-offs around 1,500 to 4,000 cells per microlitre (confirmatory 3,000 plus PMN over 80% in hip and knee), CRP above 10 mg/L suggestive, and five tissue samples at surgery.[3]
  • Drainage plus antibiotics is the treatment; empirical cover must target S. aureus, with MRSA cover where prevalence demands.[1]
  • Cartilage damage can become irreversible within about three days without drainage.[1]
  • Older and immunocompromised patients may present subtly; maintain a low threshold.
  • Prevention centres on sterile technique, prophylaxis for arthroplasty, and control of predisposing conditions.

Check yourself

Answer each one in your head first, then open the answer.

  1. A 60-year-old man with osteoarthritis and type 2 diabetes wakes with sudden, severe pain in the right knee, which is hot and swollen. He is febrile. What is the most likely diagnosis and the next single most useful test?

    Show answerHide answer

    Septic arthritis of the right knee. The single most useful test is an urgent sterile aspiration of synovial fluid for Gram stain, culture, cell count and crystal examination. Do not delay antibiotics or drainage in systemic sepsis, but obtain the sample first where the patient is stable.

  2. Why is drainage considered essential even when intravenous antibiotics are given promptly?

    Show answerHide answer

    The closed joint space accumulates pus under raised pressure, which impairs perfusion, and bacterial enzymes and host-derived proteases degrade cartilage. Without timely drainage, irreversible cartilage damage can occur within about three days.

  3. Synovial fluid from a painful knee shows 45,000 cells per microlitre with 92% neutrophils and no crystals. The patient is stable. What does this establish?

    Show answerHide answer

    It establishes markedly inflamed joint fluid with neutrophil predominance, which supports septic arthritis but does not prove it. Crystal arthritis, rheumatoid flare and other conditions can look similar. The result should be combined with Gram stain, culture, blood cultures, serum CRP and clinical judgement; microbiology confirmation or compatible findings guide treatment.

  4. A 72-year-old woman has a painful swollen knee six months after total knee replacement, with a slowly rising CRP and no fever. How would you approach the diagnosis?

    Show answerHide answer

    Use the EBJIS three-level approach rather than a binary rule. Examine carefully for a sinus tract. Aspirate synovial fluid for cell count, culture, and alpha-defensin where available. Interpret CRP above 10 mg per litre as suggestive and remember that a normal value does not exclude infection. Take at least five separate tissue samples at surgery with extended culture incubation for at least 14 days, and manage with the orthopaedic and microbiology teams. [3]

  5. Which single microscopic feature of tissue histology is used to confirm prosthetic joint infection?

    Show answerHide answer

    Five or more neutrophils in each of five high-power fields on an intraoperative or biopsy tissue sample. One or more high-power fields with five or more neutrophils makes infection likely.

  6. What antibiotic principle is supported by the microbiology of native joint septic arthritis?

    Show answerHide answer

    Empirical therapy must always cover Staphylococcus aureus, since it accounts for roughly 58% of isolates, and Gram-negative cover is added on risk because these account for about 12%. Therapy is later narrowed to the identified pathogen. [1]

  • Rheumatoid arthritis
  • Osteoarthritis
  • Gout and crystal arthritis
  • Septic arthritis in children
  • Prosthetic joint infection
  • Infective endocarditis
  • Osteomyelitis

References

  1. Khodadadi RB, Matsuo T, Reasoner KC, Herrera LN, Colak C, White KD, Yuan BJ, Tande AJ. State-of-the-Art Review: Native Joint Septic Arthritis Epidemiology, Pathogenesis, Microbiology, Diagnosis, Management, and Outcomes. Clin Infect Dis 2026;83(4):499-513. doi:10.1093/cid/ciag473. Accessed .
  2. Sbampato V, Ramadani A, Tsoupras A, El Hasnaoui S, Debordes PA, de Marco G, Vazquez O, Steiger C, Dayer R, Ceroni D. What Weight Should Now Be Given to Synovial Fluid Leucocyte Count When Septic Arthritis Is Suspected? A Narrative Review. Microorganisms 2026;14(10):2189. Accessed .
  3. McNally M, Sousa R, Wouthuyzen-Bakker M, Chen AF, Soriano A, Vogely HC, Clauss M, Higuera CA, Trebse R. The EBJIS definition of periprosthetic joint infection. Bone Joint J 2021;103-B(1):18-25. doi:10.1302/0301-620x.103b1.bjj-2020-1381.r1. Accessed .
  4. Momodu II, Savaliya V. Septic Arthritis (StatPearls chapter). StatPearls Publishing, Treasure Island (FL), 2023. Accessed .

Educational content for medical students and doctors. It is not a substitute for clinical judgement or local guidelines.