How should I evaluate and rule out septic arthritis?

Ruling Out Septic Arthritis

Start with plain radiographs of the affected joint, then proceed directly to arthrocentesis with synovial fluid analysis—this is the definitive diagnostic step you cannot skip. 1

Initial Imaging Approach

Begin with plain radiographs of the affected joint. While these are often normal in early septic arthritis (<14 days), they provide essential baseline information by excluding fractures and tumors, and may show joint effusion or soft tissue swelling 1. Radiographs are the recommended first-line imaging for all suspected musculoskeletal infections 1.

The Critical Diagnostic Step: Arthrocentesis

Imaging alone cannot distinguish infected from noninfected joints—you must perform arthrocentesis with culture to definitively rule out septic arthritis 1. This is non-negotiable because:

• Culture identifies the causative organism and directly guides antibiotic selection 1
• Clinical findings alone are insufficient (the classic triad of fever, pain, and diminished mobility occurs in only ~50% of cases) 2
• Any isolated arthritis in patients with inflammatory arthritis should be considered septic until proven otherwise 3

Synovial Fluid Analysis Interpretation

The synovial white blood cell (sWBC) count is your most powerful diagnostic tool 4, 5:

Likelihood ratios by sWBC count:

• <25,000/μL: LR 0.32 (makes septic arthritis less likely)
• 25,000-50,000/μL: LR 1.06 (doesn't change probability)
• 50,000-100,000/μL: LR 7.7 (significantly increases probability)

• 100,000/μL: LR 28.0 (strongly suggests septic arthritis) 4

Critical caveat: More than one-third of adults with confirmed septic arthritis have sWBC <50,000/μL 6. Do not use the traditional 50,000/μL cutoff to rule out septic arthritis—this is a common and dangerous pitfall 5, 6.

Polymorphonuclear cell percentage:

• ≥90% PMNs: LR 3.4 (suggests septic arthritis)
• <90% PMNs: LR 0.34 (makes it less likely) 4

Emerging Diagnostic Markers

Synovial lactate shows promise with LR ranging from 2.4 to infinity for ruling in septic arthritis 5. A composite scoring system (combining sWBC, PMN%, serum CRP, and leukocytes) demonstrated 95.6% sensitivity in recent data 7, though this requires prospective validation.

Risk Factors That Significantly Increase Probability

Only two historical factors substantially raise septic arthritis likelihood:

• Recent joint surgery: LR 6.9 5
• Skin infection overlying a prosthetic joint: LR 15.0 5

Other risk factors include age, diabetes, rheumatoid arthritis, hip/knee prosthesis, and HIV infection 4.

Serum Laboratory Tests: Limited Utility

Serum markers (WBC, ESR, CRP) are not useful for ruling out septic arthritis acutely 5:

• Peripheral WBC sensitivity: only 48% 6
• ESR sensitivity: 96% (useful for ruling out, but not specific) 6
• All patients with septic arthritis have at least one abnormal test, but normal values don't exclude the diagnosis 6

Advanced Imaging: When and Why

If initial radiographs show effusion or soft tissue swelling and you need to plan arthrocentesis:

Use image-guided aspiration (ultrasound, CT, or fluoroscopy) because it:

• Confirms accurate needle placement 1
• Significantly reduces vascular and nerve complications 1
• Allows planning of needle trajectory to avoid contaminating adjacent tissues 1

MRI or CT with IV contrast may be considered if:

• You need to assess extent of infection for surgical planning 1
• You suspect associated osteomyelitis or abscess 1
• However, these should not delay arthrocentesis 1

Algorithmic Approach

1. Obtain plain radiographs of the affected joint
2. Perform arthrocentesis (preferably image-guided) immediately if clinical suspicion exists
3. Send synovial fluid for:
   • Cell count with differential
   • Gram stain and culture (gold standard)
   • Consider lactate if available
4. Interpret results cautiously: Even low sWBC counts don't exclude septic arthritis
5. When in doubt, treat as septic arthritis—the mortality risk (2-15%) 8 and joint destruction risk mandate aggressive management

The test threshold for arthrocentesis is approximately 5% probability, and the treatment threshold is 39% probability 5, meaning if your clinical suspicion reaches even 5%, you should perform arthrocentesis.