Hot Swollen Knee on Apixaban: Aspirate or Treat Blindly
A 68-year-old man presents with 2 days of severe right knee pain and swelling. He cannot bear weight. T 38.1 C, HR 96, BP 138/82, RR 18. The knee is warm, tense with a large effusion, held at about 20 degrees of flexion with range limited by pain; there is no overlying cellulitis or wound. He takes apixaban 5 mg twice daily for atrial fibrillation (last dose 9 hours ago), metoprolol and atorvastatin. Platelets 210,000, INR 1.1, creatinine 1.3 mg/dL (CrCl about 48 mL/min), WBC 13,400. He had two uncomplicated arthrocenteses in the past. The same knee was injected with corticosteroid by his rheumatologist 3 weeks ago. He asks, "Can't you just give me antibiotics and skip the needle?"
1. Decide whether to aspirate now despite apixaban, and describe the precautions you take.
Read the worked answer
Aspirate. The needle is the diagnosis, and delay is the main risk here.
Reasoning:
- Septic arthritis is a time-critical emergency; cartilage destruction can occur within days. Empirical antibiotics alone are imprecise: they cover the wrong organism, miss crystals, and commit him to weeks of therapy without a pathogen.
- Apixaban is not a contraindication to arthrocentesis. Joint aspiration is a superficial, compressible procedure with low bleeding risk; the consequences of a missed septic joint far outweigh a small hematoma. Reversal or holding is not indicated, and there is no threshold INR/anti-Xa to wait for - INR does not reflect apixaban effect.
- His CrCl of 48 means somewhat prolonged apixaban exposure, but this does not change the decision.
Precautions:
- Use a 18–20 G needle with a landmark or, preferably, ultrasound-guided approach, staying away from the saphenous vein and the superolateral vessel; prep and drape sterilely, use lidocaine for the skin.
- Apply firm pressure for at least 5 minutes afterwards, then a compressive dressing; check distal pulses and neurovascular status before and after.
- Do not delay for a repeat coagulation panel. Document the shared decision.
- Have a low threshold to involve orthopedics if the tap is dry, the knee is tense after aspiration, or he is systemically unwell.
2. What do you send on the synovial fluid, how do you interpret it, and when do you give antibiotics relative to the tap?
Read the worked answer
Send: cell count with differential, Gram stain and culture, crystal analysis under polarized light, and (if there is any chance of a prosthesis or prior surgery) additional studies per orthopedics. Simultaneously draw blood cultures x2, CBC with differential, CRP, ESR, and glucose.
Interpretation:
- Synovial WBC greater than 50,000/mcL with more than 90% neutrophils strongly suggests septic arthritis; 2,000–50,000 suggests an inflammatory process (gout, pseudogout, reactive, or post-injection synovitis).
- Apixaban does not alter the cell count, so anticoagulation is not a reason to distrust the result.
- The corticosteroid injection 3 weeks ago muddies the picture: a post-injection flare can produce an inflammatory fluid, and it also slightly raises infection risk. A moderately elevated WBC does not exclude infection in that setting.
- Crystals do not exclude sepsis - concurrent gout and infection occur. Gram stain is insensitive; a negative stain does not rule it out.
Antibiotic timing: do not delay for the result if suspicion is high. Aspirate first, then start empirical IV therapy (for example vancomycin plus ceftriaxone, or per local protocol) rather than giving antibiotics and then tapping - antibiotics reduce culture yield. If the tap is delayed for any reason, treat first and document. Admit for IV therapy and orthopedic evaluation, with repeat aspiration or washout if he does not improve within 24–48 hours.
3. Explain to him why skipping the needle is not in his interest, and address his anticoagulant.
Read the worked answer
Plain-language script: "I understand why the needle sounds worse than a pill. But arthritis in a joint can be caused by at least three very different things - bacteria, crystals like gout, or inflammation from your recent steroid injection - and the treatment for each is completely different. Antibiotics will not help crystals, and gout medicine will not cure an infection. If we guess wrong and it is an infection, the bacteria can damage the cartilage in your knee permanently within days, and the wrong antibiotic or no antibiotic can let it spread into your blood. The needle takes about a minute and is the only way to know which one we are treating."
On the blood thinner: "You do not need to stop apixaban for this. This is a shallow needle into a joint we can press on afterwards, and the risk of a significant bleed is low. Stopping it would raise your stroke risk for no real gain, and we do not need to wait or reverse it." Explain the pressure dressing and what to watch for at home (rapid re-swelling, increasing pain, numbness, bleeding that does not stop).
Close the loop: confirm his understanding with a teach-back ("tell me in your own words why we are doing the tap"), state the plan for results and admission, and document the discussion. Invite questions and acknowledge that his preference is reasonable - the goal is informed agreement, not compliance. If he still declines, document the informed refusal, treat empirically, and arrange urgent orthopedic follow-up; that is a worse but not unreasonable path.