Knee osteoarthritis flare in a patient with CKD and anticoagulation
Mr E, 72, has a 4-year history of right knee osteoarthritis. He presents with a 3-week flare: pain 8/10 at night, knee swelling, warmth, and inability to walk more than 50 m. He takes apixaban for atrial fibrillation, has CKD stage 4 (eGFR 24), and takes bendroflumethiazide. He has tried paracetamol and topical diclofenac. He asks for strong tablets like his wife's and for a steroid injection. He has an outpatient MRI report from 6 months ago showing bone-on-bone medial compartment, and asks whether keyhole surgery would cure it. Temperature is 36.8°C, BP 150/85, knee is warm with a moderate effusion, no erythema, ROM 10–100°.
1. What is your differential, and what immediate action is required before treating this as a simple osteoarthritis flare?
Read the worked answer
The differential includes septic arthritis, crystal arthritis, and a severe osteoarthritis flare. A warm swollen knee with night pain and inability to walk 50 m is a red flag even without fever; absence of fever does not exclude infection, especially in older adults.
I would arrange same-day assessment with bloods and knee aspiration before giving antibiotics or a steroid injection. I would check CRP, full blood count, blood cultures, and send synovial fluid for cell count, Gram stain, and crystals. I would not inject steroid until infection is excluded, because intra-articular steroid in a septic joint worsens outcomes.
I would also review the medication list: bendroflumethiazide may contribute to dehydration and renal impairment, and apixaban increases bleeding risk for aspiration, though the procedure is usually still possible with precautions. The MRI report is historical and does not explain the acute warmth. The key reasoning is to treat the flare as potentially infective until proven otherwise, while recognising that crystal arthritis can mimic infection.
2. How would you manage his medication and the request for a steroid injection?
Read the worked answer
I would avoid systemic NSAIDs because of CKD stage 4, heart failure risk, and apixaban. I would also avoid regular codeine without a clear plan because of renal accumulation and falls risk.
- Continue paracetamol at a safe dose (usually ≤3 g/day in this context) and consider topical NSAID only if the skin is intact and renal function is stable, with a review.
- If infection is excluded, an intra-articular corticosteroid is reasonable for a severe flare; it gives short-term relief and is safer than systemic NSAIDs. I would check renal function and coagulation, counsel about bleeding and infection, and document consent.
- If septic arthritis is confirmed, admission for intravenous antibiotics and orthopaedic review is required.
I would explain that a steroid injection is a bridge, not a cure, and pair it with exercise and weight management. I would review at 48–72 hours and give safety-net advice to return if pain, swelling, or fever worsens.
3. How would you answer his question about keyhole surgery, and what would you document?
Read the worked answer
I would explain that arthroscopic knee surgery does not cure osteoarthritis and is not recommended for routine degenerative knee pain. Trials show little or no benefit over exercise and education, with small surgical risks. His 'bone-on-bone' MRI describes structural change, not the whole pain experience, and imaging findings correlate poorly with pain.
I would redirect to what does help: land-based exercise, strength and balance work, weight management, walking aids, and analgesic optimisation. If he develops end-stage disease with severe night pain, deformity, or failed conservative care, referral for joint replacement assessment would be appropriate.
I would document the shared decision, his expectations, the risks of injection and surgery, and the safety-net plan. This is an evidence-based practice issue as well as a communication issue: the clinician must avoid overtreatment while respecting the patient's distress and maintaining trust.