Anticoagulation and Antiplatelet Therapy — Written Clinical Reasoning

Model answer — Part A: the warfarin hip replacement

Frame the decision as risk arbitration. His stroke risk off anticoagulation is modest and brief; the surgical bleeding risk of arthroplasty is significant. The evidence base is BRIDGE: in warfarin-treated AF patients, therapeutic-dose LMWH bridging produced no reduction in thromboembolism but significantly more major bleeding — so the plan is interruption WITHOUT bridging [1].

The protocol: last warfarin dose 5 days before surgery (day minus 6); INR on the day before or day of surgery, with low-dose oral vitamin K only if the INR remains elevated; no bridging LMWH; prophylactic-dose heparin post-op until full anticoagulation is safe; restart warfarin the evening of surgery or next day if haemostasis is secure (warfarin's slow onset is itself the bridge); full therapeutic effect returns over days [1].

If he were on apixaban: apply the PAUSE schedule — hold for about 2 days before a high-bleed-risk procedure (last dose 3 days pre-op, longer only if renal impairment), no bridging, and resume full-dose apixaban at 48–72 hours post-op given the bleeding risk of arthroplasty, using prophylactic dosing in between [3].

If the procedure were pacemaker insertion: reverse the logic entirely — BRUISE CONTROL showed continued warfarin through device surgery causes markedly less pocket haematoma than interruption with heparin bridging, so operate on therapeutic warfarin with a pre-procedure INR check [2].

Named exceptions to no-bridging earn the final marks: mechanical mitral valve, stroke/TIA or VTE within 3 months, or very high thrombotic risk — decided with haematology, not habit [1].

Model answer — Part B: GI bleed on rivaroxaban

Immediate management follows the bleeding-patient rule — thresholds do not apply, resuscitate: two large-bore cannulae, high-flow oxygen, group and crossmatch, FBC/coagulation/fibrinogen/biochemistry, lactate, catheterisation, and activation of the major haemorrhage pathway if shock persists. Establish the time of her last rivaroxaban dose — drug persistence is renal-function-dependent and drives how much active drug remains [4].

Reversal reasoning: this is life-threatening bleeding on a factor Xa inhibitor with refractory shock — reversal is indicated. Andexanet alfa (bolus then 2-hour infusion, dosed by agent and time since last dose) restored haemostasis in most ANNEXA-4 patients but carries about a 10% thrombotic event rate, reserving it for exactly this severity; 4-factor PCC is the pragmatic alternative where andexanet is unavailable. FFP has no useful role. Reversal supports — but never substitutes for — securing the source [4].

Definitive management: urgent upper GI endoscopy once resuscitated, with endoscopic haemostasis for the culprit lesion; high-dose PPI infusion for ulcer bleeding; tranexamic acid is not routine in GI bleeding. Investigate the source completely — a new bleed on anticoagulation still requires the underlying lesion to be found and treated [5].

Resumption is the physician's mark. Her unprovoked PE was only 5 months ago — she is still within the highest-recurrence-risk window and her indication has intensified, not disappeared. Restart anticoagulation once haemostasis is secure and the source is treated — typically within 1–2 weeks, individualised with gastroenterology, documented as a shared decision — and reconsider the agent and dose (apixaban's better GI bleeding profile makes it a reasonable switch) [5].