The default position for a patient on anticoagulation who needs a peripheral joint or soft tissue injection is that the procedure can usually go ahead without interrupting treatment, because the bleeding risk of a low-risk injection is small while the thrombotic risk of stopping anticoagulation is real and sometimes catastrophic. That default is not automatic, and the sport and exercise medicine (SEM) clinician must still check the indication for anticoagulation, assess the bleeding risk of the specific procedure, confirm that anticoagulation is at a therapeutic rather than supratherapeutic level, and follow local policy. This page covers how anticoagulants and antiplatelets differ in relevance, how to assess procedural bleeding risk, and how to make and document a safe decision, including when to seek advice rather than proceed.
Which drugs matter, and how much?
The drugs are not equivalent and it helps to separate them. Antiplatelet agents such as aspirin and clopidogrel modestly impair platelet function and are generally continued for peripheral joint and soft tissue injections, since stopping them for a low-risk procedure exposes the patient to cardiovascular risk without meaningful benefit. Dual antiplatelet therapy, and treatment following a recent coronary stent, must never be stopped without specialist advice. Vitamin K antagonists, principally warfarin, are monitored by the international normalised ratio (INR), and the practical requirement is that the value is known, recent and within the intended therapeutic range; a supratherapeutic INR is a reason to defer and correct rather than to proceed. A direct oral anticoagulant (DOAC), such as apixaban, rivaroxaban, edoxaban or dabigatran, presents a different problem, because there is no routine, widely available test equivalent to the INR that reliably quantifies the anticoagulant effect. Rather than describing the level as therapeutic, the checks are that the dose is correct for the indication and renal function, that the patient is adherent and when the last dose was taken, what the renal function is, whether interacting medicines are being taken, and whether there is any evidence of overdose or active bleeding. Low molecular weight heparin is similarly timing-dependent, and the appropriate interval differs between prophylactic and therapeutic dosing and is also influenced by renal function and the specific procedure. In all cases the underlying indication is the key context: anticoagulation for a mechanical heart valve, a recent venous thromboembolism or atrial fibrillation with high stroke risk carries a very different consequence of interruption than a low-risk indication.
Antiplatelets are generally continued, though dual therapy or a recent coronary stent needs specialist advice; warfarin requires a checked in-range international normalised ratio, and direct oral anticoagulants have no equivalent routine test, so dose, last dose and renal function matter. Superficial compressible targets are lower risk than deep or spinal ones.
Procedural risk is the other half of the equation, and it is determined by depth and compressibility rather than by the joint's name. A superficial injection into a compressible site, where direct pressure can control bleeding and where a haematoma would be visible and usually amenable to direct pressure, is low risk: a subacromial injection, a small joint of the hand, a superficial bursa or a tendon sheath. Higher-risk procedures are those where the needle passes deep, where bleeding cannot be compressed, or where a haematoma would be confined and damaging: deep gluteal and iliopsoas injections, deep hip procedures and, above all, spinal and neuraxial procedures, where a haematoma can cause neurological compromise and where specialist protocols with defined dosing intervals apply. Peripheral arthrocentesis, including aspiration of a large joint, is generally regarded as low bleeding risk and is usually acceptable on anticoagulation when performed carefully, subject to individual and procedural factors, with the recognition that a haemarthrosis is a recognised though uncommon consequence. Practical technique reduces risk further: use the smallest needle appropriate to the task, avoid multiple passes, take a route that avoids visible vessels, use ultrasound with colour Doppler where vascular structures are near, and apply firm direct pressure afterwards.
How should the decision be made and recorded?
A defensible decision follows a consistent sequence. Establish the indication for anticoagulation and what would happen if it were interrupted, since this determines whether interruption is even on the table. Establish the current anticoagulation status: for warfarin this means a recent in-range INR, while for a direct oral anticoagulant it means checking the dose is correct, adherence and the timing of the last dose, renal function, interacting medicines, and any evidence of overdose or active bleeding. Assess the bleeding risk of the specific procedure using depth and compressibility. Check local policy, because organisations differ and the safest practice is the one your service can support if something goes wrong. Then decide, and where the answer is not clear, discuss with the clinician or service managing the anticoagulation rather than making a unilateral change; for spinal or other high-risk procedures the decision belongs with the specialist performing them and follows the relevant protocol. Specialist advice is also sought where anticoagulation is combined with antiplatelet treatment, or where there is thrombocytopenia, significant renal or liver disease, or a previous major procedural bleed. The overriding principle is that anticoagulation is not stopped without a clear reason, and where any interruption is genuinely required, the plan for stopping and restarting is made with the team responsible for it and communicated to the patient in writing.
The sequence is indication and thrombotic risk, therapeutic level, procedural bleeding risk, local policy, then decide and seek advice if unclear. The record should show the reasoning, not simply that the injection went ahead.
The consent conversation and the record should reflect the reasoning. The patient is told that bleeding and bruising are more likely than usual, what a significant haematoma would look and feel like, and when to seek review, and this safety-net advice is more important than usual in this group. The record should show that the indication, current anticoagulation status, procedural risk and local policy were considered, what was decided and why, and that the patient was informed; a note that simply says the patient was on anticoagulation and the injection went ahead does not demonstrate a decision. Two further points are worth carrying into practice. First, a bleeding tendency is not confined to prescribed drugs: liver disease, thrombocytopenia, an inherited bleeding disorder or a personal history of easy bruising and prolonged bleeding all warrant the same care, and a personal or family bleeding history is a more useful screen than a routine clotting test in an otherwise well patient. Second, a patient who has had a haemarthrosis or a significant haematoma after a previous injection deserves a rethink rather than a repeat, since the previous event is better evidence of their individual risk than any general rule.
Exam Tips
•Low-bleeding-risk peripheral injections can usually proceed without interrupting anticoagulation, after checking the indication, the bleeding risk, that anticoagulation is therapeutic rather than supratherapeutic, and local policy.
•Antiplatelets such as aspirin and clopidogrel are generally continued; stopping them for a low-risk injection adds cardiovascular risk without benefit, and dual antiplatelet therapy or recent coronary stenting must not be stopped without specialist advice.
•Warfarin requires a recent international normalised ratio (INR) within the therapeutic range; a supratherapeutic value is a reason to defer rather than proceed.
•A direct oral anticoagulant (DOAC) has no routine, widely available test equivalent to the international normalised ratio, so check the dose, adherence and last dose, renal function, interacting medicines and any evidence of overdose or bleeding.
•Procedural risk depends on depth and compressibility: superficial compressible targets and peripheral arthrocentesis are generally low risk, whereas deep gluteal, iliopsoas and especially spinal or neuraxial procedures are high risk and follow specialist protocols.
•Do not stop anticoagulation without a clear reason; where interruption seems necessary, discuss with the team managing it, and record the indication, status, procedural risk, policy and decision.