DAPT Duration After PCI: Decisions at the Bedside
A 71-year-old woman attends your clinic with a discharge summary in a language you do not read, a stent card she cannot find, and a surgeon who wants to repair an incarcerated hernia next week. She had “a heart procedure” about five months ago and takes two tablets that she describes by colour. Deciding how long her dual antiplatelet therapy should continue, and whether it can safely be interrupted, is one of the commonest and least standardised decisions in international practice, and it is usually made without the implanting operator on the phone. This reference works through it as a sequence of bedside choices rather than as a summary of guideline tables.
On this page
Framing the question properly
The old question was how many months of treatment a stent requires. The modern question is different: how do the ischaemic and bleeding hazards for this individual cross over in time, and which drug should be carrying the load once they do. Ischaemic risk after intervention is front-loaded and decays over the first months, while bleeding risk is broadly constant and accumulates with exposure. Most of the trial programme of the past decade has been an attempt to find the crossing point, and the practical answer has shifted away from uniform durations towards shorter combined therapy followed by a single potent agent.
The decision points
1. Establish what was implanted, where, and when
Before any duration decision, reconstruct the index procedure. You need the date, the vessel or vessels, the number and length of stents, whether the left main or a bifurcation was treated, and whether the indication was an acute coronary syndrome or stable disease. Where records are unobtainable, a plain fluoroscopic look or the operator’s institution email will often yield more than the patient’s recollection. Treat an undocumented procedure as complex until proven otherwise.
2. Separate acute from chronic coronary syndrome, because the default differs
An acute presentation carries a higher residual thrombotic hazard than elective intervention for stable disease, and current guidance favours a longer default combined course after acute coronary syndrome than after elective stenting, alongside a preference for a potent P2Y12 inhibitor in the acute setting where bleeding risk permits. Elective intervention in a patient with low complexity anatomy can generally be managed with a shorter combined course.
3. Score the bleeding risk formally, not impressionistically
Clinicians consistently underestimate bleeding risk in older patients and overestimate it in patients who simply look frail. Use PRECISE-DAPT at the point of discharge or review, and apply the high bleeding risk criteria set out by the Academic Research Consortium, which capture the features that actually drive events: anaemia, chronic kidney disease, prior spontaneous bleeding, active malignancy, planned surgery and long-term steroid or anti-inflammatory use. In patients who also require anticoagulation, HAS-BLED adds a structured view of the modifiable elements. Grade any bleeding that does occur with the BARC scale so that the next clinician inherits a comparable number.
4. Score the ischaemic side with equal seriousness
Complexity matters more than any single comorbidity. Three or more stents, three or more lesions treated, total stent length beyond about sixty millimetres, bifurcation stenting with two stents, left main intervention, chronic total occlusion recanalisation and last remaining patent vessel all shift the balance towards longer therapy. Diabetes, chronic kidney disease, prior stent thrombosis and recurrent events do the same. The anatomical complexity language of the SYNTAX and EXCEL programmes is useful here, even though those trials addressed revascularisation strategy rather than antiplatelet duration.
5. Choose the agent before you choose the duration
Clopidogrel remains the default in stable disease and in patients with meaningful bleeding vulnerability, with the caveat of variable metabolism. Ticagrelor and prasugrel deliver more consistent and more potent inhibition after acute coronary syndrome; prasugrel is avoided after stroke or transient ischaemic attack and used cautiously in low body weight and advanced age, while ticagrelor requires twice-daily dosing and carries dyspnoea as a recognised nuisance effect that drives discontinuation if not anticipated and explained.
6. Decide between shortening, standard and extending
Three strategies now coexist. Abbreviated combined therapy followed by monotherapy with the P2Y12 agent was the approach tested in TWILIGHT and, in a bleeding-vulnerable population, in MASTER DAPT; both support dropping aspirin early rather than dropping the potent agent. Standard duration remains reasonable for the average patient with an acute presentation and no bleeding flags. Extension beyond a year is for the patient with recurrent events, diffuse disease and low bleeding risk, and is a decision to review annually rather than to set and forget. De-escalation from a potent agent to clopidogrel after the first months is a further legitimate option in patients who tolerate the combination poorly.
7. Handle concurrent anticoagulation as its own problem
When atrial fibrillation or a mechanical valve coexists with recent stenting, the direction of travel is clear: keep triple therapy short, then continue an oral anticoagulant with a single antiplatelet agent, usually clopidogrel. The AUGUSTUS and PIONEER AF-PCI programmes established the shape of that strategy, and the practical consequence is that aspirin is the component that leaves first in most patients. Anticoagulant dose reduction to a level not validated for stroke prevention is a common and serious error in this group.
8. Plan interruptions before they are demanded
Elective surgery should be deferred past the highest-risk window wherever possible. When it cannot be, negotiate the specific procedure rather than the abstract principle: many endoscopic, dental, ophthalmic and superficial procedures proceed safely on continued therapy. Where interruption is genuinely required, stop for the shortest interval, keep aspirin running where the surgeon will accept it, and book the restart date into the operation note. Bridging with parenteral agents is reserved for the small group at genuinely prohibitive thrombotic risk. Wider perioperative and shared-care context is set out in our cardiology handbook for physicians.
When to escalate, transfer or call for help
- Contact the implanting service before any interruption within the first six months, and before de-escalating in a patient with left main or complex bifurcation stenting. Reference material on why those substrates behave differently is collected in our advanced PCI techniques volume and the interventional cardiology collection.
- Escalate urgently if chest pain recurs after any reduction in therapy. Treat it as stent failure until angiography or a convincing alternative says otherwise; the acute pathway is set out in the acute coronary syndromes reference.
- Involve haematology or gastroenterology for recurrent BARC type 2 bleeding rather than serially shortening therapy on your own. Proton pump inhibition, iron replacement and source treatment often preserve the regimen.
- Seek specialist input for the patient who needs anticoagulation, has had a stent thrombosis, and bleeds. That combination has no clean answer and should not be resolved by a single clinician in a corridor.
What to document
- Index procedure date, indication, vessels, number and total length of stents, and any complexity features.
- The bleeding score used, its value, and the individual high bleeding risk criteria met.
- The agent chosen with the reason, particularly when a potent agent is avoided or a de-escalation is made.
- The planned stop date, written as a date rather than a duration, and who is responsible for reviewing it.
- Any interruption: the reason, the interval, and the confirmed restart.
- The conversation with the patient about not stopping treatment on the advice of a non-cardiac clinician without contact.
Bench card
| Clinical situation | Direction of travel | Practical note |
|---|---|---|
| Elective stenting, simple anatomy, low bleeding risk | Shorter combined therapy, then monotherapy | Clopidogrel usually sufficient |
| Acute coronary syndrome, low bleeding risk | Longer default with a potent agent | Review complexity before shortening |
| High bleeding risk by ARC criteria | Abbreviated combined therapy | Drop aspirin, retain the P2Y12 agent |
| Complex or multivessel intervention | Favour longer therapy | Stent length and bifurcation technique drive this |
| Atrial fibrillation with recent stenting | Short triple, then anticoagulant plus clopidogrel | Never reduce the anticoagulant below the stroke prevention dose |
| Prior stent thrombosis | Prolonged therapy with a potent agent | Abbreviated strategies do not apply |
| Surgery required within six months | Defer if possible; negotiate if not | Write the restart date in the operation note |
Three recurring errors worth naming: treating duration as a property of the stent rather than of the patient, stopping the potent agent and keeping aspirin when the modern evidence points the other way, and allowing a non-cardiac clinician to discontinue therapy without a documented conversation. Broader risk-factor management alongside these decisions sits in the prevention and lifestyle reference hub. A guideline-structured account of the same subject is available as this explanation of DAPT duration after PCI.
Questions from the floor
The patient has been on treatment for eleven months with no problems. Do I simply stop at twelve?
Not automatically. Twelve months is a convention derived from trial design, not a biological boundary. Review complexity, residual disease and bleeding history, and make an explicit decision either to stop, to continue a single agent indefinitely, or to extend the combination.
Which single antiplatelet agent should continue long term?
The field has moved away from assuming aspirin. In patients who have completed a combined course after stenting, continuing the P2Y12 inhibitor alone is increasingly favoured, particularly where bleeding risk is a concern, and is supported by the monotherapy arms of the abbreviated-therapy trials.
Is platelet function or genotype testing worth doing?
Routine testing has not shown consistent outcome benefit, but selective genotyping has a place after an event on clopidogrel, in complex left main intervention, and where a potent agent is being avoided for cost reasons and you want to know whether that decision is safe.
How do I handle a patient who cannot afford a potent agent?
State the compromise openly in the notes, use clopidogrel with a lower threshold for genotyping if available, extend rather than shorten the combined course in complex anatomy, and address adherence support directly. An interrupted potent agent is worse than a continuous weaker one.
