Clopidogrel beat aspirin after PCI in women — but the interaction test says treat everyone the same

Design

Prespecified sex-stratified substudy of an open-label randomised trial

Population

5,506 South Korean patients at high ischaemic risk who had completed standard dual antiplatelet therapy after PCI (1,002 female, 4,504 male)

Primary outcome

Composite of all-cause death, myocardial infarction or stroke

Effect

Women 4.3% vs 9.4%, adjusted HR 0.45 (95% CI 0.24–0.85); men 4.4% vs 6.0%, adjusted HR 0.77 (0.56–1.07); P for interaction 0.288

SMART-CHOICE 3 randomised patients at high risk of recurrent ischaemic events who had finished standard dual antiplatelet therapy after PCI to clopidogrel or aspirin monotherapy. This prespecified substudy split the 5,506 South Korean patients into 1,002 women and 4,504 men. Despite a worse baseline risk profile, women did not have a significantly higher rate of the primary composite of death, myocardial infarction or stroke (6.7% vs 5.2%, adjusted HR 0.89, 95% CI 0.62–1.28).

On treatment, clopidogrel reduced the primary endpoint in women (4.3% vs 9.4%, adjusted HR 0.45, 95% CI 0.24–0.85, P = 0.014) and did not reach significance in men (4.4% vs 6.0%, adjusted HR 0.77, 0.56–1.07, P = 0.116). The temptation is to read that as a sex-specific effect. The interaction test says otherwise: P for interaction = 0.288, which is the number that governs whether two subgroup results genuinely differ.

So the practice-changing part is not about women. It is that in both sexes clopidogrel monotherapy pointed the same way against aspirin after the dual-therapy period, and the aggregate case for switching a high-ischaemic-risk post-PCI patient from aspirin to clopidogrel now has consistency behind it. Treat the sex difference as noise and the direction as signal.

• At the end of standard dual antiplatelet therapy in a high-ischaemic-risk post-PCI patient, consider clopidogrel rather than aspirin as the single agent.
• Do not make that choice on the patient's sex — the interaction test does not support it.
• Check for interacting drugs and for prior clopidogrel non-response before switching.
• Cost favours aspirin in Indian practice, and generic clopidogrel is cheap but not free — confirm the patient can sustain it before switching.
• Document the switch and the reason, so the next prescriber does not revert it.

Why it matters

A subgroup result this eye-catching usually changes prescribing before the interaction test is read; here the interaction test is the part that matters.

Don't overread it

This does not establish that clopidogrel works better in women — the sex-by-treatment interaction was not significant.

The statistics, in plain English

The women's hazard ratio of 0.45 looks far more impressive than the men's 0.77, but with 1,002 women the confidence interval runs from 0.24 to 0.85 — wide enough to be compatible with the men's estimate. That is what P for interaction = 0.288 is telling you: the difference between the two subgroups is within what chance produces. A subgroup result that is significant on its own and not significant on interaction is a hypothesis, not a finding.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.