Immediate non-culprit PCI in STEMI linked to higher one-year mortality

Design

Frequentist pairwise meta-analysis of four randomised trials, PROSPERO registered

Population

2,651 patients with STEMI and multivessel disease treated with second-generation or newer drug-eluting stents

Primary outcome

All-cause death at 1 year

Effect

Immediate vs staged: HR 1.50 (95% CI 1.02 to 2.21), prediction interval 0.80 to 2.81, I² 0%

A meta-analysis pooled four randomised trials (2,651 patients) comparing immediate with staged complete revascularisation in STEMI with multivessel disease. All used second-generation or newer drug-eluting stents and followed patients for at least a year. The protocol was registered.

One-year all-cause death was higher with immediate treatment (HR 1.50, 95% CI 1.02 to 2.21, P = 0.04), but the prediction interval ran from 0.80 to 2.81. Recurrent infarction was lower with immediate treatment when procedural infarction was counted (HR 0.58, 0.36 to 0.94), but spontaneous infarction did not differ (HR 0.91, 0.55 to 1.48). Cardiac death and unplanned revascularisation were similar.

The authors warn that few deaths drove the mortality result, so it is fragile. Even so, it removes the reassurance that doing everything at once is safe, and it supports staging non-culprit lesions in stable patients unless a clear reason exists to treat immediately. Current guidelines recommend complete revascularisation and have not changed on this evidence.

• Treat the culprit lesion first and consider staging non-culprit lesions in stable patients
• Immediate treatment of non-culprit lesions may suit cardiogenic shock or refractory ischaemia; this analysis did not address those cases
• Avoid long delays: plan the staged procedure before discharge or within the index admission, per local practice
• Discuss the uncertainty with the heart team when a patient has a complex anatomy
• Existing guidelines have not changed, so use local protocol and judgement

Why it matters

It turns the timing of non-culprit PCI from a matter of convenience into a question with a possible mortality cost.

Don't overread it

The mortality signal rests on few deaths and a prediction interval that includes no difference; guidelines on complete revascularisation have not changed.

The statistics, in plain English

A hazard ratio of 1.50 with a lower confidence limit of 1.02 only just clears 1, and the prediction interval (0.80 to 2.81) crosses it. That is why the authors call the finding statistically fragile: a handful of deaths either way could change it. The lower infarct rate with immediate treatment came from counting procedural infarctions, which are less clinically meaningful than spontaneous ones.

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.