Routine screening for silent ischaemia has not shown benefit

Design

Systematic review and random-effects meta-analysis of randomised trials; not prospectively registered

Population

3,315 asymptomatic adults with diabetes and no known coronary disease across 5 randomised trials (4 pooled)

Primary outcome

Major adverse cardiovascular events, cardiovascular and all-cause mortality

Effect

Pooled HR/RR 0.76 (95% CI 0.48 to 1.19), not significant

A systematic review looked at randomised trials that screened asymptomatic adults with diabetes and no known coronary disease for silent myocardial ischaemia. Five trials with 3,315 participants (DIAD, DYNAMIT, FACTOR-64, DADDY-D and a pilot) were included. Four were pooled: screening did not significantly reduce major cardiovascular outcomes (pooled HR/RR 0.76, 95% CI 0.48 to 1.19).

Screening did find occult ischaemia, but this did not translate into fewer events compared with the medical management used in the trials. Reporting of false positives and procedure harms was incomplete. The review was not prospectively registered, and one trial was analysed narratively.

This fits with the approach of treating risk factors and reserving testing for symptoms or situations where a result would change management. The point estimate favoured screening, so a modest benefit cannot be ruled out, and the trials predate current therapy such as SGLT2 inhibitors and GLP-1 receptor agonists.

• Spend the effort on risk factors first: blood pressure, lipids, glucose, smoking, and cardioprotective drugs where indicated
• Test for ischaemia in people with symptoms, including breathlessness or fatigue that could be an atypical presentation
• Consider selective testing before major surgery or when the result would change management
• Discuss false-positive risk and downstream angiography with patients who ask for a screening stress test
• Existing guidelines have not changed on the basis of this review

Why it matters

It challenges the common habit of ordering a stress test as part of an annual diabetes check.

Don't overread it

The trials are older and no benefit shown is not proof of no benefit; the guideline position has not changed.

The statistics, in plain English

The pooled estimate of 0.76 points towards fewer events, but the interval of 0.48 to 1.19 includes 1, so the data are compatible with a sizeable benefit or with none. A review of only a few trials cannot rule out a small effect. The review itself was not prospectively registered, which allows more room for analytic choices.

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.