Design: Cluster-randomised trial, 10 centres
Population: 329 patients with ACS (315 analysed), median LDL 110 mg/dL
Primary outcome: LDL cholesterol <70 mg/dL at 6 months
Effect: 86.4% vs 73.7%; difference 12.6 points (95% CI 3.8–21.5)

BRIDGE was a cluster-randomised trial across 10 Japanese centres. Protocol hospitals started or intensified lipid-lowering therapy during the index ACS admission using an algorithm of high-intensity statin, ezetimibe and PCSK9 inhibitor, rechecked LDL cholesterol at four weeks and escalated if it was above about 55 mg/dL.

In 315 analysed patients (median baseline LDL 110 mg/dL), 86.4% of protocol patients reached LDL below 70 mg/dL at six months against 73.7% with usual care (difference 12.6 percentage points, 95% CI 3.8–21.5). For below 55 mg/dL the gap was larger: 60.8% vs 34.5%. A hospital-level sensitivity analysis gave a similar estimate but a wider interval that crossed zero.

This is an implementation trial, not an outcomes trial. Its value is showing that a simple rule — start combination therapy before discharge and check at four weeks — closes much of the gap between guideline and practice.

Start high-intensity statin before discharge after ACS, and add ezetimibe early if LDL is well above target.
Book an LDL cholesterol check at four weeks, not at the next routine clinic visit.
Escalate if LDL is above 55 mg/dL (1.4 mmol/L) at four weeks.
Put the algorithm in the discharge order set so it does not depend on individual memory.
In India, generic ezetimibe is inexpensive; cost should not delay a statin–ezetimibe combination.

Why it matters

It shows the gap in post-ACS lipid control is a systems problem that a simple protocol can close.

The statistics, in plain English

In a cluster trial whole hospitals are randomised, so patients in one hospital are more alike than patients across hospitals. The hospital-level sensitivity analysis accounts for that more strictly, and its interval (−2.0 to 27.6) crosses zero — meaning the result is less certain than the headline interval suggests. With only 10 centres, that caution matters.