Design: Pragmatic randomised non-inferiority implementation trial
Population: 3,543 adults with suspected acute coronary syndrome in two emergency departments
Primary outcome: Discharge within four hours; 30-day major adverse cardiac events for safety
Effect: Discharge 21.8% vs 19.2% (p=0.07); sensitivity 93.7% vs 89.5%, non-inferior

MACROS2 randomised 3,543 patients with suspected acute coronary syndrome at two North-West England emergency departments to the European Society of Cardiology 0/1-hour high-sensitivity troponin pathway or a 0/3-hour pathway, with an independent assay used for adjudication.

The faster pathway did not translate into faster discharge: 21.8% of patients left within four hours on the 0/1-hour pathway against 19.2% on 0/3-hour, not a significant difference. The 0/1-hour pathway was non-inferior for safety, with a sensitivity of 93.7% against 89.5%, but both pathways were less safe than observational studies had suggested. Laboratory turnaround of about 81 minutes and system constraints absorbed the theoretical time saving.

The lesson is operational. A shorter sampling interval only helps if the laboratory and the department can act on the earlier result; where turnaround is slow, moving to 0/1-hour sampling alone will not clear the department faster.

Pragmatic randomised trial of 3,543 patients with suspected ACS across two emergency departments.
Discharge within four hours was similar: 21.8% with 0/1-hour versus 19.2% with 0/3-hour (not significant).
The 0/1-hour pathway was non-inferior for safety (sensitivity 93.7% versus 89.5%).
Both pathways were less safe than earlier observational data implied.
Laboratory turnaround of about 81 minutes was the main bottleneck, not the sampling interval.

Why it matters

The efficiency case for 0/1-hour sampling rests on fast turnaround that many real departments do not have.

The statistics, in plain English

Non-inferiority here means the 0/1-hour pathway was no worse than 0/3-hour within a pre-set 3% margin, not that it was better. A sensitivity of 93.7% means roughly 6 in 100 true events were missed at discharge, a reminder that neither pathway is perfectly safe.