Fascinating. Are the data in this forest plot also limited to NSTEMI subgroup?

Something I've got to ask: Achieving Balloon time of ~5 hours in NSTEMI seems far better than most centers do, so aren't you afraid that the true effect of OMI detection will be masked? ( If you think of the typical 24 hours it takes to get NSTEMI to PCI)

The important part is NSTEMI-OMIs in the AI group made it to the CCL in guideline interval <90 mins. It is not entirely true that all NSTEMI-OMIs are cathed in 24 hours, many STEMI equivalents are cathed couple hours later as STEMI that was “not diagnostic initially”

OMI without STE includes patients with varying collateral sufficiency. Those with near-adequate collaterals (e.g., acute occlusion of chronically stenotic vessel) may benefit less from ultra-fast door-to-balloon times (similar to the troponin and LVEF on the forest plot). But catching posterior MI, often missed entirely, is a clear win. Your explainable ECG model could help distinguish these scenarios. Congrats on your work!