Effect of random deferral of percutaneous coronary intervention in patients with diabetes and stable ischaemic heart disease

Background In stable ischaemic heart disease (SIHD), measurement of fractional flow reserve (FFR) to guide selection of lesions for percutaneous coronary intervention (PCI) reduces death and myocardial infarction (MI) compared with angiographic guidance. However, it is unknown if the improved outcomes are due to avoidance of stenting of physiologically insignificant lesions or are a by-product of placing fewer stents. Methods We developed a Monte Carlo simulation using the PCI strata of the Bypass Angioplasty Revascularization Investigation 2 Diabetes study to investigate how random deferral of PCI impacts outcomes. To simulate deferral, a randomly selected group of patients randomised to PCI were removed and replaced by an equal number of randomly selected patients randomised to intensive medical therapy (IMT) using a random number generator in Python's NumPy module. The primary endpoint was the rate of death or non-fatal MI at 1 year. Results Death/MI at 1 year occurred in 8.3% of 798 patients in the PCI group and 5.1% of 807 patients in the IMT control group (p=0.02). Following 10 000 iterations of random replacement of 10%, 20%, 30% or 40% of PCI patients with randomly selected IMT patients, the rate of death/MI at 1 year progressively declined from 8.3% to 8.0%, 7.6%, 7.3% and 7.0%, respectively. Conclusions In this simulation model, random deferral of PCI procedures in SIHD progressively reduced death/MI as the percentage of procedures deferred increases. FFR-guided deferral of PCI may improve outcomes as a result of placing fewer stents and be unrelated to the haemodynamic severity of lesions.