Coronary disease & ACSResearch article
Comparative validation of risk scores for in-hospital complications following chronic total occlusion percutaneous coronary intervention
- retrospectively analysed
- single tertiary center
Authors' conclusion
In CTO-PCI, BMC2 offers superior performance for in-hospital mortality prediction. Their complementary features suggest BMC2 as a powerful, data-demanding, and patient-focused tool, while PROGRESS-CTO as an easy-to-calculate, procedure-oriented model.
Study context
- Population
- Chronic total occlusion percutaneous coronary intervention (CTO-PCI) procedures performed at a single tertiary center between 2019 and 2023
- Participants
- 1,157 CTO-PCI procedures
- Comparison
- We compared the performance of 2 risk models (BMC2 and PROGRESS-CTO) for the prediction of in-hospital outcomes after CTO-PCI
- Outcome
- In-hospital mortality
- Follow-up
- In-hospital
Main finding
As reported in the abstract
AUC
0.96 vs 0.71
P < 0.001
By group
BMC2
0.96
AUC
PROGRESS-CTO
0.71
AUC
- Mortality
- 1.0% (n = 11)
Further findings
- BMC2 discrimination for complications
- 0.93 · Acute kidney injury AUC · Dialysis AUC 0.91 · Stroke AUC 0.84
- PROGRESS-CTO discrimination for complications
- 0.71 · Periprocedural myocardial infarction AUC · Pericardiocentesis AUC 0.71 · Coronary perforation AUC 0.72 · major adverse cardiac and cerebrovascular events (MACCE) AUC 0.62
MACCE · in-hospital
- MACCE occurred
- 3.0% (n = 35)
Calibration and Brier score
- Better calibration for the PROGRESS-CTO model
- Both models showed low overall prediction error (Brier score)
Original publication
Comparative validation of risk scores for in-hospital complications following chronic total occlusion percutaneous coronary intervention. Am Heart J. Published 19 March 2026. doi:10.1016/j.ahj.2026.107427
Summarised from the published abstract; every number is checked against it