Predicting left main stenosis in stable ischemic heart disease using logistic regression and boosted trees
Authors' conclusion
In this large population-based study of patients with stable ischemic heart disease using extensive clinical data, only modest prediction of left main coronary artery disease was possible with traditional and machine learning modelling techniques.
Study context
- Population
- Adults with stable ischaemic heart disease who had stress testing before elective coronary angiography, Ontario, Canada, April 2010–March 20199,225 (6.1%) with left main stenosis
- Participants
- 150,423 adults
- Exposure
- Logistic regression vs boosted trees
- Outcome
- Left main coronary artery stenosis ≥50%
Methodological detailspopulation-based cohort
- Population-based cohort
Main finding
As reported in the abstract
Prediction of left main stenosis ≥50%
| Group | Optimism-adjusted c-statistic | C-statistic | Optimism-adjusted Integrated Calibration Index | Integrated Calibration Index |
|---|---|---|---|---|
| Logistic regression | 0.72 | — not stated in the abstract | 0.0044 | — not stated in the abstract |
| Boosted tree | — not stated in the abstract | 0.72 | — not stated in the abstract | 0.0054 |
Optimism-adjusted c-statistic
Logistic regression0.72
Boosted tree— not stated in the abstract
C-statistic
Logistic regression— not stated in the abstract
Boosted tree0.72
Optimism-adjusted Integrated Calibration Index
Logistic regression0.0044
Boosted tree— not stated in the abstract
Integrated Calibration Index
Logistic regression— not stated in the abstract
Boosted tree0.0054
— not stated in the abstract
Observational data: associations, not effects.
Further findings
Logistic regression performance by subgroup (sex, DM, extent of ischaemia)
- Results consistent in subgroups of males and females, DM and non-DM, and extent of ischaemia
Original publication
Predicting left main stenosis in stable ischemic heart disease using logistic regression and boosted trees. Am Heart J. Published 11 November 2022. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →