Article Figures & Data
- Figure 1
Schematic overview of methods. CAS, carotid artery stenting; CEA, carotid endarterectomy; CREST, Carotid Revascularisation Stenting versus Endarterectomy; MI, myocardial infarction; SVS, Society for Vascular Surgery; VQI, Vascular Quality Initiative.
- Figure 2
Distribution of predicted net benefit for each individual in crest with treatment on Primary Stroke and MI Treatment Models (A) and the Sensitivity Stroke and Primary MI Treatment Models (B). The y-axis represents the net change in the risk of MI with CEA vs CAS (higher values indicate greater risk with CEA compared with CAS) calculated by taking the predicted probability with CEA on the Primary MI Treatment Model and subtracting the predicted probability with CAS. The x-axis represents the change in stroke risk with CEA versus CAS (higher values indicate greater risk with CAS compared with CEA), calculated by taking the predicted probability with CAS on the primary stroke (A) or sensitivity stroke (B) and subtracting the predicted probability with CEA on the same model. The red line has a slope of 1 and intercept of 0, representing the break-even point where stroke and MI risks are comparable. To the left of the line, the net benefit is greater with CAS and to the right, net benefit is greater with CEA. CAS, carotid artery stenting; CEA, carotid endarterectomy; MI, myocardial infarction.
- Figure 3
Calibration of baseline treatment models in SVS-VQI. (A) displays the predicted stroke risk from the CREST treatment model (red bars), the actual stroke risk measured in SVS-VQI (blue bars) and the estimated risk in a model that recalibrated the intercept of the CREST model (green bars) as well as a model that estimated both the slope and intercept of the CREST model in SVS-VQI (yellow bars) . (B) is analogous to (A), but the models are for MI prediction. CREST, Carotid Revascularisation Stenting versus Endarterectomy; MI, myocardial infarction; SVS, Society for Vascular Surgery; VQI, Vascular Quality Initiative.
- Table 1
Summary of regression coefficients for stroke/MI treatment models in crest
Stroke MI Regression coefficient P value Regression coefficient P value Baseline Risk Index 0.2944396 0.403 0.6315831 0.001 CAS −1.529699 0.441 −0.6956389 0.035 Symptomatic 0.6403783 0.084 −0.4983009 0.148 CEA-Age Interaction 0.0235438 0.297 CAS-Age Interaction 0.0542187 0.001 Constant −5.986339 0 −5.150785 0 SVS Recalibrated Constant −6.486095 −5.622736 CAS, carotid artery stenting; CEA, carotid endarterectomy; MI, myocardial infarction; SVS, Society for Vascular Surgery.
- Table 2
Predicted optimal event rates applying MI and stroke treatment models (with recalibrated slopes and intercepts) in the SVS-VQI dataset
Estimated stroke rate Estimated MI rate, % Total event rate, % Optimal treatment 1.31 0.73 2.04 As treated 1.39 0.73 2.12 All CEA 1.30 0.75 2.05 All CAS 2.05 0.58 2.63 CAS, carotid artery stenting; CEA, carotid endarterectomy; MI, myocardial infarction.
Supplementary Materials
Supplementary data
Supplementary data
Additional Files
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
