5-year MACE prediction model performance with or without blood test
| Method | Data used | Blood test required | AUC (95% CI) |
| Proposed models with blood test | |||
| SCORE | Age, gender, LDL cholesterol, HDL cholesterol, triglycerides, systolic blood pressure and smoking status | Y | 0.682 (0.640 to 0.719) |
| SCORE 2 | Age, gender, total and HDL cholesterol, systolic blood pressure, diabetes statues and smoking status | Y | 0.714 (0.674 to 0.753) |
| PCE | Age, gender, total and HDL cholesterol, systolic blood pressure, diabetes statues and smoking status | Y | 0.695 (0.656 to 0.73) |
| Framingham Risk Score | Age, total and HDL cholesterol, systolic blood pressure, treatment for hypertension, smoking and diabetes status | Y | 0.689 (0.648 to 0.730) |
| Logistic regression | 10 CVD risk factors* | Y | 0.758 (0.729 to 0.784) |
| Neural network | Y | 0.763 (0.734 to 0.790) | |
| XMACE+ | 10 CVD risk factors*+end-to-end retinal image DL model | Y | 0.769 (0.742 to 0.795) |
| Proposed models without blood test | |||
| Retinal image | End-to-end retinal image DL model following Poplin et al10 | N | 0.662 (0.632 to 0.694) |
| XMACE | Estimated 10 CVD risk factors* from the image | N | 0.738 (0.710 to 0.766) |
*10 cardiovascular disease (CVD) risk factors: Age, gender, BMI, systolic and diastolic blood pressure, HbA1c, HDL cholesterol, LDL cholesterol, triglycerides and smoking habits. The logistic regression and neural network used the same data source.
AUC, area under the receiver operating characteristic curve; BMI, body mass index; DL, deep learning; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MACE, major adverse cardiovascular events; PCE, Pooled Cohort Equations; SCORE, Systematic COronary Risk Evaluation.