Table 1:
Performance Characteristics of the Diagnostic Predictive Models used to Create Probabilistic Gold Standard Datasets
| Phenotype | Database | AUC | Calibration Intercept |
Calibration Slope |
Average Predicted Probability Case |
Median Predicted Probability Case |
Average Predicted Probability Non-Case |
Median Predicted Probability Non-Case |
|---|---|---|---|---|---|---|---|---|
| Chronic Kidney Disease | CCAE | 0.997 | 0 | 1.00 | 0.87 | 1.00 | 0.00 | 0.00 |
| Optum1862 | 0.996 | 0 | 1.05 | 0.84 | 1.00 | 0.00 | 0.00 | |
| OptumGE66 | 0.977 | 0 | 1.02 | 0.79 | 0.97 | 0.04 | 0.01 | |
| MDCD | 0.997 | 0 | 1.02 | 0.88 | 1.00 | 0.00 | 0.00 | |
| MDCR | 0.988 | 0 | 1.01 | 0.85 | 1.00 | 0.03 | 0.00 | |
| PanTher1862 | 0.993 | 0 | 1.02 | 0.81 | 0.98 | 0.01 | 0.00 | |
| PanTherGE66 | 0.982 | 0 | 1.01 | 0.78 | 0.94 | 0.03 | 0.00 | |
| Atrial Fibrillation | CCAE | 0.999 | 0 | 0.98 | 0.86 | 0.98 | 0.00 | 0.00 |
| Optum1862 | 0.999 | 0 | 0.97 | 0.87 | 0.97 | 0.00 | 0.00 | |
| OptumGE66 | 0.992 | 0 | 1.00 | 0.84 | 0.94 | 0.02 | 0.00 | |
| MDCD | 0.996 | 0 | 0.98 | 0.78 | 0.91 | 0.00 | 0.00 | |
| MDCR | 0.996 | 0 | 1.01 | 0.88 | 0.99 | 0.02 | 0.00 | |
| PanTher1862 | 0.998 | 0 | 1.01 | 0.84 | 0.96 | 0.00 | 0.00 | |
| PanTherGE66 | 0.994 | 0 | 1.02 | 0.84 | 0.95 | 0.03 | 0.00 | |
| Myocardial Infarction | CCAE | 1.000 | 0 | 1.00 | 0.84 | 0.97 | 0.00 | 0.00 |
| Optum1862 | 1.000 | 0 | 0.98 | 0.86 | 0.97 | 0.00 | 0.00 | |
| OptumGE66 | 0.994 | 0 | 1.05 | 0.74 | 0.91 | 0.01 | 0.00 | |
| MDCD | 0.998 | 0 | 1.01 | 0.76 | 0.92 | 0.00 | 0.00 | |
| MDCR | 0.994 | 0 | 1.04 | 0.77 | 0.94 | 0.01 | 0.00 | |
| PanTher1862 | 0.998 | 0 | 1.03 | 0.76 | 0.94 | 0.00 | 0.00 | |
| PanTherGE66 | 0.984 | 0 | 1.02 | 0.68 | 0.83 | 0.02 | 0.00 | |
| Cerebral Infarction | CCAE | 1.000 | 0 | 1.01 | 0.90 | 1.00 | 0.00 | 0.00 |
| Optum1862 | 1.000 | 0 | 1.02 | 0.90 | 0.99 | 0.00 | 0.00 | |
| OptumGE66 | 0.999 | 0 | 1.04 | 0.87 | 0.99 | 0.01 | 0.00 | |
| MDCD | 0.999 | 0 | 0.99 | 0.88 | 1.00 | 0.00 | 0.00 | |
| MDCR | 0.998 | 0 | 1.00 | 0.91 | 1.00 | 0.01 | 0.00 | |
| PanTher1862 | 0.996 | 0 | 0.99 | 0.82 | 0.99 | 0.00 | 0.00 | |
| PanTherGE66 | 0.991 | 0 | 1.01 | 0.79 | 0.97 | 0.01 | 0.00 |
AUC – Area Under Receiver Operator Characteristics Curve; CCAE - IBM® MarketScan® Commercial Claims and Encounters Database, ages 18-62 years; MDCR - IBM® MarketScan® Medicare Supplemental and Coordination of Benefits Database, ages 66 years and greater; MDCD - IBM® MarketScan® Multi-State Medicaid, ages 18-62 years; Optum1862 - Optum© De-Identified Clinformatics® Data Mart Database – Date of Death, ages 18-62 years; OptumGE66 - ages 66 years and greater; PanTher1862 - Optum© de-identified Electronic Health Record Dataset, ages 18-62 years; PanTherGE66 - Optum© de-identified Electronic Health Record Dataset, ages 66 years and greater