Summary of test performance for predicting three or more unplanned admissions within 1 year of discharge for different case finding algorithms
| Method | Country | Model | Data | Total separations (patients) | Test separations (patients) | Target: 3 or more unplanned in 12 months | ||
| AUROC (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | ||||||
| HURT | Australia | XGBoost | IP, ED, OP, ABS | 206 714 (125 743) | 34 801 (27 216) | 84.2% (83.4 to 84.9) | 57.2% (55.4 to 59.0) | 89.4% (89.4 to 90.0) |
| HLCC @ NH | Australia | Weighted score | IP, ED | 206 714 (125 743) | 34 801 (27 216) | 70.5% (69.4 to 71.8) | 48.1% (46.3 to 50.0) | 87.8% (87.5 to 88.2) |
| HLCC Victoria17 | Australia | Weighted score | IP, ED | N/A (N/A) | N/A (N/A) | N/A (N/A) | 41%* (N/A) | N/A (N/A) |
*Recall was 78%.
ABS, Australian Bureau of Statistics; AUROC, Area Under the Reciever Operator Curve; ED, emergency department; HLCC, HealthLinks Chronic Care; HURT, Hospital Unplanned Readmission Tool; IP, inpatient; N/A, not available; OP, outpatient.