Table 1.
Area under the receiver operating characteristic curves of models predicting hypotension within 6 h.
| Models | Outcomes | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| MAP Δ20 | P* | P† | P‡ | P§ | MAP Δ30 | P* | P† | P‡ | P§ | |
| SOFA | 0.500 (0.453–0.547) | 0.496 (0.435–0.557) | ||||||||
| APACHE II | 0.546 (0.499–0.593) | 0.592 (0.535–0.649) | ||||||||
| MOSAIC | 0.568 (0.522–0.615) | 0.578 (0.518–0.638) | ||||||||
| LR | 0.809 (0.774–0.844) | 0.824 (0.775–0.873) | ||||||||
| SVM | 0.807 (0.772–0.842) | < 0.001 | < 0.001 | < 0.001 | 0.686 | 0.830 (0.784–0.876) | < 0.001 | < 0.001 | < 0.001 | 0.536 |
| DNN | 0.822 (0.789–0.856) | < 0.001 | < 0.001 | < 0.001 | 0.601 | 0.835 (0.789–0.881) | < 0.001 | < 0.001 | < 0.001 | 0.755 |
| LGBM | 0.813 (0.780–0.847) | < 0.001 | < 0.001 | < 0.001 | 0.768 | 0.845 (0.802–0.888) | < 0.001 | < 0.001 | < 0.001 | 0.216 |
| XGB | 0.828 (0.796–0.861) | < 0.001 | < 0.001 | < 0.001 | 0.440 | 0.861 (0.822–0.900) | < 0.001 | < 0.001 | < 0.001 | 0.253 |
*Compared with the APACHE II model.
†Compared with the SOFA model.
‡Compared with the MOSAIC model.
§Compared with the LR model.
MAP mean arterial pressure, MAP Δ20 reduction in MAP ≥ 20 mmHg from the initial value, MAP Δ30 reduction in MAP ≥ 30 mmHg from the initial value, SOFA Sequential Organ Failure Assessment, APACHE Acute Physiology and Chronic Health Evaluation, MOSAC Mortality Scoring system for AKI with CRRT, LR Logistic regression, SVM support vector machine, DNN deep neural network, LGBM light gradient boosting machine, XGB extreme gradient boosting.