Table 3 Diagnostic performance of four models in the evaluation of mild or moderate renal function impairment
Training cohort | Validation cohort | |||||||
|---|---|---|---|---|---|---|---|---|
Models | AUC | Sensitivity (%) | Specificity (%) | Accuracy (%) | AUC | Sensitivity (%) | Specificity (%) | Accuracy (%) |
All-K | 0.888 (0.881–0.895) | 78.9 (78.2–79.2) | 89.0 (88.4–89.6) | 85.0 (84.4–85.6) | 0.820 (0.817–0.823) | 65.0 (63.8–66.2) | 74.2 (73.9–74.5) | 75.0 (74.5–75.5) |
LC-K | 0.919 (0.916–0.922) | 83.0 (81.8–84.1) | 94.5 (93.2–95.8) | 88.3 (87.5–89.2) | 0.852 (0.846–0.857) | 65.0 (64.4–65.6) | 84.4 (82.8–86.0) | 76.9 (76.3–77.5) |
TLCO-K | 0.926 (0.919–0.932) | 80.8 (80.3–81.4) | 91.8 (89.9–93.6) | 87.5 (86.9–88.1) | 0.705 (0.691–0.720) | 75.0 (73.6–76.3) | 59.4 (58.7–60.0) | 65.4 (64.8–66.0) |
PIZZA-K | 0.867 (0.860–0.874) | 63.8 (63.3–64.3) | 93.1 (93.1–93.2) | 81.7 (80.9–82.4) | 0.705 (0.690–0.721) | 25.0 (24.4–25.5) | 87.5 (87.0–87.9) | 63.5 (62.9–64.0) |