Skip to main content
. 2023 Jun 26;23(3):156–161. doi: 10.4103/tjem.tjem_79_23

Table 1.

Predictive performance of chat generative pretraining transformer for each emergency severity index triage category

Triage category Percentage (95% CI)

Sensitivity Specificity PPV NPV
ESI 1 88.9 (51.8–99.7) 95.1 (83.5–99.4) 80 (44.4–97.5) 97.5 (86.8–99.9)
ESI 2 41.7 (15.2–72.3) 92.1 (78.6–98.3) 62.5 (24.5–91.5) 83.3 (68.6–93)
ESI 3 50 (18.7–81.3) 80 (64.4–90.9) 38.5 (13.9–68.4) 86.5 (71.2–95.5)
ESI 4 50 (11.8–88.2) 70.5 (54.8–83.2) 18.8 (4–45.6) 91.2 (76.3–98.1)
ESI 5 15.4 (1.9–45.5) 97.3 (85.8–99.9) 66.7 (9.4–99.2) 76.6 (62–87.7)

ESI: Emergency severity index, PPV: Positive predictive value, NPV: Negative predictive value, CI: Confidence interval