Sensitivity, specificity and related analyses of M2 to predict the risk of sepsis in the development and external validation data sets at range of predicted probabilities cut-off points
| Predicted probabilities cut-off point | N+ | Sensitivity, % (95% CI) | Specificity, % (95% CI) | Positive predictive value (95% CI) | Negative predictive value (95% CI) | Positive likelihood ratio (95% CI) | Negative likelihood ratio (95% CI) |
|---|---|---|---|---|---|---|---|
| Development data set, YH | |||||||
| 0.05 | 9558 | 63.85 (61.43–66.21) | 75.04 (74.58–75.50) | 10.66 (10.05–11.30) | 97.80 (97.62–97.98) | 2.56 (2.45–2.67) | 0.48 (0.45–0.51) |
| 0.06* | 7337 | 55.20 (52.72–57.66) | 81.13 (80.71–81.54) | 12.01 (11.27–12.77) | 97.49 (97.30–97.67) | 2.93 (2.78–3.07) | 0.55 (0.52–0.58) |
| 0.07 | 5753 | 47.24 (44.77–49.73) | 85.39 (85.01–85.76) | 13.11(12.24–14.01) | 97.20 (97.01–97.38) | 3.23 (3.05–3.43) | 0.62 (0.59–0.65) |
| 0.08 | 4639 | 41.79 (39.36–44.26) | 88.39 (88.05–88.73) | 14.38 (13.38–15.42) | 97.02 (96.82–97.21) | 3.60 (3.37–3.84) | 0.66 (0.63–0.69) |
| 0.09 | 3834 | 36.97 (34.59–39.39) | 90.52 (90.20–90.83) | 15.39 (14.26–16.57) | 96.85 (96.66–97.04) | 3.90 (3.63–4.19) | 0.70 (0.67–0.72) |
| 0.10 | 3243 | 32.39 (30.10–34.75) | 92.03 (91.74–92.32) | 15.94 (14.70–17.25) | 96.69 (96.49–96.88) | 4.07 (3.75–4.40) | 0.73 (0.71–0.76) |
| 0.11 | 2677 | 28.13 (25.94–30.41) | 93.49 (93.22–93.75) | 16.77 (15.38–18.24) | 96.54 (96.34–96.73) | 4.32 (3.96–4.72) | 0.77 (0.75–0.79) |
| 0.12 | 2265 | 24.87 (22.77–27.07) | 94.54 (94.29–94.78) | 17.53 (15.98–19.16) | 96.43 (96.22–96.62) | 4.56 (4.14–5.01) | 0.79 (0.77–0.82) |
| 0.13 | 1913 | 21.68 (19.68–23.78) | 95.42 (95.19–95.64) | 18.09 (16.39–19.89) | 96.31 (96.11–96.51) | 4.73 (4.26–5.26) | 0.82 (0.80–0.84) |
| 0.14 | 1644 | 19.24 (17.33–21.26) | 96.09 (95.88–96.29) | 18.67 (16.82–20.64) | 96.23 (96.02–96.43) | 4.92 (4.39–5.51) | 0.84 (0.82–0.86) |
| 0.15 | 1407 | 16.92 (15.11–18.85) | 96.68 (96.48–96.86) | 19.19 (17.16–21.35) | 96.15 (95.94–96.35) | 5.09 (4.50–5.76) | 0.86 (0.84–0.88) |
| External validation data set, NH | |||||||
| 0.05 | 20 864 | 93.66 (92.73–94.51) | 43.84 (43.30–44.39) | 13.39 (12.93–13.86) | 98.68 (98.48–98.86) | 1.67 (1.65–1.69) | 0.14 (0.13–0.17) |
| 0.06 | 18 160 | 89.78 (88.63–90.84) | 51.89 (51.34–52.43) | 14.75 (14.23–15.27) | 98.21 (98.00–98.40) | 1.87 (1.84–1.90) | 0.20 (0.18–0.22) |
| 0.07 | 15 471 | 83.04 (81.64–84.37) | 59.62 (59.08–60.16) | 16.01 (15.44–16.60) | 97.43 (97.20–97.65) | 2.06 (2.01–2.10) | 0.28 (0.26–0.31) |
| 0.08 | 13 331 | 77.98 (76.44–79.45) | 65.80 (65.28–66.32) | 17.45 (16.81–18.10) | 96.99 (96.76–97.21) | 2.28 (2.23–2.34) | 0.33 (0.31–0.36) |
| 0.09 | 11 477 | 71.61 (69.95–73.22) | 70.97 (70.47–71.47) | 18.61 (17.90–19.34) | 96.42 (96.18–96.66) | 2.47 (2.40–2.54) | 0.40 (0.38–0.42) |
| 0.10 | 9956 | 65.87 (64.14–67.58) | 75.17 (74.69–75.64) | 19.74 (18.96–20.53) | 95.96 (95.71–96.20) | 2.65 (2.57–2.74) | 0.45 (0.43–0.48) |
| 0.11 | 8644 | 60.41 (58.63–62.17) | 78.74 (78.29–79.18) | 20.85 (19.99–21.72) | 95.55 (95.29–95.79) | 2.84 (2.74–2.94) | 0.50 (0.48–0.53) |
| 0.12* | 7540 | 55.88 (54.08–57.68) | 81.75 (81.32–82.17) | 22.11 (21.18–23.06) | 95.24 (94.98–95.48) | 3.06 (2.94–3.18) | 0.54 (0.52–0.56) |
| 0.13 | 6556 | 51.39 (49.58–53.20) | 84.39 (83.99–84.78) | 23.38 (22.36–24.43) | 94.93 (94.67–95.18) | 3.29 (3.15–3.44) | 0.58 (0.55–0.60) |
| 0.14 | 5776 | 47.07 (45.26–48.88) | 86.41 (86.03–86.79) | 24.31 (23.21–25.44) | 94.63 (94.36–94.88) | 3.46 (3.31–3.63) | 0.61 (0.59–0.63) |
| 0.15 | 5086 | 42.74 (40.96–44.54) | 88.16 (87.80–88.51) | 25.07 (23.88–26.28) | 94.32 (94.05–94.58) | 3.61 (3.43–3.80) | 0.65 (0.63–0.67) |
Note: N+ = the number of positive cases, NH = NLAG Hospitals, NLAG = Northern Lincolnshire and Goole NHS Foundation Trust, YH = York Hospital.
↵* Tentatively indicates a reasonable cut-off choice.