Table 4 Model performances
From: Prognostic models for complete recovery in ischemic stroke: a systematic review and meta-analysis
No. of model | Calibration | Discrimination; AUC (95%CI) | |||
|---|---|---|---|---|---|
Internal validation | External validation | Development model | Internal validation | External validation | |
1 | – | – | – | 0.73 (0.66 to 0.80) | – |
2 | – | – | – | 0.80 (0.74 to 0.86) | – |
3 | – | closely resembling perfect calibration | – | 0.84 (0.79 to 0.89) | 0.83 (0.77 to 0.89) |
4 | – | – | – | 0.74 (0.67 to 0.81) | – |
5 | – | – | – | 0.79 (0.73 to 0.85) | – |
6 | – | closely resembling perfect calibration | – | 0.84 (0.79 to 0.89) | 0.81 (0.75 to 0.87) |
7 | – | – | – | 0.87 (0.82 to 0.92) | – |
8 | – | – | – | 0.70 (0.63 to 0.77) | – |
9 | closely resembling perfect calibration, mean absolute error = 0.01 | – | – | 0.87 (0.82 to 0.92) | – |
10 | – | – | – | 0.89 (0.84 to 0.94) | – |
11 | – | – | – | 0.72 (0.65 to 0.79) | – |
12 | – | – | – | 0.89 (0.84 to 0.94) | – |
13 | – | closely resembling perfect calibration, mean absolute error = 0.33 | 0.80 (0.75 to 0.85) | – | 0.82 (0.76 to 0.88) |
14 | – | mean absolute errors = 0.4, less well calibrated | 0.79 (0.74 to 0.84) | – | 0.80 (0.73 to 0.87) |
15 | – | closely resembling perfect calibration, mean absolute error = 0.37 | 0.79 (0.73 to 0.85) | – | 0.80 (0.74 to 0.86) |
16 | – | mean absolute errors = 0.4, less well calibrated | 0.78 (0.71 to 0.85) | – | 0.76 (0.69 to 0.83) |
17 | – | – | 0.80 (0.78 to 0.82) | – | 0.78 (0.75 to 0.81) |
18 | – | – | 0.86 (0.84 to 0.88) | – | 0.74 (0.71 to 0.77) |
| Â | 0.81 (0.80 to 0.82) | ||||
19 | – | – | 0.84 (0.75 to 0.94) | – | – |
20 | – | – | 0.87 (0.79 to 0.95) | – | – |
21 | – | – | 0.74 (0.69 to 0.79) | – | – |
22 | – | – | 0.82 | – | – |
23 | – | – | 0.70 | – | – |
Median AUC (95% CI) | 0.80 (0.77 to 0.85) | 0.82 (0.73 to 0.87) | 0.80 (0.76 to 0.82) | ||