Table 5 Predictive ability of the scoring system to differentiate between non-advanced and advanced ALNM (pN2-N3) at each cutoff point in the training and validation cohorts
Total score cutoffa | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | AUC (95% CI) | P value |
|---|---|---|---|---|---|---|
Training Cohort | ||||||
≦ 2 | 100.0 | 41.3 | 23.1 | 100.0 | 0.71 (0.68–0.73) | < 0.0001 |
≦ 3 | 97.3 | 59.9 | 29.9 | 99.2 | 0.79 (0.76–0.82) | 0.0869 |
≦ 4 | 85.3 | 79.1 | 41.8 | 96.8 | 0.82 (0.78–0.87) | Reference |
≦ 5 | 60.0 | 92.3 | 57.7 | 92.9 | 0.76 (0.70–0.82) | 0.0218 |
≦ 6 | 52.0 | 96.0 | 69.6 | 91.9 | 0.74 (0.68–0.80) | 0.0044 |
≦ 7 | 22.7 | 99.3 | 85.0 | 87.9 | 0.61 (0.56–0.66) | < 0.0001 |
≦ 8 | 10.7 | 99.8 | 88.9 | 86.4 | 0.55 (0.52–0.59) | < 0.0001 |
Validation Cohort | ||||||
≦ 2 | 100.0 | 35.8 | 23.1 | 100.0 | 0.68 (0.65–0.71) | < 0.0001 |
≦ 3 | 98.0 | 55.1 | 29.6 | 99.3 | 0.76 (0.71–0.80) | 0.0366 |
≦ 4 | 87.8 | 74.8 | 40.2 | 96.9 | 0.81 (0.76–0.87) | Reference |
≦ 5 | 65.3 | 89.0 | 53.3 | 93.0 | 0.76 (0.69–0.83) | 0.1171 |
≦ 6 | 55.1 | 92.5 | 58.7 | 91.4 | 0.74 (0.67–0.81) | 0.00375 |
≦ 7 | 22.5 | 97.2 | 61.1 | 86.7 | 0.60 (0.54–0.66) | < 0.0001 |
≦ 8 | 10.2 | 99.2 | 71.4 | 85.1 | 0.55 (0.50–0.59) | < 0.0001 |