Table 4 Diagnostic performance indices of ELF for identifying LRE using thresholds with sensitivity and specificity of 80, 85 and 90%. Data are presented showing ELF thresholds with sensitivity and specificity values of 80, 85 and 90%, with data-derived diagnostic accuracy (proportion of true positives and true negatives), misclassification rate (proportion of false negatives and false positives) and the rate of indeterminate cases (ELF score between the thresholds)
Sensitivity and specificity values | ELF thresholds | Sensitivity (%) | Specificity (%) | Diagnostic accuracy (%) | Misclassification rate (%) | Indeterminate (%) |
|---|---|---|---|---|---|---|
Recruitment samples | ||||||
 80% | 8.350 | 81.0 | 19.1 | 26.0 | 19.7 | 54.3 |
9.425 | 39.7 | 80.0 | ||||
 85% | 8.270 | 84.5 | 16.5 | 26.0 | 19.7 | 54.3 |
9.750 | 27.6 | 85.2 | ||||
 90% | 8.160 | 91.4 | 14.8 | 26.0 | 19.7 | 54.3 |
9.895 | 24.1 | 90.4 | ||||
Second subsequent samples | ||||||
 80% | 9.365 | 80.4 | 34.8 | 36.4 | 19.1 | 44.5 |
10.255 | 41.1 | 80.9 | ||||
 85% | 8.975 | 85.7 | 14.8 | 36.4 | 19.1 | 44.5 |
10.540 | 28.6 | 86.1 | ||||
 90% | 8.945 | 89.3 | 13.9 | 56.6 | 28.3 | 15.1 |
10.835 | 17.9 | 90.4 | ||||