Table 8.
Test dataset performance metrics for MOdified NARanjo causality scale for ICSRs (MONARCSi) compared with clinical judgment using global introspection as the reference
| Performance metric | Value (%) |
|---|---|
| Sensitivity % Positive agreement |
64.7 |
| Specificity % Negative agreement |
93.4 |
| Positive predictive value (precision) Proportion of true ‘related’ out of all classified ‘related’ |
78.6 |
| Negative predictive value Proportion of true ‘unrelated’ out of all classified ‘unrelated’ |
87.6 |
| gKappa score Inter-rater agreement with ‘ground truth’ |
76.9 |
|
F score (F1) Harmonic mean between precision and recall |
71.0 |