Table 1.
Nine-row by three-column MONARCSi drug-event pair feature matrix (); each feature is noted as present (yes), absent (no), or unknown/not applicable (UNK/NA) by a safety professional evaluating a drug-event pair
| Feature(Row i,Column j) | Yes(1) | No(2) | UNK/NA(3) |
|---|---|---|---|
| Significant safety event(1) | F1,1 = + 1 | F1,2 = − 1 | F1,3 = 0 |
| Previous association(2) | F2,1 = + 1 | F2,2 = − 1 | F2,3 = 0 |
| Temporality(3) | F3,1 = + 1 | F3,2 = − 1 | F3,3 = 0 |
| Mechanism of action(4) | F4,1 = + 1 | F4,2 = − 1 | F4,3 = 0 |
| De-challenge(5) | F5,1 = + 1 | F5,2 = − 1 | F5,3 = 0 |
| Re-challenge(6) | F6,1 = + 1 | F6,2 = − 1 | F6,3 = 0 |
| Dose response(7) | F7,1 = + 1 | F7,2 = − 1 | F7,3 = 0 |
| Experimental data(8) | F8,1 = + 1 | F8,2 = − 1 | F8,3 = 0 |
| Confounding factors(9) | F9,1 = − 1 | F9,2 = + 1 | F9,3 = 0 |
Each feature is assigned a value for being present (+ 1), absent (− 1), or UNK/NA (0) based on the safety professional’s assessment of a specific drug-event pair narrative. Each feature element is multiplied by its corresponding element in the weighting matrix () and summed to create the aggregate score (see Table 3 and Eq. 1)