Basic Ingredients for Determining Clinical Benefit by Modeling
| Validity of change | Change | Relative frequency | Expected benefit |
| Correct | A → B | p1 | b1 |
| Correct | B → A | p2 | b2 |
| Incorrect | A → B | p3 | b3 |
| Incorrect | B → A | p4 | b4 |
Data show all 4 possible changes in decision between A and B, their relative frequencies in clinical population of interest, and their expected benefit.