Table 1:
Illustration of simulating DLT outcomes to generate operating characteristics of the 3+3 algorithm.
| Patient | Dose level | Random # x | Assumed DLT probability | DLT (yes/no) |
|---|---|---|---|---|
| Simulated trial #1 using 3+3 | ||||
| 1 | 1 | 0.88 | 0.05 | No |
| 2 | 1 | 0.56 | 0.05 | No |
| 3 | 1 | 0.18 | 0.05 | No |
| 4 | 2 | 0.64 | 0.10 | No |
| 5 | 2 | 0.53 | 0.10 | No |
| 6 | 2 | 0.49 | 0.10 | No |
| 7 | 3 | 0.66 | 0.25 | No |
| 8 | 3 | 0.14 | 0.25 | Yes |
| 9 | 3 | 0.53 | 0.25 | No |
| 10 | 3 | 0.67 | 0.25 | No |
| 11 | 3 | 0.04 | 0.25 | Yes |
| 12 | 3 | 0.55 | 0.25 | No |
| MTD = dose level 2 | ||||
| Simulated trial #2 using 3+3 | ||||
| 1 | 1 | 0.87 | 0.05 | No |
| 2 | 1 | 0.77 | 0.05 | No |
| 3 | 1 | 0.01 | 0.05 | Yes |
| 4 | 1 | 0.33 | 0.05 | No |
| 5 | 1 | 0.61 | 0.05 | No |
| 6 | 1 | 0.29 | 0.05 | No |
| 7 | 2 | 0.06 | 0.10 | Yes |
| 8 | 2 | 0.35 | 0.10 | No |
| 9 | 2 | 0.02 | 0.10 | Yes |
| MTD = dose level 1 | ||||