Table 2.
Endpoints under various designs and schemes: CRM, R-CRM, E-CRM assume a fixed sample size at n = 20. Varying sample scheme follows the stopping rule in section ‘Methods’ under subsection ‘CRM with stopping rule’. G-CRM has a minimum sample of 18. IQR: Interquartile range; s.d.: Standard deviation
| Fixed sample | Varying sample with stopping rule | |||||||
|---|---|---|---|---|---|---|---|---|
| CRM | R-CRM | E-CRM | CRM | R-CRM | E-CRM | G-CRM | SM | |
| 1. Accuracy: Percent of trials that found the true MTD | ||||||||
| S1 | 63 | 63 | 62 | 65 | 66 | 66 | 61 | 39 |
| S2 | 67 | 69 | 69 | 72 | 71 | 70 | 69 | 33 |
| S3 | 57 | 59 | 60 | 58 | 57 | 57 | 63 | 26 |
| S4 | 61 | 66 | 63 | 68 | 69 | 68 | 65 | 32 |
| S5 | 55 | 57 | 56 | 70 | 69 | 70 | 56 | 23 |
| S6 | 20 | 17 | 23 | 25 | 28 | 26 | 23 | 19 |
| S7 | 48 | 50 | 50 | 64 | 61 | 65 | 48 | 41 |
| S8 | 43 | 42 | 41 | 49 | 48 | 47 | 47 | 26 |
| 2. Optimal allocation: Percent of patients treated at the MTD (s.d.) | ||||||||
| S1 | 44 (30) | 44 (30) | 38 (28) | 47 (29) | 47 (29) | 43 (27) | 20 (16) | 18 (14) |
| S2 | 55 (32) | 50 (30) | 45 (26) | 58 (30) | 53 (29) | 49 (27) | 23 (15) | 16 (13) |
| S3 | 54 (33) | 46 (30) | 41 (27) | 56 (32) | 48 (30) | 43 (28) | 21 (15) | 13 (13) |
| S4 | 46 (30) | 39 (25) | 28 (18) | 52 (29) | 49 (28) | 43 (25) | 13 (11) | 11 (9) |
| S5 | 33 (26) | 33 (26) | 25 (20) | 41 (26) | 43 (27) | 37 (23) | 13 (11) | 7 (9) |
| S6 | 18 (13) | 14 (11) | 13 (9) | 23 (17) | 22 (17) | 21 (16) | 10 (9) | 10 (9) |
| S7 | 36 (25) | 38 (25) | 38 (25) | 44 (29) | 45 (29) | 46 (29) | 36 (22) | 35 (15) |
| S8 | 31 (23) | 31 (21) | 35 (21) | 41 (27) | 40 (26) | 43 (27) | 22 (14) | 22 (10) |
| 3. Safety: Median number of toxicities (IQR) | ||||||||
| S1 | 3 (2–1) | 3 (2–4) | 3 (2–4) | 5 (3–6) | 5 (3–6) | 5 (3–6) | 2 (1–3) | 2 (1–3) |
| S2 | 4 (3–5) | 4 (3–5) | 4 (3–4) | 6 (4–7) | 6 (4–7) | 6 (4–7) | 3 (2–4) | 2 (2–3) |
| S3 | 5 (4–5) | 4 (3–5) | 4 (3–5) | 7 (5–8) | 7 (5–8) | 6 (5–7) | 3 (2–4) | 3 (2–3) |
| S4 | 4 (3–4) | 3 (2–4) | 3 (2–3) | 7 (4–10) | 7 (4–9) | 7 (4–9) | 2 (2–3) | 3 (2–3) |
| S5 | 3 (2–4) | 3 (2–4) | 3 (2–4) | 7 (3–10) | 7 (3–10) | 7 (3–10) | 3 (2–4) | 3 (2–4) |
| S6 | 4 (4–5) | 4 (3–4) | 3 (2–4) | 10 (6–12) | 10 (6–11) | 9 (5–11) | 3 (2–4) | 3 (2–4) |
| S7 | 6 (6–7) | 6 (5–7) | 6 (5–7) | 11 (8–14) | 10 (8–13) | 10 (8–13) | 5 (4–6) | 3 (2–4) |
| S8 | 6 (5–6) | 5 (4–6) | 5 (4–5) | 12 (9–14) | 11 (9–13) | 11 (9–13) | 3 (3–4) | 3 (2–4) |
| 4. Oscillation after the nth patient | Percent of trials where Dn ≠ Dn+1 | Percent of trials that used maximum n (Percent of trials that met CI criterion) | ||||||
| S1 | 13 | 13 | 13 | 67 (34) | 68 (33) | 67 (34) | NA | NA |
| S2 | 12 | 14 | 12 | 70 (31) | 71 (31) | 72 (31) | NA | NA |
| S3 | 16 | 16 | 17 | 77 (23) | 78 (22) | 80 (22) | NA | NA |
| S4 | 12 | 14 | 15 | 19 (83) | 21 (81) | 23 (79) | NA | NA |
| S5 | 20 | 20 | 21 | 29 (77) | 30 (74) | 30 (76) | NA | NA |
| S6 | 22 | 21 | 20 | 40 (64) | 50 (55) | 52 (53) | NA | NA |
| S7 | 16 | 16 | 16 | 22 (80) | 22 (79) | 23 (79) | NA | NA |
| S8 | 23 | 26 | 25 | 51 (53) | 54 (49) | 56 (48) | NA | NA |
| 5. Median number of patients to reach the MTD (IQR) | ||||||||
| S1 | 16 (9–19) | 16 (9–19) | 16 (7–19) | 20 (9–28) | 20 (9–27) | 21 (9–28) | 13 (13–18) | 16 (13–16) |
| S2 | 15 (6–19) | 14 (5–19) | 15 (7–19) | 19 (6–27) | 19 (5–28) | 19 (7–28) | 13 (13–18) | 16 (13–16) |
| S3 | 17 (8–20) | 16 (5–20) | 16 (7–20) | 23 (8–29) | 23 (8–29) | 24 (8–29) | 13 (13–18) | 16 (13–19) |
| S4 | 16 (5–19) | 15 (8–19) | 16 (12–20) | 24 (9–33) | 25 (10–35) | 25 (12–35) | 22 (22–24) | 25 (22–28) |
| S5 | 18 (10–20) | 18 (10–20) | 18 (12–20) | 32 (11–43) | 33 (11–42) | 33 (12–13) | 22 (22–24) | 25 (22–28) |
| S6 | 18 (11–20) | 18 (10–20) | 18 (12–20) | 34 (16–43) | 38 (20–45) | 36 (20–45) | 22 (21–24) | 25 (22–25) |
| S7 | 17 (13–20) | 17 (13–20) | 17 (13–20) | 25 (17–29) | 24 (17–29) | 24 (17–28) | 16 (10–18) | 10 (10–13) |
| S8 | 19 (15–20) | 19 (16–20) | 19 (16–20) | 32 (22–42) | 35 (24–44) | 34 (22–44) | 18 (16–21) | 16 (13–22) |