Table 1 The required number of indirect comparison trials required to produce the same precision as a given number of direct (head-to-head) trials
From: Sample size and power considerations in network meta-analysis
Trial count ratio | Exact precision ratio | Number of indirect comparison trials required to match precision from the corresponding number of single trials in a pair wise meta-analysis | |||||
|---|---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 10 | ||
1:1 | 4 | 4 (2:2) | 8 (4:4) | 12 (6:6) | 16 (8:8) | 20 (10:10) | 40 (20:20) |
1:2 | 4.5 | 6 (2:4) | 9 (3:6) | 15 (5:10) | 18 (6:12) | 24 (8:16) | 45 (15:30) |
1:3 | 5.33 | 8 (2:6) | 12 (3:9) | 16 (4:12 ) | 24 (6:18) | 28 (7:21) | 56 (14:42) |
1:4 | 6.25 | 10 (2:8) | 15 (3:12) | 20 (4:16) | 25 (5:20) | 35 (7:28) | 65 (13:52) |
1:5 | 7.2 | 12 (2:10) | 18 (3:15) | 24 (4:20) | 30 (5:25) | 36 (6:30) | 72 (12:60) |
1:6 | 8.17 | 14 (2:12) | 21 (3:18) | 28 (4:24) | 35 (5:30) | 42 (6:36) | 84 (12:72) |
1:7 | 9.14 | 16 (2:14) | 24 (3:21) | 32 (4:28) | 40 (5:35) | 48 (6:42) | 96 (12:84) |
1:8 | 10.13 | 18 (2:16) | 27 (3:24) | 36 (4:32) | 45 (5:40) | 54 (6:48) | 108 (12:96) |
1:9 | 11.11 | 20 (2:18) | 30 (3:27) | 40 (4:36) | 50 (5:45) | 60 (6:54) | 120 (12:108) |
1:10 | 12.1 | 22 (2:20) | 33 (3:30) | 44 (4:40) | 55 (5:50) | 66 (6:60) | 121 (11:110) |