Table 2 Effective heterogeneity-corrected sample sizes of indirect comparison scenarios with varying degrees of patient count ratios and heterogeneity in each comparison (A vs. C and B vs. C), but with fixed total sample size of 10,000
From: Sample size and power considerations in network meta-analysis
Patient ratio | Precision ratio | IAC 2 | IBC 2 | I2 adjusted precision ratio | Effective sample size |
|---|---|---|---|---|---|
5,000:5,000 (1:1) | 4.00 | 0% | 0% | 4.00 | 2,500 |
0% | 25% | 4.60 | 2,188 | ||
0% | 50% | 5.33 | 1,875 | ||
25% | 50% | 6.40 | 1,563 | ||
50% | 50% | 8.00 | 1,250 | ||
3,333:6,667 (1:2) | 4.50 | 0% | 0% | 4.50 | 2,222 |
0% | 25% | 5.40 | 1,852 | ||
0% | 50% | 6.75 | 1,481 | ||
25% | 0% | 4.91 | 2,037 | ||
25% | 25% | 6.00 | 1,667 | ||
25% | 50% | 7.72 | 1,296 | ||
50% | 0% | 5.40 | 1,852 | ||
50% | 25% | 6.75 | 1,482 | ||
50% | 50% | 9.00 | 1,111 | ||
2,500:7,500 (1:3) | 5.33 | 0% | 0% | 5.33 | 1,876 |
0% | 25% | 6.56 | 1,524 | ||
0% | 50% | 8.53 | 1,173 | ||
25% | 0% | 5.69 | 1,759 | ||
25% | 25% | 7.11 | 1,407 | ||
25% | 50% | 9.48 | 1,055 | ||
50% | 0% | 6.09 | 1,642 | ||
50% | 25% | 7.75 | 1,290 | ||
50% | 50% | 10.7 | 938 | ||
2,000:8,000 (1:4) | 6.25 | 0% | 0% | 6.25 | 1,600 |
0% | 25% | 7.81 | 1,280 | ||
0% | 50% | 10.4 | 960 | ||
25% | 0% | 6.58 | 1,520 | ||
25% | 25% | 8.33 | 1,200 | ||
25% | 50% | 11.4 | 880 | ||
50% | 0% | 6.94 | 1,440 | ||
50% | 25% | 8.93 | 1,120 | ||
50% | 50% | 12.5 | 800 | ||
1,667:8,333 (1:5) | 7.20 | 0% | 0% | 7.20 | 1,389 |
0% | 25% | 9.09 | 1,100 | ||
0% | 50% | 12.4 | 810 | ||
25% | 0% | 7.51 | 1,331 | ||
25% | 25% | 9.60 | 1,042 | ||
25% | 50% | 13.3 | 752 | ||
50% | 0% | 7.86 | 1,273 | ||
50% | 25% | 10.2 | 984 | ||
50% | 50% | 14.4 | 694 | ||
1,000:9,000 (1:9) | 11.11 | 0% | 0% | 11.1 | 900 |
0% | 25% | 14.3 | 698 | ||
0% | 50% | 20.2 | 495 | ||
25% | 0% | 11.4 | 878 | ||
25% | 25% | 14.8 | 675 | ||
25% | 50% | 21.1 | 473 | ||
50% | 0% | 11.7 | 855 | ||
50% | 25% | 15.3 | 653 | ||
50% | 50% | 22.2 | 450 |