Table 1.
Statistical power analysis
| Frequency of antigenic allele | GVHD |
|
|---|---|---|
| 50% vs 55% | 50% vs 60% | |
| 0.1 | ||
| Unrelated donor | 7642 | 1891 |
| Sibling donor | 13 330 | 3296 |
| 0.2 | ||
| Unrelated donor | 5326 | 1319 |
| Sibling donor | 8362 | 2069 |
| 0.3 | ||
| Unrelated donor | 4723 | 1169 |
| Sibling donor | 7490 | 1853 |
| 0.4 | ||
| Unrelated donor | 4519 | 1119 |
| Sibling donor | 6296 | 1558 |
| 0.5 | ||
| Unrelated donor | 4469 | 1106 |
| Sibling donor | 6128 | 1517 |
Statistical power analysis to estimate the number of transplant donor/recipient pairs that would need to be studied to provide 90% power to detect a hypothetical 10% (50%-55%) or 20% (50%-60%) increase in the risk of developing acute GVHD attributable to donor/recipient disparity at a specific minor H locus, as a function of (1) the frequency of the antigenic allele in the population from which the donor and recipient are derived, and (2) the relationship of the donor to the recipient (unrelated vs full sibling). It is assumed for this analysis that the donor/recipient pairs are matched for HLA-A, -B, -C, and -DRB1, and that they all express the MHC-restricting allele for the minor H antigen. The requisite number of pairs when the frequency, f, of the antigenic allele is > 0.5 is the same as the requisite number when the frequency of the antigenic allele is 1 − f.