Table 3. Study endpoints (by treatment assignment).
| Selenium | Placebo | |
|---|---|---|
| Number randomized | 227 | 225 |
| Ineligible | 15 | 14 |
| Eligible | 212 | 211 |
|
| ||
| Endpoint status at 3 years: N (%) | 77 (36.3) | 77 (36.5) |
| EOS > 90 days out | 9 | 18 |
| Died | 4 | 6 |
| Non-PC primary cancer | 4 | 1 |
| Intercurrent illness | 5 | 3 |
| M.D. recommended no EOS biopsy | 6 | 5 |
| Declined EOS biopsy | 21 | 18 |
| Lost to follow-up | 13 | 12 |
| Withdrew consent | 4 | 5 |
| Other | 11 | 9 |
|
| ||
| Endpoint Known: N (%) | 135 (63.7) | 134 (63.5) |
| Prostate cancer: N (%)* | 48 (35.6) | 49 (36.6)* |
| Interim | 24 | 22 |
| EOS | 24 | 27 |
| Negative EOS biopsy: N (%) | 87 (64.4) | 85 (63.4) |
Abbreviation: EOS, end of study.
*
The P-value comparing selenium versus placebo patients is 0.73 for three-year biopsy-proven carcinoma, adjusted for the stratification factors age > 60, African American, baseline PSA, and vitamin E supplementation.