Table 4. Change in Visual Acuity at 2 Year Primary Outcome*.
| Change in Visual Acuity (letters) | Laser N=330 |
1 mg N=256 |
4 mg N=254 |
|---|---|---|---|
| Mean ± Standard Deviation (SD) | 1 ± 17 | -2 ± 18 | -3 ± 22 |
| Median (25th, 75th percentile) | 4 (-6, 11) | 1 (-11, 9) | 2 (-11, 11) |
| Distribution of Change at 2 Years - (%) | |||
| ≥ 15 letter improvement | 18% | 14% | 17% |
| 14-10 letter improvement | 13% | 11% | 11% |
| 9-5 letter improvement | 16% | 14% | 15% |
| Same ± 4 letters | 24% | 27% | 23% |
| 5-9 letters worse | 10% | 9% | 6% |
| 10-14 letters worse | 5% | 6% | 8% |
| ≥ 15 letters worse | 14% | 20% | 20% |
Visits occurring between 609 and 852 days from randomization were included as 2-year visits. When more than one visit occurred in this window, data from the visit closest to the two-year target date were used. For other eyes without any 2-year data (58 eyes in the laser group, 36 eyes in the 1 mg triamcinolone group, and 50 eyes in the 4 mg triamcinolone group), the last-observation-carried forward method was used to impute data for the primary analysis. Refraction was performed prior to the best corrected visual acuity measurement used in this analysis in 89%, 86%, 88% of eyes in the laser, 1 mg triamcinolone, and 4 mg triamcinolone groups, respectively. For other eyes, a refraction from the prior visit was used.
P values for two-group comparisons of difference in mean change: laser v. 1 mg = 0.02, laser v. 4 mg = 0.002, 1 mg v. 4 mg = 0.49
P values for two-group comparisons of proportion with 15 or more letter worsening: laser v. 1 mg = 0.03, laser v. 4 mg = 0.01, 1 mg v. 4 mg = 0.82
Hochberg's procedure was used to determine statistical significance by the following prespecified plan. After ordering the p-values of the 3 comparisons from highest to lowest (p1 > p2 > p3), statistical significance is determined as follows: If p1 ≤ 0.049 (includes an adjustment of 0.001 to account for DSMC data reviews), all comparisons are considered statistically significant; otherwise, if p2 ≤ (0.049/2) = 0.0245 then p2 and p3 are considered statistically significant; otherwise, if p3 ≤ (0.049/3) = 0.0163 then it will be considered statistically significant.