Table 2.
Change in visual acuity over 54 months by implant vs. systemic treatment group in the Multicenter Uveitis Steroid Treatment (MUST) Trial and Follow-up Study.
| Estimated Mean (SE)† | Estimated Mean Change from BL (SE)† | Estimated Treatment Effect (95% CI) †§ | |||||
|---|---|---|---|---|---|---|---|
| Sample size* | Implant | Systemic | Implant | Systemic | Implant - Systemic | P value | |
| Visual acuity (uveitic eyes) | |||||||
| Enrollment | 475 | 61.5 (2.4) | 64.8 (2.5) | n/a | n/a | n/a | |
| 12 months | 440 | 65.9 (2.4) | 68.1 (2.6) | 4.39 (1.37) | 3.35 (1.27) | 1.04 (−2.24, 4.61) | 0.569 |
| 24 months | 435 | 67.4 (2.4) | 68.0 (2.6) | 5.94 (1.38) | 3.21 (1.45) | 2.73 (−0.98, 6.70) | 0.165 |
| 36 months | 409 | 66.6 (2.4) | 69.1 (2.5) | 5.13 (1.46) | 4.31 (1.36) | 0.82 (−3.05, 5.02) | 0.675 |
| 48 months | 390 | 65.0 (2.5) | 67.7 (2.5) | 3.52 (1.51) | 2.91 (1.63) | 0.61 (−3.55, 5.17) | 0.781 |
| 54 months | 369 | 63.9 (2.5) | 67.9 (2.5) | 2.40 (1.40) | 3.12 (1.61) | −0.72 (−5.01, 3.54) | 0.737 |
| Visual acuity (better eye) | |||||||
| Enrollment | 254 | 71.6 (2.3) | 75.8 (2.0) | n/a | n/a | n/a | |
| 12 months | 234 | 74.8 (2.0) | 77.4 (2.3) | 3.13 (1.28) | 1.66 (1.07) | 1.47 (−1.80, 4.74) | 0.378 |
| 24 months | 232 | 75.5 (2.1) | 77.9 (2.2) | 3.83 (1.31) | 2.08 (1.13) | 1.75 (−1.64, 5.14) | 0.311 |
| 36 months | 217 | 74.9 (2.1) | 78.0 (2.2) | 3.25 (1.25) | 2.20 (1.19) | 1.05 (−2.34, 4.43) | 0.544 |
| 48 months | 208 | 74.1 (2.2) | 77.6 (2.2) | 2.44 (1.41) | 1.80 (1.35) | 0.64 (−3.19, 4.46) | 0.744 |
| 54 months | 197 | 73.8 (2.3) | 77.1 (2.3) | 2.20 (1.20) | 1.31 (1.50) | 0.89 (−2.88, 4.66) | 0.644 |
Number of eyes or individuals with data available at each visit.
Parameter estimates are calculated using generalized estimating equations assuming working independence between eyes and a Toeplitz covariance structure for the within-eye replicate measurements with robust standard errors. For the analysis of all uveitic eyes, bootstrap retaining the original treatment assignments with 1000 replicates is used to calculate the standard errors, confidence intervals, and p-values for the parameter estimates in order to adjust for between-eye correlation.
The treatment effect is defined as the between the two treatment groups (Implant – Systemic) at the enrollment visit and is defined as the difference in the change from baseline between the two treatment groups (Implant – Systemic) for all follow-up visits. Positive numbers favor implant treament; negative numbers favor systemic treatment.
CI = 95% confidence interval; SE = standard error.