Table 5: Non-inferiority testing of postoperative complications without clinical marginals but including statistical marginals to allow for sampling variation. The upper limits of the two-sided 90% confidence intervals coincide with the upper limits of one-sided 95% confidence intervals for the risk difference study group vs. control. P-values below 0.05 establish non-inferiority of the study group.
Postoperative complication |
Reference incidence (%) |
Reference implied marginal (%) |
Upper 90% (%) |
p-valuea) |
0-3 months postoperatively |
|
|
|
|
Postoperative CME |
1.17 |
0.36 |
0.33 |
0.03* |
Persistent anterior uveitis |
1.75 |
0.44 |
0.18 |
0.01* |
Retinal Detachment |
0.25 |
0.17 |
0.11 |
0.01* |
Corneal decompensation |
1.00 |
0.33 |
0.05 |
< 0.001* |
Dry AMD progression to exudative AMD |
0.60 |
0.26 |
0.11 |
0.002* |
Malignant glaucoma |
0.60 |
|
|
|
0-12 months postoperatively |
|
|
|
|
Postoperative CME |
1.17 |
0.37 |
0.40 |
0.08 |
Persistent anterior uveitis |
1.75 |
0.45 |
0.24 |
0.001* |
Retinal Detachment |
0.43 |
0.22 |
0.17 |
0.01* |
Corneal decompensation |
1.00 |
0.34 |
0.04 |
< 0.001* |
Dry AMD progression to exudative AMD |
2.40 |
0.52 |
0.48 |
0.03* |
Malignant glaucoma |
0.06 |
|
|
|
a) Farrington-Manning method, AMD: Age Related Macular Degeneration; CME: Cystic Macular Edema, *= statistically significant