The a priori statistical plan for the study included a subgroup analysis for patients whose baseline IOP (at 09 00) on latanoprost was ⩾20 mm Hg.
The present study was limited because tests of within-group changes from baseline IOP were not included in the a priori statistical plan for the study.
The statistical plan for the trial uses a design that evaluates the HR of febuxostat to allopurinol based on the primary CV composite end point when there are a maximum of 624 CV events.
The statistical plan for the trial uses a design that evaluates the hazard ratio (HR) of alogliptin to placebo first based on the primary CV composite end point after accrual of 80 to 150 primary CV events and again when there are 550 to 650 primary CV events.
RL, EvT, PB developed the statistical plan for this protocol.
AJ assisted in the development of the statistical plan for this protocol.
DS Developed the statistical plan for the manuscript, directed the statistical analysis and interpretation of the data, and critically revised the statistical portion of the manuscript.
We assessed differences in the statistical plan for model adjustment between the trial protocol and the trial publication for those trials with available protocols.
Furthermore, our statistical plan for model selection, which included the use of MAEs, RMSEs and tests for heteroskedasticity was in keeping with recommended diagnostic approaches [ 22].
Efficacy subgroup analysis based on any prior treatment, prior laser treatment, and no prior laser treatment was included in the statistical plan for the study.
