Choosing endpoints in clinical studies and trials

Currently, it is challenging to select endpoints for measuring disease progression in glaucoma clinical trials. The endpoints that are used are not biological, but statistical, events because there is no reference standard for progression. Progression rates depend on the criteria used to define change, and in the absence of a reference standard for progression, reasonable criteria can be determined by paying attention to specificity (ie false-positive rate). Past randomized-clinical trials (CITGS, OHTS, and CNGT) have not made a accurate estimate of specificity at the time of trial design. It is not meaningful to compare the sensitivities (or more accurately the 'hit rate') of different progression criteria unless their respective specificities are equivalent. The ideal approach to compare progression rates, therefore, would be to equalize the specificities with each approach. However, owing to the lack of an external standard, this is difficult to achieve with reasonable confidence. With the advancement of research in visual field and imaging, clinical endpoints will have to be reviewed to incorporate new criteria for progression detection and measurement.