Abstract:
|
For ophthalmic devices intended to provide improvement in uncorrected (without glasses or contact lenses) near visual acuity, uncorrected distance visual acuity may be used to assess risk, since subjects may be giving up distance vision for some gain in near vision. In this situation, one joint analysis of the benefit and risk of the device may be the assessment of how much distance vision the subjects give up (risk) for how much improvement in their near vision (benefit). We propose a method for assessing this benefit and risk that involves calculation of a ratio. A limitation is that ratios are commonly reported in the literature as a point estimates without reporting of the confidence intervals around these estimates. We compare the performance of the Delta, Fieller, and Bootstrap methods of constructing confidence intervals for our proposed ratio. A simulation is carried out to assess the performance of each of the three approaches.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.