Nonparametric inference for paired quality of
life adjusted time-to-event data
Susan Murray
University of Michigan
Department of Biostatistics
Wednesday, November 8, 2006
140 Bardeen
| ABSTRACT |
This research makes available a nonparametric quality of life adjusted survival method for testing differences in paired censored time to event data. This statistic is based on integrated quality-adjusted survival curves with the standardized test appropriately adjusted for correlation in the estimated curves. Quality of life adjusted survival analysis is desirable when there is interest not only in the overall time to event but also in the quality of that time. The asymptotic distribution of the test statistic is given along with variance estimation formulae. We conduct simulations to study finite sample properties of the proposed statistic under both null and alternative hypotheses. We apply this method to a diabetic retinopathy study comparing alternate treatments in paired eyes.
(This is joint work with PhD candidate Kristi Cooper)
Coffee and Cookies at 3:30 p.m. in Room 1210 MSC
|
Return to seminar list
