Comparing Two Kaplan-Meier Curves with the Probability of Agreement

Presentation Slides from Nathaniel Stevens Webinar 5/7/2020.

Abstract: The probability of agreement (PoA) has been used as an effective strategy for quantifying the similarity between the reliability of two populations. In contrast to the p-value approach associated with hypothesis testing, the PoA provides a more realistic assessment of similarity by accounting for a practically important difference. This talk discusses the adaptation of the PoA to the comparison of population reliabilities as estimated by Kaplan-Meier curves, when lifetime data is right censored. Three methods for quantifying uncertainty in the PoA estimate are explored: the first approach provides a convenient assessment based on large sample approximations, and the second two approaches offer nonparametric bootstrap-based alternatives. One of these relies on the traditional bootstrap while the other is based on the fractional random-weight bootstrap, which is necessitated by high censoring rates. All methods are illustrated with examples for which comparing the reliability curves of related populations is of interest.

File size
14.13 MB
Wednesday May 20th, 2020 4:44am

File Comments

No Data Available