+ Reply to Thread
Results 1 to 1 of 1

Thread: stratified log-rank test and proportional hazards

  1. #1
    Points: 5,407, Level: 47
    Level completed: 29%, Points required for next Level: 143

    Thanked 1 Time in 1 Post

    stratified log-rank test and proportional hazards

    Does stratified log-rank test overcome problem of violation of proportional hazards?

    I encountered a randomized controlled study of a new drug where survival curves for OS and PFS evidently crossed around first third of presented follow up time. Authors used stratified log-rank test (randomization was also stratified by three different variables that were later used for stratified log-rank test) and stratified Cox proportional hazards model to obtain significant overall result for overall survival. They stated that effects on survival were significant across all strata but they did not show survival curves. They also did not recognize inside the text that assumption of proportional hazards was violated nor did they explain why did they not use normal log-rank test.

    It seems like a positive trial. However, is it legitimate to use stratified log-rank test in this situation? (curves crossed in overall patient population and we do not know do they cross in different strata)
    Last edited by Markica85; 03-18-2017 at 02:58 AM.

+ Reply to Thread


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats