0

I would like to create a K-M curve for my data. My confusion is, I have some patients with multiple observed times and event status (see pic). Events under the same patients are independent. Would it be reasonable to create a K-M plot for all observations, even though some of those observations come from the same patient? Does it violate any assumptions? Thanks in advance!

enter image description here

curiousmind
  • 609
  • For ID = 1, he had two surgeries on 22-DEC-2015? Same question for ID=2. – user158565 Nov 10 '18 at 01:40
  • @a_statistician Correct, for ID=1, he had two surgeries, and the follow_up times were calculated from the 2nd surgery date for his 2nd and 3rd observations. However, for ID=2, it was only one surgery and the follow-up time was calculated from that surgery date. – curiousmind Nov 10 '18 at 18:07

0 Answers0