Entering edit mode
3.7 years ago
akhattri
▴
50
We are analyzing some TCGA Overall Survival (OS) and Progression Free Interval (PFI) data. As for some cancer types the OS could be very long as compared to others. Moreover, with a longer duration, the probability of events (death) due to unrelated reasons (like, due to a car accident) increases, and hence several studies censor the data for five years (1825 days).
I am not sure of the correct censoring strategy for OS and PFI data. Do we consider all the numbers >=1825 as 1825 days or is there any other thing to consider? I am providing a small subset of the actual TCGA data that we would like to analyze. Thanks!!
ID type OS OS.time PFI PFI.time
B6-A0IA brca 0 8391 0 8391
B6-A0RN brca 0 8008 0 8008
B6-A0RE brca 0 7777 0 7777
BH.A18T brca 1 0224 1 224
AN.A0FN brca 0 0218 0 218
B6.A400 brca 0 0215 0 215
AN.A0FK brca 0 0213 0 213
