topic Cox regression, non-significance in Statistical Procedures

Cox regression, non-significance

Gilles-Protais — Sun, 05 May 2024 22:06:09 GMT

Greetings dear members,
I have a dataset of 630 patients, with just 8 events, so when I run a Cox regression with Firth correction, I have that none of the variables taken under consideration,are significant. Please are there other ways to study or explore in such situations,

thanks.

Re: Cox regression, non-significance

PaigeMiller — Sun, 05 May 2024 23:29:17 GMT

With 8 events in 630 patients, it's really not surprising that nothing is significant. I think it is also unlikely that another analysis method would return a different result.

The real answer is that you don't have enough data to find statistical significance.