How to conduct a time-to-event survival analysis in a 1:2 matched case-control study?

Denali — Mon, 28 Mar 2022 18:39:03 GMT

Hi,

I am working on a 1:2 matched case-control study with 20 cases and 40 controls. They were matched based on age +/- 5 years, gender, race and cancer type. We would like to compare the survival rate from the date of receiving new treatment (time 0) until the time of death or last follow up date and evaluate whether infection and smoking status would be assoicated with survival. Could anyone please verify whether my code below is correct?

proc phreg data=test;

strata Matching_ID;

class group infection smoking;

model time*event(0) = group infection smoking /risklimits ;

run;

"Matching_ID" stands for the matching group ID number. In our study, there are 3 subjects coded per matching_ID (1 case + 2 matched controls)

"time" stands for the "duration from the date of receiving new treatment to endpoint death or alive on the last follow-up date"

"event" stands for "death or alive", coded as (1,0)

"group" stands for "cases or controls", coded as (1, 0)

"infection" stands for whether they had infection or not, coded as (1,0)

"smoking" stands for whether they had smoking or not, coded as (1,0)

Thanks in advance!

Re: How to conduct a time-to-event survival analysis in a 1:2 matched case-control study?

OsoGris — Tue, 29 Mar 2022 08:37:54 GMT

See Example 89.5 Conditional Logistic Regression for m:n Matching<> in the PHREG documentation.

topic Re: How to conduct a time-to-event survival analysis in a 1:2 matched case-control study? in Statistical Procedures

How to conduct a time-to-event survival analysis in a 1:2 matched case-control study?

Re: How to conduct a time-to-event survival analysis in a 1:2 matched case-control study?