turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- PROC PHREG - need help!

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-09-2010 03:57 PM

Haven't been able to find the answer to my questions about PROC PHREG at my office. I'm hoping someone here can help. Thanks!!!

I am calculating hazard ratios from a cohort (n=133,000) for a number of exposures. In the example below I am looking at intake of liquor on risk of a disease, adjusting for race and age. Example 1 is the code that was given to me to use, but according to what I’ve found online it is more common to use a calculation of TIME, i.e. number of days in the study (example 2) rather than (startage, endage) (example 1).

My first question is, what is the difference between the two? Is (startage, endage) just subtracting the two to get the same value as TIME? Or is it making some sort of adjustment for the age at entry into the cohort?

My second question is, would it be the same/different or better/worse to use (startdate, enddate) rather than (startage, endage)?

My third question, if it does in fact adjust for age at entry to the cohort, why would we use the STRATA ageatq; statement? Wouldn’t that be adjusting too much for age?

Ex. 1

PROC PHREG DATA = mm.myects MULTIPASS NOSUMMARY;

CLASS liquor (REF='0') racex;

MODEL (startage, endage)*censor(0)= liquor racex / TIES=EXACT ALPHA=0.05 RL;

STRATA ageatq;

TITLE 'Liquor, adjusting for race and age';

RUN;

Ex. 2

PROC PHREG DATA = mm.myects NOSUMMARY;

CLASS liquor (REF='0') racex;

MODEL TIME*censor(0)= liquor racex / TIES=EXACT ALPHA=0.05 RL;

STRATA ageatq;

TITLE 'Liquor, adjusting for race and age';

RUN;

I am calculating hazard ratios from a cohort (n=133,000) for a number of exposures. In the example below I am looking at intake of liquor on risk of a disease, adjusting for race and age. Example 1 is the code that was given to me to use, but according to what I’ve found online it is more common to use a calculation of TIME, i.e. number of days in the study (example 2) rather than (startage, endage) (example 1).

My first question is, what is the difference between the two? Is (startage, endage) just subtracting the two to get the same value as TIME? Or is it making some sort of adjustment for the age at entry into the cohort?

My second question is, would it be the same/different or better/worse to use (startdate, enddate) rather than (startage, endage)?

My third question, if it does in fact adjust for age at entry to the cohort, why would we use the STRATA ageatq; statement? Wouldn’t that be adjusting too much for age?

Ex. 1

PROC PHREG DATA = mm.myects MULTIPASS NOSUMMARY;

CLASS liquor (REF='0') racex;

MODEL (startage, endage)*censor(0)= liquor racex / TIES=EXACT ALPHA=0.05 RL;

STRATA ageatq;

TITLE 'Liquor, adjusting for race and age';

RUN;

Ex. 2

PROC PHREG DATA = mm.myects NOSUMMARY;

CLASS liquor (REF='0') racex;

MODEL TIME*censor(0)= liquor racex / TIES=EXACT ALPHA=0.05 RL;

STRATA ageatq;

TITLE 'Liquor, adjusting for race and age';

RUN;