topic Re: PROC MIXED REPEATED in SAS Procedures

PROC MIXED REPEATED

deleted_user — Thu, 13 May 2010 20:24:14 GMT

I am conducting an analysis on repeated measures data by using PROC MIXED with REPEATED statement.

I have an example coding available for a typical ‘drug test’ scenario below (i.e., various drugs tested on patients, and the patients’ heart rate monitored every hour):

proc mixed data = mydata;
class drug patient hour;
model heartrate = drug hour drug*hour;
repeated hour / subject=patient(drug) type=un;
run;

If instead, assuming you wanted to monitor the influence of various dosage of a given drug X (DOSE) hourly (HR) on patients (PAT) consisting of both genders (GEN), each having 3 weight categories (WC) as blocks. How do you properly modify the basic coding example provided above? I've wasted 3 days trying to get this right... Thanks for your help.

Re: PROC MIXED REPEATED

deleted_user — Thu, 13 May 2010 20:49:33 GMT

...of course, I'm still interested in the influence of drug dosage on the patients' heart rate (HEART).

Re: PROC MIXED REPEATED

SteveDenham — Fri, 14 May 2010 11:33:12 GMT

The following code is untested. I assume that dose is categorical and fixed across drugs. If not, then see part 2.

proc mixed data = mydata;
class drug patient hour dose gen;
model heartrate = gen|drug|dose|hour;
repeated hour / subject=patient(drug) type=un;
run;

If dose is continuous and varies by type of drug then:
proc mixed data = mydata;
class drug patient hour gen;
model heartrate = gen|drug|hour dose drug*dose;
repeated hour / subject=patient(drug) type=un;
lsmeans gen|drug|hour/at dose={insert value of interest};
run;

These should provide a good start, but it may be that the covariance structure contains too many parameters, in which case an autoregressive or spatial power structure might be more usable.

Steve Denham

Re: PROC MIXED REPEATED

deleted_user — Fri, 14 May 2010 12:33:51 GMT

Hi Steve,

Thanks so much for taking time to help me out (yet again!). This certainly helps me greatly to have a good starting point. I’ll try it out and see what covariance structure works best etc. Best regards!

Re: PROC MIXED REPEATED

deleted_user — Mon, 17 May 2010 13:42:04 GMT

Hi Steve,

Could you kindly tell me how I would account for the 3 weight categories (blocks) in the coding version 1? Would I use the RANDOM statement??

proc mixed data = mydata;
class drug patient hour dose gen;
model heartrate = gen|drug|dose|hour;
repeated hour / subject=patient(drug) type=un;
run;

Thanks!

Re: PROC MIXED REPEATED

deleted_user — Mon, 17 May 2010 17:23:13 GMT

Hi Steve, I tried modifying the base code you have posted for me in order to account for the blocking effect. I am not confident of the changes I have made though…how does it look? Thanks!

Scenario (just as a reminder):
A specific drug X is administered to patients at various dosages (DOSE), and the change in patients’ heart rates is monitored hourly (PATIENT, HEARTRATE, and TIME). Men and women are tested separately (GENDER). Also, prior to assigning patients to a certain dosage, three weight categories (BLOCK) are created. From all weight categories, equal number of patients is assigned to a certain dosage.

proc mixed data=mydata;
class dose patient time gender block;
model heartrate= dose|time|gender;
random block;
repeated time /subject=patient(dose*gender*block) type=un;
run;

(As this thread is getting longer, I’ll start a new one with this post)