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
- /
- Health Care and Pharma
- /
- SAS in Health Care Related Fields
- /
- A question of SAS Mixed Proc

Topic Options

- 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
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

05-26-2006 03:11 AM

I met a problem in the "Dose Proportionality" study which is a part of PK study in the Phase I clinical trial. The Power Model is oftened used to analyze the "Dose Proportionality" study. The model set up the following equation to discrib the problem:

log(yijk) = si + pj + βi •log(Dk) + εijk

The equation is base on the study consisting of a crossover design with D doses being given to N subjects over P periods. Commonly D and P are the same. yijk represents the measure of PK parameter (such as AUC) on the kth dose in the jth period for the ith subject; sj is a random subject effect; pj is a fixed period effect; and εijk is a normally distributed random error with mean value zero. If the slope is assumed common for all subjects, it will be β.

The model suggests using the SAS MIXED PROC to calculate the estimate and confidence interval of β. The REML method is also recommended for the mixed effects model.

I wrote the following code:

/*The numbers are omitted*/

Did I write the correct program?

Are the "Lower" and "Upper" the results of the Confidenc Interval calulated by the REML method?

Thank you!

log(yijk) = si + pj + βi •log(Dk) + εijk

The equation is base on the study consisting of a crossover design with D doses being given to N subjects over P periods. Commonly D and P are the same. yijk represents the measure of PK parameter (such as AUC) on the kth dose in the jth period for the ith subject; sj is a random subject effect; pj is a fixed period effect; and εijk is a normally distributed random error with mean value zero. If the slope is assumed common for all subjects, it will be β.

The model suggests using the SAS MIXED PROC to calculate the estimate and confidence interval of β. The REML method is also recommended for the mixed effects model.

I wrote the following code:

proc mixed;

model ln_auct = ln_dose /alpha=0.05 cl intercept;

random subject;

run;

The SAS system gave out the following results

Solution for Fixed Effects

Effect, Estimate, Standard Error, DF, t Value, Pr>|t|, Alpha, Lower, Upper

Intercept ... ... ... ... ... ... ... ...

log_dose ... ... ... ... ... ... ... ...

/*The numbers are omitted*/

Did I write the correct program?

Are the "Lower" and "Upper" the results of the Confidenc Interval calulated by the REML method?

Thank you!