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
- /
- Regression analysis with Cost as outcome variable

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

11-25-2016 04:50 PM - last edited on 11-25-2016 04:53 PM by Reeza

I am trying to predict the impact of readmission events (continuous) and type of readmission (medical, surgical, other) on total hospital costs. Because costs are skewed by nature, I am applying gamma distribution with log link function. Below is the model

```
PROC GENMOD data = abc;
class type (ref='medical') gender (ref='F')/param=ref;
model cost = event type age gender /dist=gamma link=log type3;
run;
```

1) Is the model correct?

2) How can I get marginal effects from the coefficients for different variables. i.e. is there a SAS option that will give me hospital cost within each level of type of readmission and for male and female?

3) I know modified park test will help us determine whether to use gamma distribution or not, but I can't seem to find that option in SAS.

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

Posted in reply to samraty81

11-25-2016 11:08 PM

1) check EFFECTPLOT statement 2) http://blogs.sas.com/content/iml/2015/09/10/plot-distrib-reg-model.html http://blogs.sas.com/content/iml/2015/09/16/plot-distrib-exp.html