Solved
Contributor
Posts: 46

# Compare means of 2 variables from admission to discharge

Hi,

I have a data with glucose levels at admission and at discharge in 2 different variables. I want to calculate the P-value for the difference in means between these 2 variables. It means, I want to test if there was a significant change in the glucose level with time in this population.

Thanks

SAS Output

Variable

HBA1C_DIS

Mean

7.1

6.5

SD

1.0

0.8

Accepted Solutions
Solution
‎10-12-2017 03:35 PM
Super User
Posts: 23,292

## Re: Compare means of 2 variables from admission to discharge

PROC TTEST or PROC ANOVA

Check both of the examples sections.

I think, this should be a PAIRED T-Test, so make sure you decide explicitly if it's paired or not.

All Replies
Super User
Posts: 23,292

## Re: Compare means of 2 variables from admission to discharge

seltonsy wrote:

Hi,

I have a data with glucose levels at admission and at discharge in 2 different variables. I want to calculate the P-value for the difference in means between these 2 variables. It means, I want to test if there was a significant change in the glucose level with time in this population.

Thanks

SAS Output

Variable

HBA1C_DIS

Mean

7.1

6.5

SD

1.0

0.8

@seltonsy So what is your question to us? Which part, specifically, are you having issues with?

Contributor
Posts: 46

## Re: Compare means of 2 variables from admission to discharge

sorry if it wasn't clear enough,

I'd like to know what is the proc to use that can perform this test. Compre the means of the two variables and provide p-values (i.e. if the means of the 2 variables are significantly different or not).

Solution
‎10-12-2017 03:35 PM
Super User
Posts: 23,292

## Re: Compare means of 2 variables from admission to discharge

PROC TTEST or PROC ANOVA

Check both of the examples sections.

I think, this should be a PAIRED T-Test, so make sure you decide explicitly if it's paired or not.

Contributor
Posts: 46

## Re: Compare means of 2 variables from admission to discharge

Thanks,

Paired t test is the right answer I guess,

Here is what I did,

proc ttest data=cohort_1 ;
run;

☑ This topic is solved.