comparing surgery case loads during two periods

Discuss statistics related things
Post Reply
xav90
Posts: 8
Joined: Tue Feb 21, 2023 2:29 am

comparing surgery case loads during two periods

Post by xav90 »

Hi all,
Potentially silly question I'm hoping to get some insight on regarding how to compare trends in procedure volume in two different time periods. Essentially I have around 4500 cases(patients) for people who underwent the surgery of interest either in time period A or time period B. While I think I have a handle on comparing differences such as length of stay, hospitalization cost, etc., between the two time periods, I'm not sure how to statistically compare the changes in overall surgical volume between the two time periods.

In other words, there were ~22% fewer cases in period B than period A, and while I'm guessing that's a statistically significant decline, I don't know how exactly to test that.

Thanks for any help, I feel like this should be really straight forward but I'm struggling to grapple with this for whatever reason.

best,
Xavier
User avatar
reason180
Posts: 274
Joined: Mon Jul 24, 2017 4:56 pm

Re: comparing surgery case loads during two periods

Post by reason180 »

So you want to know if 2500 cases is significantly greater than 2000 cases?
xav90
Posts: 8
Joined: Tue Feb 21, 2023 2:29 am

Re: comparing surgery case loads during two periods

Post by xav90 »

I'm more interested in the proper way to compare the two statistically, and think for publication purposes it would be nice to share a p-value.
User avatar
reason180
Posts: 274
Joined: Mon Jul 24, 2017 4:56 pm

Re: comparing surgery case loads during two periods

Post by reason180 »

Here it is (obtained via Frequencies; Chi Square Goodness of Fit):

Image
xav90
Posts: 8
Joined: Tue Feb 21, 2023 2:29 am

Re: comparing surgery case loads during two periods

Post by xav90 »

Ah thank you! I was stuck thinking in the box of using the original file where each row is a patient, but this makes way more sense to just create another file where each row represents the time period. Thanks for humoring me and this really straight forward question.
User avatar
reason180
Posts: 274
Joined: Mon Jul 24, 2017 4:56 pm

Re: comparing surgery case loads during two periods

Post by reason180 »

FYI. You don't have to create a new data set that includes a "counts" column. "Counts" is optional. If you just leave the dataset the way it is you can run the same chi square test (without specifying any "counts").
Post Reply