Need help with statistics to compare two lab values to predict adverse outcomes in patient population

#1
Hello. I am an RN recently in a new role where I do some data abstraction. It’s be a long time since I attempted statistics and I would appreciate any form of help.
I am looking to collect two different labs to see which is a better indicator of adverse outcomes in my patient population. Any idea of how I’d go about doing this and what my calculations would look like? Or how I can look at this data on an excel spreadsheet? The site where I need to enter this data would like to know numerator and denominator? What would they be? Thanks for your feedback.
 

Karabiner

TS Contributor
#2
Admittedly, I do not quite understand what you want us to do. To suggest how to enter the data in Excel?
What do you mean by "the site where I need to enter the data"?

Can you use other software than Excel, and how large is your sample size? Can you give a little bit more
information about the nature of that lab values, and what kind of variable "adverse outcome" is (a yes/no
variable for example?).

With kind regards

Karabiner
 
#3
hello karabiner,
Obviously I don’t have a statistics background lol.
I run a joint replacement program for a hospital. Every two years we get our program recertified. In order to do so, we need to have 4 performance improvement measures and have to show meaningful use of the data we collect for those measures. Usually the calculations are fairly simple. This one has me stumped.

excel is the only program I think I can use.

What I want to do is collect two labs from each patient during preop bloodwork. The two labs are two different glycemic markets. Patients with diabetes have increased risk of infections and wound complications after joint replacement surgery. Elevated glycemic markers identify patients at risk for complications after Joint replacement surgery. Traditionally one glycemic marker (let’s call it GMA) has been used. If you get a high GMA value, surgery is usually postponed for 3 months. However new info suggests that’s a different glycemic marker, GMB, is just as good predictor of complications and that lab can be drawn in 2-3wks instead of delaying surgery 3months.

So the data I will have is:
All joint replacement patients
GMA elevated (y/n)
GMB elevated (y/n)
Complication (y/n)