Binary outcome - How to test for superiority?

#1
Hi guys!
I am doing an audit of hospital records and would be grateful for some guidance regarding the correct statistical test to use.

We recently changed from recording patient's medications on paper forms, to electronic forms.

I want to know if my data proves that the new electronic system is better.
I think this is a very straightforward thing to demonstrate, but I'm just not certain how to go about doing it, so I'd be grateful for help!

Outcomes:
Paper charts: Of 60 charts audited, the medications were documented in 37 charts (61.66%)
Electronic charts: Of 80 charts audited, medications were documented in 64 charts (80%)

Does anyone know what test I should use here?
Thanks!
 

Karabiner

TS Contributor
#2
A common test would be to consruct a fourfold table wit
system type (paper/electronic) and medication documented
(yes/no) and perform a Chi² test of association.

With kind regards

Karabiner
 

hlsmith

Less is more. Stay pure. Stay poor.
#4
How are you defining superiority? You need to define the threshold for superiority, then calculate the rate difference with one-side confidence interval, so all the alpha is on one-side, the side closest to the threshold. If the interval excludes the superiority threshold, you have evidence to support superiority.