How best to analyze a simple medical study to show statistically significant difference?

#1
How best to analyze this data to show if there is a statistically significant difference within each group?

Measuring the of detection or determination of the entry of the needle bevel into the epidural space, using two different diameter needles 25 G (smaller) and 18 G (larger).

Summary of results:

Group 1: 18G needle (20 patients - not the same patients as in Group 2)

Recognition of the epidural space was confirmed by LORT in 12 patients out of 20 (60%).

By CST in all 20 patients (100%).



Group 2: 25G needle (20 patients - not the same patients as in Group 1)

Recognition of the epidural space was confirmed by LORT in 2 patients out of 20 in (10%).

By CST in all 20 patients (100%).
 

Attachments

Karabiner

TS Contributor
#2
How best to analyze this data to show if there is a statistically significant difference within each group?
So you want to know whether confirmation rate is different between LORT and CST in group 1,
and the same question is asked in group 2? That could be done by using McNemar's test for
dependent observations (each patient was measured twice, using LORT [result yes/no], and
CST [result yes/no], respectively).

With kind regards

Karabiner
 

Dason

Ambassador to the humans
#3
I don't know what your acronyms mean. Gretagarbo probably doesn't appreciate them either. It's a good idea to give definitions when introducing these abbreviations.