PDA

View Full Version : Summary sensitivity and specificity with 95% confidence interval



Mmhr
12-25-2015, 06:52 AM
Good day,

I'm an ophthalmologist registrar busy with a thesis. I am doing a diagnostic meta-analysis of some tests, and have a table with one test's sensitivity and specificity data from different studies. I also have the true and false positives and negatives.

Is it possible to calculate a summary sensitivity and specificity (with 95% confidence interval) for the test?

A partial example below.

Regards



Test A









Study
True +
False +
False -
True -
N
sensitivity
Specificity


1
50
59
6
475
590
0.89
0.89


2
42
111
4
157
314
0.91
0.59


3
17
3
91
81
192
0.16
0.96


4
14
144
2
250
410
0.88
0.63




123
317
103
963
1506




Summary or combined value?




?
?



Combined 95% confidence interval?