Loop for calculating OR in 2x2 tables (cci)

[Katrine Lawaetz]

Hi,

I am doing a systematic review and I am comparing ORs. For most of the studies, I can only extract prevalence, calculate numbers to insert in a 2x2 table and then calculate an OR.
There is one study per observation: cases have an aaa (abdominal aortic aneurysm), exposure is dm (diabetes)

aaa_dm	aaa_nodm	naaa_dm	naaa_nodm
87	382	1554	4119
7	28	131	556
11	208	143	2235
3	12	156	480
1	18	101	634
10	158	647	13434
3	8	272	726
3	38	10	449
66	401	809	6913
43	371	853	6879
13	52	212	1292
11	45	325	2010
19	88	709	3899
40	212	312	1389
24	209	1754	12624
7	26	78	616

Does anyone know how to do that in a loop, so I do not have to calculate all by hand?
I need to save the computed ORs along with CI (my guess is return list, gen var=r(or) - but then the variable is for all the observations and not just the exact study?)

Can anyone please help me? I know, I can do it by hand, but I would prefer programming my way out of it as to making it reproducible
Let me know if you need further information or clarification.

Thank you & Happy Holidays,

Kind regards,
Katrine

[Marcos Almeida]

After a systematic review, shall you wish to measure the "effect size", you are supposed to perform a meta-analysis approach, not a loop to calculate ORs "manually".

There are several reasons for this, and I don't even mention the burden to create the loop: the "correct" calculation of dispersion parameters, selection of weights, type of effects, link to immediate graphics, etc,

This text presents most of Stata's user-written packages concerning meta-analysis.

Hopefully that helps.

[Katrine Lawaetz]

Dear Marcos,

Thank you for your reply.
I thought, I could present the ORs in a forest plot as a result of my review, but perhaps this "calculation" qualifies for a meta-analysis.
I will have a look at it.

Thank you,

Best regards,
Katrine