[R-meta] Calculating odds ratios in bivariate meta-analysis.

[Olina Ngwenya]

Thank you Wolfgang, well understood.

Regards

Olina

-----Original Message-----
From: Viechtbauer, Wolfgang (SP) <wolfgang.viechtbauer using maastrichtuniversity.nl> 
Sent: 21 March 2022 10:54
To: Olina Ngwenya <olina.ngwenya using manchester.ac.uk>
Cc: r-sig-meta-analysis using r-project.org
Subject: RE: [R-meta] Calculating odds ratios in bivariate meta-analysis.

Hi Olina,

escalc() is a function to compute various outcome measures / effect sizes. How these are used in further analyses is an entirely different issue. For example, in this example, a bivariate model is used with log odds as outcomes computed with escalc():

https://www.metafor-project.org/doku.php/analyses:vanhouwelingen2002

Best,
Wolfgang

>-----Original Message-----
>From: Olina Ngwenya [mailto:olina.ngwenya using manchester.ac.uk]
>Sent: Monday, 21 March, 2022 11:14
>To: Viechtbauer, Wolfgang (SP)
>Cc: r-sig-meta-analysis using r-project.org
>Subject: RE: [R-meta] Calculating odds ratios in bivariate meta-analysis.
>
>Dear Wolfgang
>
>Thank you for your response. I felt like the escalc() with measure 
>="OR" , is used for univariate meta-analysis and I thought there is 
>another function used to calculate odds ratios in bivariate/multivariate meta-analysis.
>
>Regards
>
>Olina
>
>-----Original Message-----
>From: Wolfgang Viechtbauer 
><wolfgang.viechtbauer using maastrichtuniversity.nl>
>Sent: 20 March 2022 22:36
>To: Olina Ngwenya <olina.ngwenya using manchester.ac.uk>
>Cc: r-sig-meta-analysis using r-project.org
>Subject: Re: [R-meta] Calculating odds ratios in bivariate meta-analysis.
>
>Dear Olina,
>
>Not sure where exactly the problem is, but escalc() with measure="OR"
>computes log odds ratios and measure="PLO" computes log odds.
>
>Best,
>Wolfgang
>
>> Hello R-sig-meta-analysis folks,
>>
>> I am doing a random effects multivariate model (bivariate)  for 
>> binary outcomes, pregnancy and live birth (similar to this analysis:
>> https://www.metafor-project.org/doku.php/analyses:berkey1998, the 
>> difference is that berkey1998 outcomes are continuous). Now for 
>> bivariate analysis of binary outcomes, I want odds ratios and have 
>> not found the way of getting log odds and odds ratio like in the 
>> univariate analysis (for example the output for rma.peto).
>>
>> Thank you
>>
>> Regards
>>
>> Olina Ngwenya Research Assistant Centre for Biostatistics | School of 
>> Health Sciences | Faculty of Biology, Medicine and Health | 
>> University of Manchester