[R-meta] meta-analysis of ratios or ratios/ratios? can you help me ?

[Nicky Welton]

Hi Wolfgang,

I replied to Martin (below), but not the list - apologies. Not sure if it answered his question though.

Best wishes,

Nicky

Dear Martin,

You might find the following GMD2 document helpful that we wrote for NICE Guidelines developers on synthesis with continuous outcomes, including ratios or ratios of means (RoRoM):
https://www.bristol.ac.uk/population-health-sciences/centres/cresyda/mpes/nice/guideline-methodology-documents-gmds/

Keep meaning to write it up as a paper, but never seems to make it up my to do list!

Best wishes,

Nicky

-----Original Message-----
From: R-sig-meta-analysis <r-sig-meta-analysis-bounces using r-project.org> On Behalf Of Viechtbauer, Wolfgang (NP) via R-sig-meta-analysis
Sent: Wednesday, January 17, 2024 11:30 AM
To: R Special Interest Group for Meta-Analysis <r-sig-meta-analysis using r-project.org>
Cc: Viechtbauer, Wolfgang (NP) <wolfgang.viechtbauer using maastrichtuniversity.nl>; Martin Lobo <mlobo4370 using hotmail.com>
Subject: Re: [R-meta] meta-analysis of ratios or ratios/ratios? can you help me ?

Dear Martin,

I suspect you are not getting any responses as it is a bit difficult (at least for me) to follow your question. But if I understand your question correctly, then here are some thoughts:

It is not clear to me if the authors report mean(FINAL)/mean(BASAL) or if they report mean(FINAL/BASAL). In other words, do they report the ratio of the means of the mean of the (individual) ratios?

If it is the former, then mean(FINAL) / mean(BASAL) is the ratio of the means, which is a well-known effect size measure in meta-analysis (which is log transformed for the analysis). This measure is sometimes called the (log) response ratio. But to compute the sampling variance of log(mean(FINAL) / mean(BASAL)), one needs to known the means and SDs of the FINAL and BASAL values.

If it is the latter, then mean(FINAL/BASAL) is really just a mean and could be meta-analyzed as such (assuming that SD(FINAL/BASAL) or the corresponding standard error are reported). Note though that ratios such as FINAL/BASAL are not 'symmetric' around 1. For example, if FINAL/BASAL is 2 for one person and FINAL/BASAL is 0.5 for another person (which is really the same magnitude on such a ratio scale, but in opposite directions), then mean(FINAL/BASAL) is not 1, which is counter-intuitive.

As for the analysis approaches: You can do approach 1 using a multilevel/multivariate model with rma.mv().

Best,
Wolfgang

> -----Original Message-----
> From: R-sig-meta-analysis <r-sig-meta-analysis-bounces using r-project.org> 
> On Behalf Of Martin Lobo via R-sig-meta-analysis
> Sent: Wednesday, January 10, 2024 14:07
> To: r-sig-meta-analysis using r-project.org
> Cc: Martin Lobo <mlobo4370 using hotmail.com>
> Subject: [R-meta] meta-analysis of ratios or ratios/ratios? can you help me ?
>
> Hello everyone.
>
> I'm going to expand the explanation to see if someone can help me. I 
> want to investigate the changes in uCRP with a specific treatment vs 
> placebo. The difficulty I have is that very few report this 
> quantitative variable (uCRP). But I have found many papers that report 
> the FINAL/BASAL relationship in both branches. From this data I came up with two strategies.
> 1- Analyze the FINAL/BASAL relationship in the two branches with their 
> respective confidence intervals.
> 2- Analyze the ratio of ratios (FINAL/BASAL TREATMENT BRANCH / 
> FINAL/BASAL PLACEBO BRANCH) in both branches with their respective confidence intervals.
>
> I can't find how to analyze option one with metagen, since I can only 
> enter the TE and the seTE. I don't know if with this data the two 
> branches could be analyzed directly in some way or directly with metafor.
>
> Although option 2 could be analyzed with metagen using OR or RR as 
> effect measures, I would not know if it would be conceptually correct, 
> since the data are the percentage changes of a continuous variable and 
> not the proportion of binary events.
> I would appreciate help with this analysis.
>
> Have a great year everyone.
>
> Martin from argentina
>
> Lorenzo Martín Lobo MTSAC, FACC, FESC
> Especialista Jerarquizado en Cardiología Jefe de Dpto Enf. 
> Cardiovasculares y Cardiometabolismo Hospital Militar Campo de Mayo.
> Jefe de Cardiología Hospital Militar Campo de Mayo Ex Jefe de Unidad 
> Coronaria Hospital Militar Campo de Mayo Miembro Titular de la 
> Sociedad Argentina de Cardiología Fellow American College of 
> Cardiology Fellow European Society of Cardiology Ex Miembro del Area 
> de Investigación de la SAC Ex Director del Consejo de Aterosclerosis y 
> Trombosis de la SAC Miembro Asesor del Consejo de Aterosclerosis y 
> Trombosis de la SAC Ex Director del Consejo de Epidemiología y 
> Prevención Cardiovascular de la SAC Miembro Asesor del Consejo de 
> Epidemiología y Prevención Cardiovascular de la SAC Experto en Lipidos 
> de la Sociedad Argentina de Lipidos.
> Miembro de la Sociedad Argentina de Lipidos.
> Instructor de ACLS de la American Heart Association
_______________________________________________
R-sig-meta-analysis mailing list @ R-sig-meta-analysis using r-project.org To manage your subscription to this mailing list, go to:
https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis