[R] Help in meta-analysis (URGENT please)

Michael Dewey lists at dewey.myzen.co.uk
Tue Feb 9 19:08:27 CET 2016


Dear Rosa

I suggest reformatting your dataset to have four variables (a) study (2) 
MSM yes or no (3) number with HIV (4) total number. Then you can use any 
of the options in metafor (look for outcomes for individual groups in 
?escalc) and use MSM as a moderator.

Is there any reason why you did not just do a logistic regression?

On 08/02/2016 18:40, Rosa Oliveira wrote:
> Dear all,
>
> I’m conducting a met analysis and I usually use Revman, bur as I’m trying to use R more and more, I would like to conduct the met analysis here, in R (R-studio).
>
> One off my problems, I think, is that:
> 1st. it’s the first time :)
> 2. I only have data for 1 arm as you can see on the data that follows.
>
> ARTIGO	qt	tt	qc	tc	Personal Notes	qt2
> Giuliani M. (2014)	-	-	1515	1862	only MSM.	347
> Diaz A. (2015)	-	-	-	-	only MSM (n=3081)	2499
> Niedźwiedzka-Stadnik M. (2015)	828	1098	-	-		326
> Hoenig M. (2015)	-	-	8506	-	MSM (n=8925)	419
> Wu H. (2015)	58	145	-	16713	n=16892	87
> Pan X. (2015)	-			-	only MSM (n=1316)	-
> Ma Q. (2015)	-			-	only MSM (n=424)	-
> Op de Coul E. (2015)	8596			-	HIV-infected patients (n=20965)	12369
> Liu G. (2015)	-	-	1003 (?)	-	only MSM (n=1041) - some converted to HIV+ during the study	-
> Hoenigl M. (2015)	-			-	only MSM (n=8935) analysis HIV tests repetitions	-
> Moller L. M. (2015)	-	-	469	-	only MSM (N=561)	92
> watkins (2015)	-	-	-	-	only MSM (n=1154) only analysis believes concerning the risk	-
> den Dass C. (2015)	-	-	2408	-	only MSM (n=3787)	589
> Jia Z. (2015)	-	-	5314	-	only MSM (n=5800)	486
> solomon S. (2015)	-	-	10875	-	only MSM (n=12022)	1147
> Diez M. (2014)	-	3599	-	-	n=145337	
>
>
>
> legend:
> qt the number of hiv subjetcs who are not sms.
> qt2 the number of hiv subjetcs who are  sms.
> tt the total number of hiv subjects.
> qc the number of subjetcs who are sms without being hiv.
> tc the total number of subject not hiv.
>
>
>
> Is it possible to conduct a met analysis concerning the risk of hiv among MSM relative to the ones that are not MSM?
>
> or simply concerning the risk of hiv among MSM????
>
> If yes, How?
>
> Metafor? I’ve tried, but wasn’t succeed :(
>
>
> Best,
> RO
>
>
>
> Atenciosamente,
> Rosa Oliveira
>

-- 
Michael
http://www.dewey.myzen.co.uk/home.html



More information about the R-help mailing list