[R] (Meta-analysis) How to build|fake a [n]lm[e] object ?

[Thomas Lumley]

On Wed, 5 Dec 2001, Emmanuel Charpentier wrote:

>
> B) Alternatives to the odds-ration for event data ?
> ===================================================
>
> The usual way to assess effects for categorical variables is to compute the
> log(odds-ratio) for each study and to pool them using inverse variance as
> weights (that's what meta.DSL and meta.MH do, respectively for random and fixed
> effect model).
>
> However, in some trials, some event have a frequency of zero in one or both
> groups. In the first case, one can neglect the said trial for the assessment of
> the treatment effect, on the basis that it is not informative. In the second
> case, however, the data cannot be used (because the OR is either zero or
> infinite, with infine asymptotic variance). The treatment assessment by OR
> pooling dismisses these trials (see meta.DSL source, for example ; and this is
> also the case in other meta-analysis packages, such as Cochrane's RevMan).

meta.MH doesn't have this problem -- it's quite happy with zero cells.

> But the asymetry (some events in one group and none in the other) is indeed an
> information, and I do not feel at ease with discarding it. The best I can think
> of is the ordinary test of independance (Fisher's test, in this case) on a
> contingency table "summing" the individual trials' contingency tables. This
> analysis confirms the results iof the meta-analysis. But it does not account
> for trials' heterogeneity, which is a large part of the point of a
> meta-analysis.

Either meta.MH or conditional logistic regression (clogit in the
survival package) would fix this

> Someone suggested to me to add a "small" quantity (say 1, or 0.5, as in the
> case of Yate's correction for continuity) to the event counts in these groups,
> ant to see if the inclusion of these study would entail a modification of the
> results, but I'm "isntinctively" not satisfied with this approach.
>

If you want a fixed effect of treatment there's no problem (and I
personally don't like meta-analyses where a random-effects model makes a
difference)

If you need a random effects model that doesn't object to zero cells then
lme() and variants aren't going to work, and you need a real generalized
linear mixed model with random intercept and random treatment effect.
Logistic mixed models are a hard problem.  Jim Lindsey's 'repeated'
package may handle this, though.

A little simulation would tell you what the properties of the `continuity
correction' approach are.


	-thomas

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._