This is a continuation of a long topic, addressed by post number rather than by page. Start at post 1 · go to the accepted answer.
Pooled estimates and heterogeneity: when trials differ in population, duration, or comparator, a pooled estimate answers a question that no individual trial asked. High heterogeneity means effects genuinely differ across studies. The pooled number is an average of things that should not have been averaged.
Flagging that the sources on this are thinner than the confidence in the thread suggests.
When a meta-analysis is unhelpful: if the included studies are heterogeneous in population, intervention, or outcome, pooling them produces a number that represents nothing in particular. Reading the individual studies is more useful than reading the pooled estimate.
Confirming post #60 from a second method, which matters more than confirming it from a second person.
Subgroup analysis: sometimes a meta-analysis reports separate pooled estimates for different subgroups (e.g., by baseline body mass index or by trial duration). Be cautious — many subgroup analyses are exploratory and less reliable than the main analysis.
The part I am sure of is shorter than the part I have written.
Post #62 describes the usual case. This is about the unusual one.
Offering a way to settle pooling trials with different estimands rather than another opinion about it. Two measurements, taken the same way, a fortnight apart. If the difference is within the noise, the question was not answerable at this precision.
Reading rather than contributing, but this is the most useful thread I have found on it.
A meta-analysis of four small trials is not stronger evidence than one adequately powered trial, whatever the summary statistic looks like.
It took me longer than it should have to see that.
Taking post #68 at face value and following it one step further.
An observation about pooling trials with different estimands that I cannot explain and am posting anyway, on the principle that unexplained observations are more useful public than private.
Pooling trials with different estimands: I would want to see the raw numbers rather than the summary before agreeing. Summaries lose exactly the information that would settle this.
Where the included trials share a sponsor and a protocol template, their errors correlate and pooling does not average them out.
Everything in post #69 holds. The case it does not cover is the one I have.
Fixed-effects versus random-effects models: fixed-effects assumes all studies are estimating the same thing and variation is sampling error. Random-effects assumes studies are estimating effects from different distributions and allows between-study variance. Choice matters if heterogeneity is high.
That matches what I was told, which is not the same as knowing it.
I came in to disagree and I am leaving without a disagreement.
Answering the question post #73 raises rather than the one it answers.
Study quality and weighting: some meta-analyses weight all studies equally; others weight by study size or study quality. The choice affects the result and should be stated and justified.
Written in the hope of being told what I have missed.
Two people in this thread mean different things by pooling trials with different estimands and are disagreeing about the definition while believing they are disagreeing about the facts. Worth pausing to define it.
On post #78 — agreed on the reasoning, with one qualification.
On pooling trials with different estimands: the maintained page in the documentation commons covers the general case with citations and a review date, which is more reliable than any reply here including this one.
Picking up post #78: that is the part I would want checked first.
Double extraction with disagreement resolution is standard and is worth checking for, because single extraction errors are common and non-random.
I am aware this is the third time this month I have made this point.
Narrowing post #78, because the general version has more than one answer.
Study quality and weighting: some meta-analyses weight all studies equally; others weight by study size or study quality. The choice affects the result and should be stated and justified.
Worth saying I have only my own numbers here, and n is small.
On pooling trials with different estimands the community has more anecdote than the confidence in this thread implies, and I include my own contribution in that.
Pooling trials with different estimands looks different depending on whether you are reading the primary literature or the summaries of it, and the difference is not in our favour.
Post #82 answers the question as asked. The question underneath it is different.
Quality assessment of included trials should change the analysis rather than sit beside it. A sensitivity analysis excluding the weakest studies is the minimum.
Noting that the question and the thing people usually mean by it are different.
Careful with the language on pooling trials with different estimands. "Not detected" and "not present" are different findings and the first is a statement about the method.
Overlapping populations across included trials inflate the apparent sample size. It happens more than people expect where the same programme reports multiple papers.
This is where my knowledge stops and I would rather mark the edge than blur it.
Pooling trials with different estimands is worth one more sentence than it usually gets, and the sentence is the one about how the number was arrived at.