List of articles: exploring trial variables and effect on meta-analysis

I’m increasingly drawn to the concept of what is ‘correct’ in relation to systematic reviews and meta-analyses.  If we’re wanting to demonstrate the worth of a ‘rapid’ reviews we need to compare it to a ‘gold standard’ ‘correct’ answer.  There does not appear to be such a thing.

I’m particularly drawn to the 2014 Deschartres paper that:

…compare treatment outcomes estimated by meta-analysis of all trials and several alternative analytic strategies: single most precise trial (ie, trial with the narrowest confidence interval), meta-analysis restricted to the 25% largest trials, limit meta-analysis (a meta-analysis model adjusted for small-study effect), and meta-analysis restricted to trials at low overall risk of bias.

So, same question, different samples of trials (one of which was ‘all’ trials) and different results.  The authors conclude:

Estimation of treatment outcomes in meta-analyses differs depending on the strategy used. This instability in findings can result in major alterations in the conclusions derived from the analysis and underlines the need for systematic sensitivity analyses.

They further give practical considerations (heuristics?):

Estimation of treatment outcomes in meta-analyses differs depending on the strategy used. This instability in findings can result in major alterations in the conclusions derived from the analysis and underlines the need for systematic sensitivity analyses.

An important topic and some interesting papers:

 

Advertisements

4 thoughts on “List of articles: exploring trial variables and effect on meta-analysis

  1. If I ruled the world, meta-analysis would be restricted to trials at overall low risk of bias. I would also be limited to trials measuring patient-important (core) outcomes using validated outcome measures. Yes, I know that would mean most systematic reviews would be empty, but at least they could be produced rapidly, would correctly reflect generally poor state of evidence and could act as campaign documents to generate protocols for good quality research.

    Liked by 1 person

  2. I agree about the patient outcomes, absolutely and I’m currently writing a paper that includes this element.
    I’d do think the 2014 Dechartres was interesting in that respect of only using low risk of bias trials. It does make me wonder if there is such a thing as a ‘right’ answer. But, for a given intervention which is better an unbiased trial or a meta-analysis with biased trials. I think I’d prefer the former!

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s