If you’ve only recently bumped into this blog, you may not be aware of the near-obsessive compulsion for us to ask “Is there a systematic review of that?” at any opportunity. Why is this the case? Well, partly it’s the job. Mostly it’s because systematic reviews give us the best answers to our questions … and in this Basics blog we’ll cover
- What is a systematic review?
- What is meta-analysis?
- What can a systematic review do for us?
A systematic review is a formal, scientific study, which seeks to answer a specific question by collating, appraising and synthesising information from other research studies. In health care, it often seeks to answer a PICO question, uses risk of bias assessments and meta-analysis. In collating all the evidence, it avoids cherry picking the studies that tell us the answer we want, and in appraising it, it seeks to remind us that not all evidence is created equal (~ly valid).
With meta-analysis, we are really just doing a special version of a weighted average … most of the time … that first of all tells us what the average effect size is going to be. Then it can also tell us how varied the result might be if it were done in different study settings or populations.
Or perhaps tell us how taking an average might be unhelpful (like this graph, showing how the proportion of CF-diagnosis in babies with echogenic bowel changed dramatically as the reporting era altered).
So the results of a systematic review might be to
- Give us the most precise effect size of an intervention (how likely it is to work)
- Tell us if it varies between groups, doses, techniques etc – or not
- Explore where the research is centred, and where the gaps in research are
- Command an oversight of all information collected so far and so demand more consistent, patient-centred and meaningful outcomes
- Give us an idea of how to put it into practice
You can appraise a systematic review in a fairly FAST way… (click on the link to appreciate that one) … and do feel free to tweet/FB or comment to us about your own thoughts on how we can understand SRs better.