Evidence pyramid
The evidence pyramid suggests a hierarchy of research and evidence. Find out more about his concept and how you can potentially use this hierarchy to help you to decide where to search and how to prioritise your sources.
Evidence pyramid explained
Image source: Grandma Gan - Wikimedia commons: Drawn image illustrating the Hierarchy of Evidence
Creative commons license: CC BY-SA
This illustration shows us the various types of evidence that you might come across and their rank in terms of reliability. We call this the research hierarchy.
The top three are: systematic review; critically appraised topics; critically appraised individual articles
These summarise a range of research and tell you how strong the evidence is so that someone else has done some of the appraisal work for you – people whose job it is to do that, who do it every day.
The bottom four are: randomised control trials; cohort studies; case reports/studies; expert opinion and anecdotal evidence.
Many of us are familiar with randomised control trials (RCTs) as these are useful for quantifying the effects of an health or care intervention.
Cohort studies are less beneficial for effects of interventions but good for answering questions about prognosis, diagnosis and frequency
Case reports, expert opinion and anecdotal evidence are incredibly valuable but they do tend to be more at risk of bias and the findings might not be transferable or easily explained.
Criticisms and other approaches
The evidence hierarchy has been around since the late 70s with lots of variations, but there are some criticisms of this that that are worth considering.
- The hierarchy only looks at study design and does not look at sample sizes or the robustness of the research.
- There are also some criticisms that systematic reviews don’t belong in this hierarchy because these are filtered results rather than a study design.
Other approaches
Murad, M.H. et al. (2016) suggested an evidence pyramid that looked not just at study type. Their article proposed a pyramid with sections separated by wavy lines, indicating that in certain circumstances a study type at the bottom of the hierarchy may be better evidence than something nearer the top.
The real underlying message is that, regardless of where a study design sits in the research hierarchy, research needs to be done well. It doesn’t matter if it’s a massive randomised control trial, if it’s done badly, it’s value must be reduced.
References
Murad, M.H. et al. (2016) ‘New evidence pyramid’, Evidence-based medicine, 21(4), pp. 125–127. doi:10.1136/ebmed-2016-110401