Skip to main content
Drexel Library

REFRESH Online Module

Traditional Evidence Pyramid

This pyramid shows various types of evidence, all of which can be used for evidence-based practice. The key is finding what information is available for the topic you are researching. Traditionally, this image can be broken down into two groups, filtered information and unfiltered information.

Unfiltered information sources are things like randomized controlled trials, cohort studies, or case reports. Filtered information sources are things like meta-analysis and systematic reviews, which take all available unfiltered information and focus it.

So what’s the difference between a meta-analysis and a systematic review? A systematic review is created by gathering, appraising, and summarizing all available evidence. A meta-analysis goes a step further and uses a statistical procedure for combining numerical data from multiple separate studies.

This pyramid can be a little misleading, because if automatically gives the connotation that some types of information are inherently superior to others. This is not always true.

A New Pyramid

Consider this pyramid. It gives a better idea of the purpose behind systematic reviews and meta-analysis. They are used, almost like a magnifying glass, to look at all available evidence. In this case, all evidence is approached equally. After all, sometimes a well-designed cohort study is more helpful than a poorly conducted randomized controlled study. Other times, certain types of studies may not be available for your research, depending on population groups. For example, randomized controlled studies are rarely done on pregnant women. That is why it is so important to examine all available evidence. Not all evidence is created equally.

(http://ebm.bmj.com/content/21/4/125)

When examining your evidence, you want to keep two questions in mind. How high is the likelihood that the effect of the treatment will achieve clinically relevant benefits (or harms)? And "How appropriate is the outcome measure for the healthcare problem, and how useful is it in measuring the benefits (or harms) of the treatment?”

For example, in the case of asthma-practice guidelines, the US FDA has required increases in lung function and reduction of exacerbation as primary outcomes in clinical trials of new asthma drugs. However, the NIH asthma guidelines also suggest the use of patient-reported outcomes, including health-related quality-of-life measures, to assess asthma control. So when looking at evidence, it is important to consider the outcome measure being described.

https://doi.org/10.1586/erp.11.48

http://consumers.cochrane.org/levels-evidence

DynaMed Levels of Evidence

Fortunately, there is a rating system for levels of evidence. Levels of evidence are a hierarchical system of classifying evidence. They can be helpful in determining which evidence is most authoritative, by considering things like research study design, how rigorous the methodology was, and other contributing factors. In this instance, level 1 is the highest. This photo is an example from DynaMed, a resource we will explore more later. 

http://www.dynamed.com/home/content/levels-of-evidence

In addition, some professional organizations, resources, or associations also create their own levels of evidence. Some examples are the American Academy of Pediatrics or the American Association of Critical Care-Nurses. Take a moment to open a new tab in your browser and search for the levels of evidence guidelines that are used by the main society or association that is related to your area of expertise.