The information you use in your research and study must all be credible, reliable and relevant. Part of the Evidence Based Medicine process is to critically appraise scientific papers, but in general, all the resources you refer to should be evaluated carefully to ensure their credibility.
Once you have asked the clinical question and searched for evidence, it's often not enough that you've checked for peer review if you want to find the very best evidence - it will ensure that studies with scientific flaws are disregarded, and the ones you include are relevant to your question.
In the Evidence Based Medicine process, and especially in the process of evaluating primary research (which hasn't be pre-appraised or filtered by others), we need to go beyond the usual general information evaluation and make sure the evidence we are using is scientifically rigorous. The main questions to address are:
For a more detailed look at Critical Appraisal, head to the Systematic Review Guide - Critical Appraisal and the Evidence-Based Practice Guide - Appraise.
Ask - is the evidence you have acquired valid? To what extent can the observed association between the exposure and the outcome be attributed to bias and confounding?
After selecting appropriate papers that address your specific clinical question, then you need to assess them for quality and bias.
1. You will need to read the full paper for each study carefully. Different study designs are prone to different types of bias. There are a number of tools which can help you to assess your papers for the main types of bias associated with each study design. Correctly identifying the level of evidence is important in deciding the baseline strength of the paper, before adjusting this assessment based on the internal validity of the study.
2. A key characteristics table can be a good way to organise and concisely summarise the information you have found, and the main findings from each study. Clearly summarising the population, study design, exposure/intervention and outcome measure for each paper helps to show how well the paper addresses your clinical question.
3. Synthesise the information, outlining the extent to which the association observed may be attributed to bias and confounding.
Synthesis means to discuss the evidence as a whole, explaining the main findings clearly and concisely, while weighing up the strengths and the weaknesses of the papers to help you to come to an overall conclusion which addresses the clinical question you originally posed – including consideration of the direction of effect, the size of the effect for each paper, overall strength of the papers you have chosen to include. It goes beyond just stating what the authors' findings were.
Meta-analyses aim to provide a numerical synthesis of the results, but this is only possible when equivalent outcome measures or reporting is used. When this is not possible, then a narrative synthesis is done.
The following article gives a good example of synthesising the evidence from 5 different studies, using the GRADE framework. You can see how the authors are giving a clear summary of how the papers perform against a number of criteria, comparing and contrasting the different papers, before coming to an overall conclusion as to how well the papers as a whole answer the clinical question.
Murad MH, Mustafa RA, Schünemann HJ, Sultan S, Santesso N. Rating the certainty in evidence in the absence of a single estimate of effect. BMJ Evidence-Based Medicine. 2017 Mar 20.
While there is no requirement to use the GRADE framework in your synthesis, this article gives a good idea of what is meant by synthesis of the information in the task description.
Fortunately, there have been some great checklist tools developed for different types of studies. Here are some examples:
Introduction to Critical Appraisal - This short video from the library at the University of Sheffield in the UK looks at the background to critical appraisal, what it is, and why we do it. A very useful introduction to the topic.
How to read a paper: The basics of evidence-based medicine and healthcare (6th ed.) by 
"Critical Appraisal" In Searching skills toolkit: Finding the evidence (2nd ed.) by 
"Critically Appraise the Evidence" In Evidence-based practice workbook: Bridging the gap between health care research & practice (2nd ed.) by Charles Sturt University acknowledges the traditional custodians of the lands on which its campuses are located, paying respect to Elders, both past and present, and extend that respect to all First Nations Peoples.
Charles Sturt University is an Australian University, TEQSA Provider Identification: PRV12018. CRICOS Provider: 00005F.
