Making a systematic literature review is large feat, and figuring out how to start can be tough. We have collected a handful of resources to help you out.
How do I create a search string?
Once you have figured out what to research, the next step is to find relevant literature from databases. We think this video from Johns Hopkins University does an excellent job at describing the process from research question to search string.
How do I get started?
Unsure of whether to do a systematic literature review, a scoping review, or a narrative review? Finding out what type of literature review to do can be a challenge in itself.
The Charles Sturt University Library has made an informative overview explaining the different types of literature reviews and their different purposes.
A full systematic literature review is one of the most rigorous academic disciplins, and following it from start to publication can be quite complicated. To get a full guide, we recommend looking at these sources:
If you want to do a Meta-analysis but are unsure of it is the right choice, Cochrane South Africa has made a thorough guide to meta-analyses, including when these are appropriate.
Risk of Bias tools
Not all scientific studies are created equal. Thus, a risk of bias assessment might be important for your project. There are many tools to report the risk of bias. We want to highlight RoB2.
RoB2 (Risk of Bias 2) is the risk of bias tool developed by Cochrane. RoB2 is specifically developed for randomized control trials (RCTs), but versions for other types of studies such as cluster-randomized trials and crossover trials have been developed as well. It focused on factors such as trial design, trial conduct, and data reporting. After the judgement of all questions in RoB2, the RCT is either judged as high risk of bias, low risk of bias, or some concerns.
To access RoB2, click the button below.
If you need a quick overview of the estimated evidence quality level, we recommend the Oxford Centre Levels of Evidence.
The Oxford Centre Levels of Evidence, last updated in 2009, were a guide created for clinicians who do not usually have the time to evaluate the risk of bias before making clinical decisions. The guideline rates evidence types on a scale from 1-5 based on five different outcome types in the studies: Therapy, Prognosis, Diagnosis, Differential Diagnosis, and Economic and Decisions Analysis.
Learn more about the Oxford Centre Levels of Evidence here:
If you are doing a review of reviews, we recommend QUADAS-2.
Some risk of bias tools may not be applicable for reviews, as one is not necessarily addressing the concrete data, but how the review was conducted and the data for it was chosen. QUADAS-2 is a specific risk of bias tool for reviews. It has four domains: patient selection, index test, reference standard, and flow and timing. It is meant to be tailored to each specific review, leaving a lot of flexibility.
Read more about QUADAS-2 here: