You should have an RIS file from each of your three databases.
Before uploading your RIS files to Rayyan, name the file in a way that makes it clear which database it's from and the date of your search. For example, my search in Academic Search Premier will be named ASP_Mar15.ris.
We will do two rounds of screening: an initial screening, in which we screen the abstracts only, and a final screening, in which we screen the full text.
Rayyan is the tool we'll be using to screen our articles. Start by signing up for a free account. Then create a new review.
You will be prompted to enter a title and then upload your references, which are the RIS files you exported from each database. They should be named after the database and search date before you upload. You do not need to invite members.
The first step is to de-duplicate your articles, in case some of the same articles showed up in your searches from different databases. Click "detect duplicates."
If duplicates were detected, click "start resolving" in the "unresolved duplicates" box. You may need to wait a few minutes and refresh the website if it doesn't come up.
The two possible duplicates will appear. If you can see that they are the same article, you can decide which one you want to keep. Look for the one that has the most complete information (i.e. not missing things like publication date). The red text shows the differences between the two. If they look like different articles, click "keep both."
For our initial screening, we will be reviewing article abstracts to see whether they meet our inclusion criteria (PICOS, see documentation worksheet from first meeting).
Go to the "screening" tab at the top of the page, which should be default take you to your "undecided" articles.
Because we are doing a systematized review, rather than a full systematic review, we are going to start by screening the most recent articles first, and continue until we get about 10-15 articles.
Start by sorting the sources by publication date, newest first.
The articles with no publication date will come up first. Try to Google these to find the publication date.
Select the first article that comes up and read the abstract. There are buttons to include or exclude the article. This decision should be made based on your PICOS. Is the population correct? Is the intervention correct? Is the comparison correct? Is the outcome measured correct? Is the study type correct?
If you exclude an article, you always need to record the reason in the reason box, rather than just clicking "exclude." You can choose from the reasons offered, or add your own. Otherwise, choose "wrong population," "wrong study type," etc. You only have to choose one reason for exclusion, so as soon as you identify a reason to exclude a source, mark it and move on.
If you are considering limiting your articles by geography (this may or may not be appropriate, depending on your research question), mark articles outside your potential geographic area of focus as "maybe." You can revisit these later, once you have decided whether to limit by geography.
At this stage, your goal is to screen until you've included about 10-15 articles. Important: Your decisions about which articles to include and exclude should be made by consistently applying your inclusion/exclusion criteria (PICOS). We are NOT making inclusion/exclusion decisions based on our subjective criteria for what we think sounds interesting, fits our topic, or will help us write our papers.
You can check how many articles you've included using the drop down arrow by "undecided."
If you've reached your 10-15 included articles, continue screening the rest of the articles published in that year so that your date range has a clean cut-off point.
Revisit any "maybes" to determine whether or not to include them. If you don't have enough information from the abstract to determine whether it meets your criteria, include it. You will do a more thorough screening in the next step.
For our final screening, we will be reviewing the full text of the articles to see whether they meet our inclusion criteria. Most of this information will be in the methods sections.
Click on the "full text screening" tab at the top of the page. Click to "enable" full text screening, and then "add data." You should add your included articles and your maybe articles, if you have any. This will import all of the data from those articles into your full text screening section.
You will need to obtain the full text for all of the articles you will be screening in this round. Start by opening on an article record and clicking on the listed URL. If that doesn't work, here's how to obtain full text:
You want to upload the full text PDFs into your full text review in Rayyan so that they are easily accessible. Select an article and use the "upload" button.
Once the PDFs are uploaded, you can start screening by clicking the PDF link in the article record, which opens the PDF in Rayyan. Without the PDF, it doesn't allow you to make a screening decision.
Based on the methods section of the paper, decide whether each paper meets all your PICOS criteria, and exclude those that don't.
You can use the notes field at the bottom to record the population, intervention, outcomes, study type, and other relevant information.
Use the include and exclude buttons to record your decisions, and make sure to mark reasons for exclusion. Through this round of screening, your criteria should narrow down your final group of articles to 5-8. If you have too many, you can exclude the oldest for being "outside of date range," though make sure you have a clean "cut off" year.
A PRISMA flow diagram will be included in the methods or results section of your paper and will document all of your steps that brought you to the final set of articles that you will analyze. You can see that you'll need to record the necessary numbers to populate this table.
Here's an example of a PRISMA diagram:
You can generate your diagram manually, or using a PRISMA flow diagram generator, such as this one.