Screening decisions

The platform suggests, you decide

For each paper, the platform suggests a label against your written criteria and gives a reason. The suggestion is only a proposal: a person decides every paper.

When you change a label, both the suggestion and your decision are kept, and your decision is the one that counts. A later reader can see where people and suggestions disagreed.

Three labels

A paper is labelled from its title and abstract only, against this project’s criteria only. The same paper can be relevant in one project and irrelevant in another.

• Relevant: meets the inclusion criteria on the information available.
• Irrelevant: meets an exclusion criterion, which is named in the reason.
• Maybe: the title and abstract are not enough to decide. This is a valid answer.

Papers that need a closer look

A paper is flagged for careful reading when it has no abstract, when the suggestion is "maybe", or when the confidence of the suggestion is below the project threshold. A paper with no abstract cannot be judged from its title alone, so get its full text first.

You can filter the list to unscreened, flagged or reviewed papers, or to a single label, and work through one group at a time.

When you keep disagreeing

Many changes in the same direction usually mean the criteria are unclear. Fix the criteria in the study configuration and run screening again instead of correcting papers one at a time.

Running screening again does not delete earlier labels. Each run is stored with the criteria it used, so two runs can be compared.