Eligibility criteria have to be predefined. Some criteria are easy to define (e.g., language, time frame, etc). However, it is sometimes not possible to predefine all the criteria and you may need to read the first references retrieved by the search strategy to refine them.

Most of the time, we define the eligibility criteria (both inclusion and exclusion) for all type of studies:

• inclusion criteria
• exclusion criteria

Sometimes, it is easier to present the eligibility criteria according to the type of studies:

• Eligibility criteria for QUAN studies (or QUAN component of mixed methods studies)
• Inclusion criteria
• Exclusion criteria

• Eligibility criteria for QUAL studies (or QUAL component of mixed methods studies).
• Inclusion criteria
• Exclusion criteria

Inclusion criteria

It is easier to decide up-front what will be the inclusion criteria as they are directly linked with the research question or objectives.

The most popular way to define the inclusion criteria is by using the PICO structure. In addition, type of studies, time frame (if any) and language(s) should be specified.

Examples of inclusion criteria

• Intervention or process or topic. It has to be clearly defined in terms of components. Indeed, if you are interested in “transitional care interventions”, you want to include studies that share the same component but have other names.
• Population targeted by the intervention or process or topic: patients, clinicians, etc. You can also specify the country and the setting.
• Type of study: empirical study using quantitative, qualitative or mixed methods design.
• If needed: Type of participants. Sometimes, it is useful to specify the type of participants according to each study type in addition to the targeted population. For instance, if you are interested in interventions aimed at reducing teenage pregnancy, the target population is teenagers. However, if you want to assess barriers to implementation of these interventions from the point of view of clinicians, your  type of participants is thus clinicians. Hence, you might be interested in two types of participants:  
• Type of participants for QUAN studies  (or QUAN component of mixed methods studies): teenagers
• Type of participants for QUAL studies (or QUAL component of mixed methods studies): clinicians
• Outcomes (if any)
• Language(s)
• Time frame (to be justified)

Exclusion criteria

The exclusion criteria are harder to define. Most of the time, we refine them after reading the first articles retrieved by the search strategy. For instance, if you are interested on the impact of primary care models for patients with a specific chronic disease, you may retrieve many articles focusing on education (initial or continuing) of family physician and then decide which interventions should be part or not part of your review. 

Examples of exclusion criteria:

• Studies not on the topic-intervention-process. This criterion can be refined as the review progresses (see example  about education above).
• Type of papers. You may decide to reject all non-empirical study:     

• Editorials, description of an intervention (without empirical data, no data collection, and no method), methodological paper (e.g., validity of a tool that is often used in the studies you want to include, if your review does not focus on this tool). Also, you may or may not want to exclude the following articles: abstract poster, thesis, dissertation, history of a case (case report).
• Literature reviews, systematic review and meta-analysis: it is useful to identify these articles using a specific exclusion code because you may want to use them for the background or the discussion.
• Protocol: it is useful to exclude these articles using a specific code because you may want to check if the results of the study have been published in order to update your review.
• If you are interested in a specific intervention, add an exclusion criterion for all studies with no intervention (e.g., epidemiology, statistics on services usage, costs of illness, current practices, etc.).