9.2.5 Inclusion criteria

For the purposes of an Umbrella Review, the term “studies” refers exclusively to syntheses of research evidence including systematic reviews and meta-analyses. The “Inclusion criteria” of the protocol detail the basis on which studies will be considered for inclusion into the Umbrella Review and should be clearly defined.

These criteria provide a guide for the reader to clearly understand what is proposed by the reviewers and, more importantly a guide for the reviewers themselves to base decisions about the studies to be included in the Umbrella Review.

Types of participants

Important characteristics of participants should be detailed, including age and other qualifying criteria that make them appropriate for the objectives of the Umbrella Review and match the review question. In the example question above these characteristics include elderly people with dementia. Umbrella Reviews that aim to encompass multiple population groups should define each group clearly. Justification for the inclusion or exclusion of participants should be explained. In many cases, defining characteristics of the participants for a review may also include details of the setting of interest such as acute care, primary health care, or the community.

Interventions/phenomena of interest

The interventions or phenomena of interest for an Umbrella Review should be defined in detail and should be congruent with the review objective and intervention(s) or the phenomena of interest. Umbrella Reviews that aim to address multiple interventions and treatments should define each potential intervention of interest clearly.

Outcomes

Outcomes of interest should be predefined in Umbrella Reviews that lend themselves to quantitative evidence. Outcomes should be relevant to the question of the Umbrella Review and also the important outcomes for the participant group of the review. Surrogate outcomes should be explained and presented where there is a clear association with patient relevant outcomes. To provide a balanced overview of the evidence base related to a particular topic and fully inform decision-making, an Umbrella Review should attempt to report both beneficial and adverse outcomes.

Context

Context will vary depending on the objective(s)/question(s) of the review. The context should be clearly defined and may include but is not limited to consideration of cultural factors such as geographic location, specific racial or gender based interests. In some cases, context may also encompass details about the specific setting (such as acute care, primary health care, or the community).

Types of studies

As mentioned at the outset, the unit of analysis for an Umbrella Review is a completed research synthesis; therefore, the types of studies included in an Umbrella Review are exclusively syntheses of  existing research from  systematic reviews (using internationally accepted methodologies) and meta-analyses. Research syntheses included in an Umbrella Review should represent syntheses of empirical research evidence. Due to the enormous range of “review” types and articles available in the literature (Grant and Booth, 2009), authors of Umbrella reviews will have to stipulate clearly which review types should be included a priori in the protocol. Reviews that incorporate theoretical studies or text and opinion as their primary source of evidence should not be included in a JBI Umbrella Review and should be listed as an explicit exclusion criterion in the protocol.