PRISMA 2020 requires a discussion of the limitations of the evidence and the limitations of the review process — as separate obligations. That split is worth keeping, because they lead to different next work: the first says the field does not know something, the second says this review could not establish it.
For most reviews this slot, not the summary, is the deliverable. A review that only reports what is known has produced a longer version of an introduction somebody already wrote.
Three kinds of gap worth separating:
- Not asked. Nobody has studied it. Cheap to state, and worth stating precisely enough that a reader could design the study.
- Asked and inconsistent. Studies disagree. The useful note says what differs between them — population, measure, scale — because that difference is usually the real finding rather than an annoyance.
- Asked and unanswerable as posed. The construct is not measured consistently enough across studies for the results to be comparable at all. This is the most useful gap to name and the least often named, because naming it means conceding the question needs rebuilding before more data would help.