Our search generated 3,803 articles, which after removing 711 duplicates, left a total of 3,092
articles. Next 1807 abstracts where both reviewers agreed on exclusion were excluded.
Abstracts where there was disagreement or uncertainty, or which were selected for inclusion
(1285) were re-checked and resulted in the removal of an additional 763 abstracts, leaving 522articles for full-text examination (26 review articles and 496 studies). At the full-text reading
stage, an additional 236 studies were excluded after being examined by two reviewers, leaving
260 studies with some level of community participation in health systems research studies.
This included articles that aimed to engage communities more fully, but failed to do so (Fig 1
and S1 Table).