BackgroundInaccurate, false or incomplete research publications may mislead readers including researchers and decision-makers. It is therefore important that such problems are identified and rectified promptly. This usually involves collaboration between the research institutions and academic journals involved, but these interactions can be problematic.MethodsThese recommendations were developed following discussions at World Conferences on Research Integrity in 2013 and 2017, and at a specially convened 3-day workshop in 2016 involving participants from 7 countries with expertise in publication ethics and research integrity. The recommendations aim to address issues surrounding cooperation and liaison between institutions (e.g. universities) and journals about possible and actual problems with the integrity of reported research arising before and after publication.ResultsThe main recommendations are that research institutions should: develop mechanisms for assessing the integrity of reported research (if concerns are raised) that are distinct from processes to determine whether individual researchers have committed misconduct;release relevant sections of reports of research integrity or misconduct investigations to all journals that have published research that was investigated;take responsibility for research performed under their auspices regardless of whether the researcher still works at that institution or how long ago the work was done;work with funders to ensure essential research data is retained for at least 10 years.Journals should: respond to institutions about research integrity cases in a timely manner;have criteria for determining whether, and what type of, information and evidence relating to the integrity of research reports should be passed on to institutions;pass on research integrity concerns to institutions, regardless of whether they intend to accept the work for publication;retain peer review records for at least 10 years to enable the investigation of peer review manipulation or other inappropriate behaviour by authors or reviewers.ConclusionsVarious difficulties can prevent effective cooperation between academic journals and research institutions about research integrity concerns and hinder the correction of the research record if problems are discovered. While the issues and their solutions may vary across different settings, we encourage research institutions, journals and funders to consider how they might improve future collaboration and cooperation on research integrity cases.