BackgroundPatient and public involvement (PPI) is a prerequisite for many funding bodies and NHS research ethics approval. PPI in research is defined as research carried out with or by the public rather than to, about or for them. While the benefits of PPI have been widely discussed, there is a lack of evidence on the impact and outcomes of PPI in research.ObjectivesTo determine the types of PPI in funded research, describe key processes, analyse the contextual and temporal dynamics of PPI and explore the experience of PPI in research for all those involved. Mechanisms contributing to the routine incorporation of PPI in the research process were assessed, the impact of PPI on research processes and outcomes evaluated, and barriers and enablers to effective PPI identified.DesignA three-staged realist evaluation drawing on Normalisation Process Theory to understand how far PPI was embedded within health-care research in six areas: diabetes mellitus, arthritis, cystic fibrosis, dementia, public health and learning disabilities. The first two stages comprised a scoping exercise and online survey to chief investigators to assess current PPI activity. The third stage consisted of case studies tracked over 18 months through interviews and document analysis. The research was conducted in four regions of England.ParticipantsNon-commercial studies currently running or completed within the previous 2 years eligible for adoption on the UK Clinical Research Network portfolio. A total of 129 case study participants included researchers and PPI representatives from 22 research studies, and representatives from funding bodies and PPI networks.ResultsIn the scoping 51% (n = 92) of studies had evidence of PPI and in the survey 79% (n = 80), with funder requirements and study design the strongest influence on the extent of PPI. There was little transparency about PPI in publicly accessible information. In case studies, context–mechanism–outcome configurations suggested that six salient actions were required for effective PPI. These were a clear purpose, role and structure for PPI; ensuring diversity; whole research team engagement with PPI; mutual understanding and trust between the researchers and lay representatives; ensuring opportunities for PPI throughout the research process; and reflecting on, appraising and evaluating PPI within a research study. PPI models included a ‘one-off’ model with limited PPI, a fully intertwined model in which PPI was fully embedded and an outreach model with lay representatives linking to broader communities. Enabling contexts included funder, topic/design, resources, research host, organisation of PPI and, most importantly, relationships. In some case studies, lack of coherence in defining PPI persisted, with evidence of a dual role of PPI representative/study participant. Evidence of PPI outcomes included changes to study design, improvements to recruitment materials and rates, and dissemination.ConclusionsSix salient actions were required for effective PPI and were characterised by a shared understanding of moral and methodological purposes of PPI, a key individual co-ordinating PPI, ensuring diversity, a research team positive about PPI input and fully engaged with it, based on relationships that were established and maintained over time, and PPI being evaluated in a proactive and systematic approach. Future work recommendations include exploring the impact of virtual PPI, cost analysis and economic evaluation of the different models of PPI, and a longer-term follow-up study of the outcomes of PPI on research findings and impact on services and clinical practice.FundingThe National Institute for Health Research Health Services and Delivery Research programme.