How are health research partnerships assessed? A systematic review of outcomes, impacts, terminology and the use of theories, models and frameworks
Kelly J Mrklas,Sera Merali,Masood Khan,Sumair Shergill,Jamie M Boyd,Lorelli Nowell,Lisa M Pfadenhauer,Kevin Paul,Amelia Goertzen,Liam Swain,Kathryn M Sibley,Mathew Vis-Dunbar,Michael D Hill,Shelley Raffin-Bouchal,Marcello Tonelli,Ian D Graham
DOI: https://doi.org/10.1186/s12961-022-00938-8
2022-12-14
Abstract:Background: Accurate, consistent assessment of outcomes and impacts is challenging in the health research partnerships domain. Increased focus on tool quality, including conceptual, psychometric and pragmatic characteristics, could improve the quantification, measurement and reporting partnership outcomes and impacts. This cascading review was undertaken as part of a coordinated, multicentre effort to identify, synthesize and assess a vast body of health research partnership literature. Objective: To systematically assess the outcomes and impacts of health research partnerships, relevant terminology and the type/use of theories, models and frameworks (TMF) arising from studies using partnership assessment tools with known conceptual, psychometric and pragmatic characteristics. Methods: Four electronic databases were searched (MEDLINE, Embase, CINAHL Plus and PsycINFO) from inception to 2 June 2021. We retained studies containing partnership evaluation tools with (1) conceptual foundations (reference to TMF), (2) empirical, quantitative psychometric evidence (evidence of validity and reliability, at minimum) and (3) one or more pragmatic characteristics. Outcomes, impacts, terminology, definitions and TMF type/use were abstracted verbatim from eligible studies using a hybrid (independent abstraction-validation) approach and synthesized using summary statistics (quantitative), inductive thematic analysis and deductive categories (qualitative). Methodological quality was assessed using the Quality Assessment Tool for Studies with Diverse Designs (QATSDD). Results: Application of inclusion criteria yielded 37 eligible studies. Study quality scores were high (mean 80%, standard deviation 0.11%) but revealed needed improvements (i.e. methodological, reporting, user involvement in research design). Only 14 (38%) studies reported 48 partnership outcomes and 55 impacts; most were positive effects (43, 90% and 47, 89%, respectively). Most outcomes were positive personal, functional, structural and contextual effects; most impacts were personal, functional and contextual in nature. Most terms described outcomes (39, 89%), and 30 of 44 outcomes/impacts terms were unique, but few were explicitly defined (9, 20%). Terms were complex and mixed on one or more dimensions (e.g. type, temporality, stage, perspective). Most studies made explicit use of study-related TMF (34, 92%). There were 138 unique TMF sources, and these informed tool construct type/choice and hypothesis testing in almost all cases (36, 97%). Conclusion: This study synthesized partnership outcomes and impacts, deconstructed term complexities and evolved our understanding of TMF use in tool development, testing and refinement studies. Renewed attention to basic concepts is necessary to advance partnership measurement and research innovation in the field. Systematic review protocol registration: PROSPERO protocol registration: CRD42021137932 https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=137932 .