Knowledge synthesis is often required to describe what is known on a particular topic and identify the need for further research. We will conduct a realist review based on the methods described by Pawson et al. . This five-step approach was developed to explore how complex interventions work for whom and in what contexts. A realist review focuses on describing theoretical and contextual details about why the intervention did or did not work that can be used to fine-tune its design. It draws upon a wide range of quantitative and qualitative study designs. This approach was recently used to examine the interaction between context, intervention, outcome, and underlying theory to understand the efficacy of school feeding programs .
To be feasible, realist reviews must be bounded by focusing the question on either particular processes or groups of recipients; thus, we will restrict our review to arthritis and cancer. Both represent prevalent conditions that have generated considerable research on patient involvement through education and self-management, but systematic reviews revealed variable impacts of these outcomes and called for further investigation of the factors that influence their effectiveness [15–17]. By focusing on two conditions, the review will be limited to a manageable number of studies, while still allowing for comparison by patient and contextual factors.
Step 1: clarify scope--refine purpose of review/key theories to be explored
To refine the research questions and theories of interest, we will consult with stakeholders, including arthritis and cancer researchers, clinicians, managers, and patients, after first conducting an exploratory scan of the literature. A useful starting point is provided by another realist-inspired analysis in which 26 behaviour-change techniques were identified in a Cochrane Library review of interventions to promote physical activity . In this study the taxonomy was validated by using it to code strategies and associated theories in a review of studies to encourage healthy eating. Additional theories relevant to patient-mediated KT will be assembled by searching indexed sources of literature, including MEDLINE and CINAHL, for [(models, theoretical or models, educational or models, psychological) AND patient education as topic or (information dissemination and patient participation)]. The research team will review the assembled theories to refine review questions, guide the selection of relevant theories, and confirm or expand the conceptual framework upon which a more comprehensive literature review will be based.
Step 2: search for evidence
A comprehensive literature search developed by an information specialist will be conducted by using several indexed sources. Search strategies will combine concepts reflecting [(arthritis or neoplasms) AND patient education as topic or (information dissemination and patient participation)]. Searches will be executed for the years 1995 to current to encompass a nearly 15-year span during which research on patient involvement became prevalent. Databases include MEDLINE (North American), the Cochrane Library (systematic reviews, trials), EMBASE (European), and CINAHL (nursing, allied health). To augment these searches, we will examine five years of tables of contents for journals likely to publish patient-mediated KT interventions, including Patient Education and Counseling, Health Expectations, Implementation Science, Journal of Cancer Education, Psycho-Oncology, Arthritis Care & Research, and Communication & Medicine. To ensure that all relevant literature is captured, we will scan the references of eligible studies. Quantitative (meta-analyses, systematic reviews, guidelines, surveys, observational studies, randomized trials) or qualitative (interviews, focus groups) studies published in the English language that focus on developing, implementing, or evaluating patient-mediated KT interventions are eligible. Abstracts, letters, or editorials are ineligible. Two individuals will independently review titles and abstracts and select articles for inclusion based on eligibility criteria using a screening tool. Articles selected by at least one reviewer will be retrieved since ultimate judgment about inclusion must be reserved until the full text is examined.
Step 3: extract data and appraise primary studies
A data-extraction form will be developed based on the refined version of the conceptual framework (Step 1) to collect information on intervention (format, content), outcome (intended, reported), mechanism of action (theory explicitly reported by authors or referred to implicitly in objectives or methods), and application (clinical indication, setting of care, attributes of those delivering and receiving the intervention). As a pilot, data will be extracted independently by the principal investigator and research associate for 10 randomly selected articles. They will compare congruence of extracted data and determine whether and how the form should be revised, then independently extract data from remaining studies. Most details will be noted on the form by checking the appropriate box. Qualitative details, including description of implicit theory, will be highlighted in the article, which will be copied and attached to the data extraction form. Study quality will be assessed using criteria relevant to study design to describe the nature of this literature but will not be used to exclude studies from review [18–20].
Step 4: synthesize and interpret
The research associate will tabulate extracted/highlighted quantitative and qualitative data, noting any differences between independently extracted information for the same article and resolving those differences through discussion with the principal investigator. The total number of eligible and included studies from each source will be reported, along with reasons for exclusions. Tabulated findings will be examined to discuss the quantity, design, and quality of studies. The nature of patient-mediated KT interventions will be described according to the elements of the refined conceptual framework, including purpose, context, recipient characteristics, intervention design and delivery, explicit theory, and outcomes. Contextual information will be further examined thematically according to May's narrative review approach . This involves directly summarizing relevant details as they are reported to identify recurring or important issues, without any attempt to transform them into a common metric or interpreted theme as in a standard qualitative analysis. Qualitative details will be independently examined by the principal applicant and research associate. They will compare findings and resolve differences through discussion. Findings will be summarized to describe interventions, how they work, for whom, and in what context and identify explicit and implicit theories relevant to interventions. These data will be used to expand the conceptual framework and to create a separate taxonomy of patient-mediated KT strategies and associated relevant theories.
The research team will review and interpret the findings and confirm or further refine the products, which include the following: (a) a conceptual framework of patient-mediated KT interventions and outcomes; (b) a description of patient-mediated KT interventions and the degree to which they have been evaluated in different settings or patients, highlighting key elements of design or implementation that contribute to or detract from their impact; (c) a taxonomy listing the variety of patient-mediated KT interventions and associated theories; (d) recommendations for systematic review of particular patient-mediated KT interventions where evidence is found to be sufficient; and/or (e) identification of research gaps that warrant further investigation through primary research study by comparing findings and the nature/quality of that evidence to concepts in the conceptual framework.