Below, we describe our methods for obtaining internal team input, conducting key informant interviews, reviewing the peer-reviewed and grey literature, and identifying and assessing relevant tools and interventions.
Internal Team Input
Because our project team represents the perspectives of patients, families, hospitals, health professionals, and health care organizations, we gathered input from internal team members to refine search strategies, identify tools and resources, and ensure equal input from all partners. We talked with representatives from each organization represented on the project Steering Group: Consumers Advancing Patient Safety (CAPS); the Institute for Family-Centered Care (IFCC); Carilion Clinic; the Joint Commission; the Health Research and Educational Trust (HRET); Aurora Health Care; Planetree; and the Maryland Patient Safety Center (MPSC). We also spoke with our two patient advisor Steering Group representatives. These informal conversations focused on the following topics: conceptualization of patient and family engagement around patient safety and quality; the perspectives and behaviors of patients, families, and health care professionals related to patient and family engagement; hospital interventions and strategies for patient and family engagement—what works and what does not work; recommendations for research (peer-reviewed and grey literature), tools, and resources to include in the scan; and recommendations for key informant interviews.
Key Informant Interviews
We conducted eight telephone interviews with key informants; one interview was still pending at the time this report was prepared, owing to scheduling conflicts. Based on conversations with our internal project team and in consultation with AHRQ, we developed a list of key informants who were knowledgeable about patient and family engagement and patient- and family-centered care and who could contribute a diverse and complementary set of perspectives. Our key informants were a patient, a patient and family educator, a hospital administrator, a physician, a nurse, a representative from the Veteran's Health Administration, an expert in engagement and other initiatives, and an expert in safety and quality measurement.
For these key informant interviews, we developed a semi-structured interview protocol (Appendix A). The protocol focused on the following areas:
- Patient, family, and health care professional interventions to encourage engagement, along with barriers and facilitators to these interventions.
- Educational strategies that are most appropriate for each audience.
- Tools and resources that the interviewee uses or has used to engage patients and families and obtain feedback on their effectiveness.
- Priority areas in which the Guide could make a difference (i.e., where it can effect change and in what topic and content areas patients, families, and health professionals most need assistance).
All interviews were audiotaped, and detailed notes were taken. Analysis focused on major themes related to information needs, priorities, intervention strategies, and format preferences.
Here, we describe the search strategies used for the peer-reviewed literature, grey literature, and tools.
The team scanned the medical and social science peer-reviewed literature—including both descriptive qualitative and quantitative studies—using PubMed®, PsycInfo, CINAHL, and Cochrane databases. We searched these databases using keyword search terms and MeSH headings (in PubMed®) for each category of interest. The searches were limited to English-language articles published since 1999 and included resources developed domestically and internationally—especially in the United Kingdom, Australia, and Canada. We reviewed 6,315 abstracts that resulted from these searches. We retrieved 303 articles and reviewed and abstracted 188 of these. Detailed information about search terms is presented in Appendix B.
We conducted an initial review of all abstracts identified through searches of the databases listed above. Using the criteria noted in Exhibit 3, we determined which abstracts were of higher priority for review. Next, we retrieved the full text and reviewed each article. This review allowed us to further assess whether the article would be relevant in determining the content for the Guide, and if found to be relevant, we conducted an in-depth data abstraction of its content. We also reviewed reference lists in the articles that were retrieved and abstracted for relevant citations. In this review, we included significant (i.e., seminal) articles from prior to 1999, if appropriate. As we analyzed the literature, we conducted additional literature searches and reviewed articles to address possible gaps in knowledge. These additional articles were reviewed but not abstracted in the Access database.
We defined "grey literature" as any non-peer-reviewed literature that met the study's inclusion criteria noted in Exhibit 3. This included presentations, articles, press releases, white papers, trade publications, issue briefs, and book anthologies. We did not include articles from magazines or newspapers.
To scan the grey literature, we gathered documents from AHRQ, the project team, and external experts during the key informant interviews. In addition, we conducted a more targeted search of 110 Web sites recommended by our partners for non-peer-reviewed literature and potential tools (Appendix C presents a list of Web sites reviewed). We also reviewed citation lists or bibliographies related to patient and family engagement found on these Web sites for additional peer-reviewed or grey literature.
The grey literature scan also included an inventory of tools that could be included in the Guide. Tools included handouts, flyers, brochures, posters, videos, and guides that addressed opportunities for patient and family engagement in safety and quality. We downloaded and reviewed tools that were available in the public domain and could be applied to the hospital setting. Tools were assessed on our ideal criteria for inclusion in the Guide: focusing on a hospital setting; reflecting target audience needs and priorities (user-centered); being actionable, that is, focusing on specific behaviors; having been developed with input from the target audience; and having been evaluated for effectiveness and feasibility in a hospital setting. However, we did not exclude tools that did not meet these criteria.
Data Abstraction and Analysis
Peer-Reviewed and Grey Literature
All documents were entered into a Reference Manager® database. We exported the references from Reference Manager into an Access® database to abstract key dimensions and identify key themes; the data abstraction protocol is shown in Appendix D. We developed an abstraction form in Microsoft® Access® for the peer-reviewed and grey literature. The main elements of the abstraction form were document information, purpose, methods, and main findings. Exhibit 4 shows the numbers and types of documents abstracted. We also noted the type of documents found in the literature (e.g., systematic reviews, single descriptive studies). Each findings section includes an exhibit that summarizes the strength of the evidence of documents reviewed in that section.
Once abstracted, the information was synthesized by main themes related to the conceptual framework: target audience perspectives, organizational context, and strategies and interventions.
As with the literature review, we entered all tools and resources into a Reference Manager database and then exported these resources to the Access database. Once the tools were entered into the Access database, we classified them according to predefined dimensions. We developed a form in the Access database specifically for tools. The form included general information about the tool (e.g., content, target audience, mode of delivery) and also addressed the usability of the tool (e.g., literacy/use of plain language, organization of information, format, layout, cultural inclusivity, languages available, developed with input from target audiences, evaluated for effectiveness). We assessed the content and usability of approximately 80 tools. For the remaining tools, we assessed content but did not address usability. We will conduct a usability assessment on the remaining tools, depending on their correspondence to the priority topic areas identified for inclusion in Task 3. Exhibit 5 shows the numbers of tools abstracted.
The analysis of the tools identified what existing materials can be adapted or modified and what new materials need to be developed. We reviewed materials to identify areas in which resources are limited or nonexistent and areas in which tools will need to be revised or updated if we want to focus on a particular substantive area or on a particular point in the care process.