In WP 1, we will conduct a scoping review, which adheres to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analysis Extension for Scoping Reviews) [33] and follows the updated methodological guidance for the conduct of scoping reviews proposed by the JBI (formerly Joanna Briggs Institute) [34].

To provide an overview of methods to foster SDM via PDAs in the CPG development process, we will use the GIN Public Toolkit—Patient and Public Involvement in Guidelines as a starting point [4]. To account for the progress after the publication of the GIN Public Toolkit, we will update the current state of knowledge by conducting a scoping review in relevant scientific databases (PubMed and Embase; see Multimedia Appendix 1 for full search strategy). To account for research-relevant information that is not formally published [35], we will additionally search in Google and Google Scholar. The search will be restricted to studies published since 2000, as the link between CPGs and PDAs is relatively new [3,4]. References will be eligible for inclusion if they meet all the following criteria: they are published from the year 2000 onward, provide information about the methodology of developing and integrating PDAs in the context of CPGs, are available in German or English, and have full texts available. The exclusion will apply to studies not focusing on PDAs but on other patient-directed knowledge tools, such as the patient version of CPGs.

As part of the scoping review, handbooks, method manuals, and other documents of international CPG organizations that are known for providing patient-directed knowledge tools [36] (see Multimedia Appendix 2 for a list of organizations) will be sought via their websites. To obtain additional information on methods for the development and integration of PDAs in the context of CPGs, we will contact experts via the earlier-mentioned CPG organizations [36] and via GIN. Ideally, this will help to identify CPGs that have already systematically integrated PDAs, as the applied methods for doing so cannot be easily identified using other research techniques. After removing potential duplicates, 2 reviewers (LF and LVS) will independently screen all identified literature based on title and abstract for eligibility. They will independently continue with a full-text screening of references found to be eligible at the title and abstract screening stage. A list with the reasons for excluding full texts will be compiled, while disagreements will be resolved through discussion and by contacting a third reviewer (DP) for advice.

One reviewer (LF) will extract findings into standardized tables, which the second reviewer (LVS) will double-check. The extraction tables will contain categories of general study and method manual characteristics, such as title, publication year, authors, and their corresponding CPG organization. Additionally, information concerning the methods for developing and integrating PDAs in the context of CPGs will be extracted. Disagreements will be again resolved through discussion, and consensus will be reached by the judgment of a third reviewer (DP). The generated information will be narratively synthesized.

In WP 2, we aim to explore experts’ perspectives on the identification of preference-sensitive CPG recommendations and the successful integration of SDM strategies into CPGs by conducting a qualitative study via semistructured interviews. We will use COREQ (Consolidated Criteria for Reporting Qualitative Research) for planning and reporting [37].

For recruitment, we will develop a sample plan to ensure that relevant experts of national and international CPG organizations are adequately presented (ie, selective sampling). By contacting the 17 CPG organizations mentioned earlier [36] and potential other identified relevant organizations, we will continue to recruit experts until we reach data saturation. However, we expect that 10-15 experts will be sufficient to reach data saturation. Their expertise concerns the areas of CPGs, PDAs, and SDM. For example, experts can be either methodologists or health care professionals involved in CPG development.

The consortium and all collaborators will conceptualize an interview guide before the interview. The developed interview guide will be piloted in the first interviews and adapted if necessary. The guide will include preformulated open questions related to the project topic but will focus in particular on questions concerning (1) What is meant by preference sensitivity? and (2) How can SDM be successfully integrated into CPGs? All interviewees will receive the interview guide in advance.

The interviews will be conducted by telephone or videoconference and digitally recorded. The interviewer will take notes during or immediately after the interview. Conducted interviews will be transcribed, coded, and analyzed using MAXQDA analysis software (standard version 18.2.4; Verbi GmbH). The interviews will be analyzed using qualitative content analysis according to Mayring [38]. A category system will be developed deductively based on the core themes or questions. Further subcategories may be developed inductively. Category building will be discussed within the research team.

WP 3 will start with a workshop to synthesize the results of WP 1 and preliminary results of WP 2, based on which a multistage Delphi study including Delphi focus groups will be conducted. As there has been limited research in the area of preference sensitivity and prioritization of CPG questions with regard to the provision of PDAs, WP 3 aims (1) to provide a clear and generic definition of preference sensitivity and (2) to identify clear criteria for prioritizing CPG recommendations where PDAs may be required. To systematically develop statements on the topics of preference sensitivity and prioritization of CPG questions regarding the provision of PDAs, all members of the core team and the patient partners of the project will participate in the workshop. The subsequent modified Delphi study will be based on the internationally proven RAND/UCLA Appropriateness Method [39]. We will use ACCORD (Accurate Consensus Reporting Document) for reporting [40].

Expert panelists will be recruited through GIN and the International Shared Decision Making Society and invited by email to participate in the Delphi study. We will select the panelists on the basis of their expertise while ensuring representation of relevant perspectives (ie, selective sampling of CPG or SDM methodologists and health care professionals). The Delphi survey rounds will each consist of 20-30 participants, approximately half of whom will actively participate in the Delphi focus groups.

Based on the workshop conducted in advance, statements on the topic of preference sensitivity will be developed and put to vote. Before distributing the survey to the panelists, the developed statements will be pilot-tested by the Brandenburg Medical School Theodor Fontane and consortium partners to ensure clarity, usability, and functionality. As a first step in the Delphi study, a group of national SDM experts will evaluate and, if necessary, expand criteria that are considered appropriate for SDM. In the second step, international panelists will be asked to consider and vote on the inclusion of statements in web-based Delphi rounds in order to achieve a high level of agreement. Voting will take place on a scale from 1=no agreement to 8=full agreement. To specify the statements for the second Delphi round and the Delphi focus group discussion, a free-text option will be offered for each item. For a statement to be accepted, 75% (n=15-23) of the participants must agree with the statement, defined as a value of 5 or higher. Participants in the Delphi rounds will be sent the respective preliminary results and asked for feedback. As the number of Delphi rounds is difficult to estimate a priori, we will flexibly adapt the number of rounds needed to reach an agreement for statements put to vote. However, most Delphi studies reach agreement within 2 or 3 rounds, which is also assumed to be the case here. A new vote will be taken in each round.

In addition to the written feedback, the Delphi focus group will be held via videoconference. The purpose of the focus group is to reach an agreement on statements that were not agreed upon in the previous survey rounds. With an agenda, background information, consent form, and survey results provided in advance, participants will discuss the statements before a verbal summary and final vote. The results of this Delphi focus group will be recorded and incorporated into the methods to be developed.

The findings from the previous WPs will form the basis for the development of the tool to identify recommendations in CPGs that are most relevant to SDM. The methodology for developing the tool is based on the methods from the AGREE II (Appraisal of Guidelines for Research and Evaluation II) portfolio [41]. The process that we will follow, as described below, is geared on the development of the AGREE-REX (Appraisal of Guidelines for Research and Evaluation-Recommendations Excellence) [42,43]. A small working group (consisting of the project partners) will compile a preliminary list of initially relevant items, which will mainly be derived from the results of WPs 1-3. Subsequently, the list will be revised and reduced in terms of focus and objectives as well as potential overlaps between items. At the same time, a user manual will be produced, presenting the items with background, rationale, and explanation.

Items will reflect relevant aspects of preference sensitivity as identified in WPs 1, 2, and 3 as well as aspects of prioritization such as relevance of the clinical decision, burden of disease or treatment, and others. Items will be assessed using a Likert scale. An overall score may be calculated by summing up the individual scores. The use of weighing different items will be assessed by the development team. The higher the total score, the greater the need to provide SDM support for the respective recommendation. The possibility to generate a total score will facilitate the use and interpretation of the tool for subsequent application.

The number of PDAs developed or integrated into a CPG will depend highly on the resources available. The tool will reflect this by providing the possibility to prioritize those recommendations for which decision support seems most needed. Whether a cut-off value or a ranking according to total scores is more helpful will have to be considered.

The items on the list (ie, the preliminary tool), including the user manual, will be tested on 5 completed CPGs within the working group and then revised. We will ensure that the selected CPGs cover a broad spectrum of (1) conditions and (2) areas of application (diagnosis, prevention, and treatment).

The tool will be applicable to all recommendations of a CPG, regardless of the grade of recommendation. However, it may be discussed whether strong negative recommendations, probably recommendations for emergency settings, are worth rating, given that the addressed intervention is no sensible alternative (no benefit but relevant harm) or the urgency to act does not allow for SDM. The resulting version of the tool will be circulated among the project partners. They will be asked to provide feedback with regard to clarity, completeness, relevance, and user-friendliness. This includes both the items and the overall structure. The group will use the feedback received to revise the tool. This version will then be circulated to the project partners to test the tool against 2 CPGs and provide feedback, which will be followed by another revision and refinement loop. The project partners will finalize the items in a concluding workshop.

A key strategy to promote SDM in CPGs is to attach or link PDAs to SDM-relevant recommendations so that health care professionals can directly find the PDA and use it as support for the SDM process when their patient is facing a medical decision. In addition, there is currently no platform in Germany that (1) collects already existing PDAs, (2) makes them easy to find, (3) systematically compiles them, and (4) facilitates the creation of machine-readable PDAs through a user-friendly web-based editor. Taken together, we recognized the need to develop a platform for PDAs. With this platform, it will be possible to develop PDAs semiautomatically and to digitize as many of the existing formats of German-language PDAs as possible in a database in order to make them ubiquitously available via an interface for other applications (eg, hospital or physician information systems). The aim is to create a valuable possibility, in particular for those who create CPGs, to link existing PDAs and create new PDAs on key issues with little effort from the evidence syntheses that have been done in advance as part of the CPG work. WP 5 is of central importance for the project, as the development of the tool for CPGs alone misses the point of care practice. Without such a platform, the CPG tool would be able to identify and prioritize the recommendations for which SDM seems most relevant in the CPG development process, but the bridge to support the practical SDM process between health care professionals and patients would be missing. The platform also corresponds with the idea that all CPGs should be digitized so that PDAs can ultimately be better integrated or linked with CPGs and thus be more readily available at the point of care. An example of such a platform is PADA (formerly Collaborative Platform for Authors of Healthcare Decision Aids), which can be used to create new PDAs [44]. However, our target database aims to take another step forward by integrating already available PDAs, which may also differ in their formats, into the database.

To this end, a data structure is being developed that can be used to map the majority of existing PDA formats and make them digitally available. As a central function, the database will allow PDAs to be linked to each other and to external resources, for example, to the respective recommendation in the CPG and vice versa, when possible.

Three forms of presentation for PDAs are planned for the first version of the platform: (1) digital user-friendly option overviews, (2) multistage and partially digital user-friendly PDAs, and (3) option for information based exclusively on a sequence of texts, tables, and graphics.

To fill the database, known authors of PDAs will be contacted and invited to enter their PDAs in the German language. Additionally, a simultaneous search will be conducted in bibliographic databases such as PubMed and Embase, complemented by a Google search. A selection of PDAs identified in this way will be entered into the database when agreed by the PDA authors. The PDAs will be selected according to quality criteria defined during the project phase.

For the specific implementation of important functions, such as import, maintenance, linking of literature sources, data structures for study results or effect measures, and word export, previous knowledge of the consortium partners can be used. The platform should allow registered authors to independently enter PDAs into the system. For this purpose, a back end will be developed that includes a team-based user administration, an input system for PDAs in the 3 modes described earlier, as well as a release workflow for the PDAs and a versioning system for tracking changes. A literature management system for citing literature references in the PDA and a glossary function are also planned.

The front end with the interface for searching and retrieving PDAs should include a simple and comprehensive search with filter options using keywords and metadata (eg, subject area, year, and publisher). Users of the platform should be able to rate and review the PDAs (nonpublic). Possible improvements in presentation and completeness can be suggested by automated feedback mechanisms aimed at continuously evaluating and improving the quality of PDAs. The interface will be developed in a responsive design, making it possible to be displayed on mobile devices, tablets, and PCs. In addition, it should be possible to export the PDA as a PDF and other formats.

The system should provide an interface (application programming interface), through which third-party systems can access the system and retrieve PDAs. For example, integration into hospital information systems, decision support systems, or similar would be possible. The interface should have an authentication mechanism and allow access to the PDA and its metadata in a structured form.

To meet the needs of CPG developers and patients, they will be involved in the planning and subsequent test phase. To ensure the continuation of the platform after the end of the funding phase, it is intended to evaluate possibilities for further development and continued operation. A viable concept for this will be developed during the funding phase. In our view, possible institutions that could be interested in the continuation of such a platform include the EbM-Network, the German Network for Health Literacy, and the Association of the Scientific Medical Societies (AWMF), in addition to their digitization strategy.

For the integration of PDAs into the platform, inclusion criteria, in terms of minimum requirements, will be developed in advance and applied to the identified PDAs. The Good Practice Health Information of the EBM-Network [45] and the International Patient Decision Aid Standards criteria [46], among others, will be used to develop the quality criteria.

The last WP will serve to test the tool on approximately 12-15 German CPGs. Applying the tool in practice will allow to gain insights into its handling and application. At the same time, the actual integration of PDAs to CPG recommendations identified by the tool as sensitive to individual preferences will be realized.

For pragmatic reasons, the selected CPGs should be close to completion in order to test the developed tool. It is a prerequisite that the CPG group is currently working on the CPG, but at the same time, new or updated CPG recommendations have already been established so that individual recommendations or entire CPG chapters can already be used. Given the increasing number of living guidelines [47], this selection process seems realistic and has the advantage of not having to wait for the entire CPG to be finalized or updated. We will consider S3 CPGs (which represent the highest level of a systematic approach in the CPG development process according to the AWMF) registered in the AWMF registry or CPGs under revision, including the National Program for Disease Management Guidelines (NDMG) and CPGs of the German Guideline Program in Oncology (GGPO). For this purpose, an overview of these CPGs will be obtained from the AWMF registry and sorted by planned publication date (or alternatively expiry date of the CPG to be updated). In this way, the CPG organizations that are closest to finalizing the respective CPG will be contacted first. The relevant CPG organizations will be contacted consecutively but in such a way that at least 3 CPGs of the NDMG and GGPO will be included. Pragmatic aspects, such as the willingness to participate in the testing of the tool, will be particularly important in the selection process. The number of CPG recommendations included in the testing will vary from chapter to chapter, both within and across the CPGs. Approximately 30 recommendations per CPG should be included in the test. To achieve the planned number of recommendations for testing the tool, we may also decide to use several chapters of a CPG.

Similar to the development of quality indicators [48] or patient versions of CPGs [49] during the initial CPG development process, a small multidisciplinary team will be formed for each CPG, on which the tool will be tested. For this purpose, in addition to the person responsible for the respective CPG chapter (ie, CPG methodologist and health care professional), the involvement of patients and patient partners is planned. Patients or patient partners who are already involved in the CPG development will be contacted first. Otherwise, alternatives will be sought through patient organizations. The team will be completed by a person involved in the development of the tool (ie, one member of our project team) and an SDM expert. At the beginning of the testing phase, the latter 2 group members will present the SDM concept and the developed tool to the multidisciplinary team. Using the tool, the multidisciplinary team approaches a CPG that is being revised or newly developed and screens all CPG recommendations. This means that the tool will be applied to every single recommendation. Ratings will be given by health care professionals and patients. These ratings will be used to prioritize preference-sensitive recommendations. The SDM expert and a member of our project team will support the rest of the team in gathering continuous feedback with the intention of refining the tool during the application phase, if necessary. Each test group will provide (1) verbal feedback directly during the testing phase and (2) written feedback via a questionnaire after the testing phase.

In a final workshop, the findings and experiences will be reflected upon with representatives of the NDMG, GGPO, and AWMF. The aim of the workshop (or several workshops, if necessary) is to implement the tool in the respective manuals of the CPG organizations, taking into account their digitization strategies.