Background. Despite advancements in medical research and methodology, the credibility of the guideline development process is prone to compromise. Currently, no comprehensive tool exists that evaluates the trustworthiness of guidelines with a focus on single recommendations.
Objective. This project aims to develop and validate TRUSTGUIDE - a tool for assessing the trustworthiness of health guidelines, including credibility, appropriateness, and methodological quality.
Methods. To develop the TRUSTGUIDE Tool, we will follow a structured process and framework, including a multi-step process to generate and validate the new tool. The multistep development process includes: 1. Establishment of a Working Group; 2. Evidence synthesis; 3. TRUSTGUIDE development, including item generation, reduction, and response phrasing; 4. TRUSTGUIDE validation and AI implementation. The framework will also integrate AI in three main aspects: 1. To support the creation of tool items; 2. To include items related to using AI for each guideline development step; and 3. To support validation and application. The AI output will be validated by comparing it with human validation. Descriptive statistics for each item in the TRUSTGUIDE Tool will be calculated. Cronbach α and correlations, if item-changed will be calculated to assess the internal consistency. Intraclass correlations will be calculated to evaluate the TRUSTGUIDE reliability.
Results. The preliminary results will be presented during GIN Conference 2025.
Discussion. The TRUSTGUIDE tool will be focused and concise, evaluating the trustworthiness of guidelines. We expect good external validity, a strong correlation between TRUSTGUIDE and AGREE II, and adequate AI support for assessing Clinical Practice Guidelines.

Background: The development of practice guidelines requires that panelists approach the process with an open mind. Conflicts of interest (COI) of individuals, e.g., financial relationships, can lead to fixed views. The same applies to the attachment to a specific viewpoint that is not tied to financial interests (i.e., intellectual, or nonfinancial interests).
Objective: We would like to make the case for accounting for open-mindedness as part of the COI management policies, in addition to the declaration of financial interests, and in lieu of the declaration of non-financial interests.
Methods: Concept paper.
Results: Traditionally, we have relied on reported “interests” as surrogate measures of the “extent of fixed-mindedness”. However, the surrogate measure (interests) might not be fully correlated with the target measure (extent of fixed-mindedness). Accordingly, individuals judged to have no COI or COI that does not preclude their participation should be assessed for their open-mindedness. While valid methods for assessing open-mindedness have not been developed, one could ask prospective panel members about their (1) preferences for the direction and strength of the recommendations being considered; and (2) willingness to change their preferences if the evidence and panel discussions favor such changes.
Discussion: Accounting for open-mindedness as part of COI management policies can complement the declaration of financial COI, and negate the need for the non-financial COI category and the challenges of operationalizing it. There is a need to develop and evaluate tools and processes to assess and account for open-mindedness.

Background
In Germany guideline group members (GGMs) who have moderate conflicts of interest (COI) in specific subject areas should abstain from voting on recommendations in those areas. In the German Guideline Program in Oncology (GGPO), this regulation is implemented through an electronic voting system where GGMs with and without moderate COI vote separately and anonymously.

Objective:
Our aim is to investigate the frequency of moderate COI during voting at consensus conferences and the influence of these conflicts on voting outcomes regarding approval rates.

Methods:
The protocols from consensus conferences conducted over the last 4 years were analyzed. The total number of votes as well as the proportion of votes by GGMs with moderate COI were determined. For votes with involvement of GGMs with moderate COI, the potential impact on voting outcomes was assessed.

Results:
We analyzed 1,328 votes across 70 conferences held between 2021 and 2024 for 21 guidelines. In 338 (25.5%) of these votes, GGMs with moderate COI were involved. In 40 cases (11.8%), the non-abstention of GGMs with COI would have significantly altered the voting outcome. In 22 (6.5%) cases, a rejected recommendation would have been accepted, and in 18 (5.3%) cases, an accepted recommendation would have been rejected. In 118 cases (34.9%), the consensus strength of the vote would have changed while maintaining consensus.

Discussion:
The involvement of individuals with moderate COI may influence voting outcomes in guideline conferences. Abstention rules thus help reduce the impact of COI on guideline recommendations.

Background. Systematic reviews (SR), which form the foundation of clinical guidelines, are widely regarded as the most reliable resources for informed decision-making in healthcare. However, they often focus on research published in peer-reviewed journals, risking overlooking study results not published there.

Objectives. To evaluate the impact of including data not published in journals (e.g. from trial registries, only available upon request from authors) on the results of network meta-analyses in efficacy and harm outcomes.

Methods. This ongoing case-study is based on a SR (CRD42023441146) evaluating the pharmacologic treatment of acute episodic migraine attacks in adult outpatients. To expand the analysis, we supplemented the original literature search with additional searches targeting unpublished data (i.e., trial registries, unpublished drug approval data, preprints, conference abstracts). We screened the additional records and extracted relevant data dually.
We will re-run the network meta-analysis (NMA) for all main outcomes of the SR incorporating both published and unpublished data and compare results with the original NMA including only published data.
The primary outcomes of interest are changes in the certainty of evidence and overall conclusions. We will also assess the equivalence of effect estimates from network meta-analyses with and without unpublished data, along with changes in network geometry and ranking probabilities.

Preliminary Results. In addition to the 169 publications (174 randomised controlled trials) included in the original SR, we identified 50 relevant unpublished studies, 12 with posted results. Currently we are preparing the main analysis for this study. We expect to complete the study in April 2025.

Background: Internationally accepted standards for trustworthy guidelines include the necessity to ground recommendations in values and preferences. Early anecdotal reports suggest that COVID-19 public health guidelines did not consider values and preferences.
Objective: To systematically evaluate the extent to which COVID-19 public health guidelines considered values and preferences in their development and to assess the methods used for their inclusion.
Methods: We performed a systematic review of COVID-19 public health guidelines. We searched the eCOVID-19 RecMap platform up to July 2023 and the Guidelines International Network (GIN) Library up to May 2024. We included guidelines that made recommendations addressing vaccination, masking, isolation, lockdowns, travel restrictions, contact tracing, infection surveillance, and school closures. Reviewers worked independently and in duplicate to review guidelines for consideration of values and preferences.
Results: Our search yielded 130 eligible guidelines, twenty-eight (21.5%) guidelines considered values and preferences. Among guidelines that considered values and preferences, most did so to assess the acceptability of recommendations (23; 82.1%) and by referencing published research (25; 89.3%). Guidelines only occasionally engaged laypersons as part of the guideline development group (8; 28.6%).
Discussion: Most COVID-19 public health guidelines did not consider values and preferences. When they were considered, it was often suboptimal. Disregard for values and preferences in guidelines may have contributed to unpopular COVID-19 policies and an erosion of trust in public health. Given the possibility of future health emergencies, we recommend guideline developers identify efficient and effective methods for considering values and preferences in crisis situations.

Background
In 2021, the Japan Stroke Society revised its clinical practice guidelines (CPGs) to strongly recommend mechanical thrombectomy (MT) for acute ischemic stroke.
Objective
This study aimed to assess the impact of Japanese stroke CPGs on the adoption of MT and its implications for inpatient outcomes.
Methods
We analyzed national administrative claim data managed by the Japanese government in 2017 and 2022. The analysis focused on hospitals performing endovascular treatment (ET), including MT, and its impact on length of stay (LOS). Statistical significance was determined using chi-square and t-tests.
Results
The number of hospitals and ischemic stroke inpatients was 3,701 hospitals with 207,841 inpatients in 2017, and 5,919 hospitals with 206,382 in 2022. The proportion of hospitals performing ET rose from 18.5% to 28.8% (p < 0.001), and the average number of ETs per hospital increased from 24.9 to 33.0 (p < 0.001). In 2022, hospitals experienced in ET since 2017 had a shorter mean LOS than those adopting it later (27.6 vs. 29.3 days, p < 0.001) and showed lower variability (coefficient of variation: 0.25 vs. 0.28).
Discussion for scientific abstracts
The Japanese government enacted the Cerebrovascular and Cardiovascular Disease Control Act to enhance stroke and cardiovascular care in 2019. The Japan Stroke Society has certified MT facilities and specialists to improve the medical system for stroke. With government and academic efforts as background, CPGs have influenced the uptake of MT. The shorter, more consistent LOS at experienced hospitals suggests that greater proficiency in MT improves patient outcomes.

Background
The GRADE approach, extensively outlined in numerous publications, has been extremely influential. Over time, however, GRADE guidance has become increasingly complex, and its evolution difficult to follow. Core GRADE initiative has introduced a new series of guidance articles (we anticipate BMJ to publish the series by the GIN meeting), which offers a solution to these problems: a single, coherent presentation of the essential GRADE principles. This workshop will familiarize participants with Core GRADE.

Objective
The primary objective is to provide participants with an understanding of Core GRADE principles and equip them with practical skills to apply these concepts to their systematic reviews and practice guidelines.

Format for workshops
The workshop will include three key segments:
Introduction (30 minutes): We will introduce the Core GRADE framework, summarize key messages from the Core GRADE series, and demonstrate practical applications with examples.

Group Discussions (45 minutes): Participants will work in small groups (4-10 people for each group, depending on the number of registrants), each facilitated by one of the workshop speakers. Groups will engage in discussions on one of the following topics:

Topic 1: Core GRADE in risk of bias assessment.
Topic 2: Core GRADE in addressing imprecision and inconsistency.
Topic 3: Core GRADE’s approach to subgroup analyses applying the ICEMAN instrument to the credibility of subgroup effects.
Topic 4: Core GRADE in addressing moving from evidence to decisions.

Feedback and Summary (15 minutes): Each group will present their findings and challenges, followed by expert feedback and a concluding summary.