BMJ Evidence-Based Medicine最新文献

{"title":"Effectiveness, safety, adherence and risk compensation of oral pre-exposure prophylaxis (PrEP) in men who have sex with men: a systematic review and meta-analysis.","authors":"Chengchang Xiao, Kecheng Qu, Qian He, Yuanzhong Zhu, Qingya Yu, Mengxi Zhai, Nuo Li, Rong Hu, Songjie Wu, Bin Yu","doi":"10.1136/bmjebm-2025-113648","DOIUrl":"https://doi.org/10.1136/bmjebm-2025-113648","url":null,"abstract":"<p><strong>Objective: </strong>This systematic review and meta-analysis evaluate the clinical effectiveness, safety and potential risk compensation associated with oral pre-exposure prophylaxis (PrEP) in men who have sex with men (MSM).</p><p><strong>Methods: </strong>We systematically searched PubMed, Embase and the Cochrane Library up to 17 September 2025 for randomised controlled trials (RCTs) and observational studies. The primary outcomes were HIV incidence and sexually transmitted infections (STIs). Secondary outcomes included PrEP adherence and adverse events. Meta-analyses used random-effects models, presenting results as relative risks (RRs), risk differences and calculated absolute risk reduction/increase (ARR/ARI) and number needed to treat (NNT) where applicable.</p><p><strong>Results: </strong>A total of 21 studies (7 RCTs and 14 observational) with 18 214 participants were included. PrEP use was associated with a 68% RR reduction in HIV acquisition (RR=0.32, 95% CI 0.19 to 0.53; ARR=1.14%, NNT=88). High adherence (≥80%) led to a 78% reduction in HIV risk (RR=0.22, 95% CI 0.09 to 0.54; ARR=1.51%, NNT=66), while lower adherence showed a non-significant 58% reduction (RR=0.42, 95% CI 0.16 to 1.09). Meta-regression indicated a 17% decrease in efficacy per 10% drop in adherence (β = -1.7, p=0.009). No significant differences were found in any adverse events (RR=1.02, 95% CI 0.88 to 1.18) or serious adverse events (RR=0.93, 95% CI 0.68 to 1.28; ARI=-0.00%). There was no significant increase in overall STIs (RR=1.05, 95% CI 0.95 to 1.17; ARI=0.05%) or syphilis incidence (RR=1.11, 95% CI 0.96 to 1.30; ARI=0.11%). Sexual behaviour outcomes were mixed, with no consistent evidence of increased risk-taking.</p><p><strong>Conclusion: </strong>PrEP effectively reduces HIV incidence among MSM, particularly when adherence is high, without a significant increase in overall STIs risk. Adherence support and comprehensive sexual health counselling should be integral to PrEP programmes to maximise its effectiveness and mitigate potential risk compensation behaviours.</p><p><strong>Prospero registration number: </strong>CRD42024593380.</p>","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":""},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149103949","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Assessing shared decision-making in clinical practice: a systematic review and meta-analysis of studies using OPTION-12 and OPTION-5.","authors":"Cato Caroline Bresser, Andrea Duarte-Díaz, Himar González-Pacheco, Amado Rivero-Santana, Yolanda Ramallo-Fariña, Jet Westerink, Lea Dijksman, Harm van Melick, Paul van der Nat, France Légaré, Glyn Elwyn, Mirjam Garvelink, Lilisbeth Perestelo-Perez","doi":"10.1136/bmjebm-2025-113830","DOIUrl":"10.1136/bmjebm-2025-113830","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Objectives: &lt;/strong&gt;Observing Patient Involvement in Decision Making (OPTION)-12 and OPTION-5 assess the extent to which observers score healthcare professionals' (HCPs) involvement of patients in shared decision-making (SDM). We systematically reviewed studies measuring the extent to which HCPs involve patients in the decision-making process using the OPTION instrument.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Design: &lt;/strong&gt;Informed by Preferred Reporting Items for Systematic Reviews and Meta-Analyses, we updated a previous systematic review and included new studies reporting OPTION-12 or OPTION-5 scores from recordings of real-world clinical encounters, involving patients and HCPs making healthcare-related decisions. Searches were conducted across PubMed, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), and Web of Science databases (2012-2025), supplemented by citation screening and outreach to professional networks. We extracted study characteristics, OPTION version, psychometric data and item-level score details. We also assessed the study quality using the reports of rating procedures and conducted meta-analyses, subgroup analyses using a priori hypotheses and completed meta-regressions.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;In total, 174 studies were included, comprising almost 20 000 clinical consultations: 102 studies used only OPTION-12 and 64 used only OPTION-5, while four studies reported using both scales. Mean OPTION-12 and OPTION-5 score for studies unaffected by interventions were 25.1 (95% CI 22.1 to 28.2, k=76, I&lt;sup&gt;2&lt;/sup&gt;=99.71%) and 31.8 (95% CI 26.6 to 37.1, k=42, I&lt;sup&gt;2&lt;/sup&gt;=99.55%), respectively. Subgroup analyses revealed significantly higher scores in studies with postintervention OPTION-scores for both OPTION-12 (38.4 vs 25.1, p&lt;0.001, k=91, I&lt;sup&gt;2&lt;/sup&gt;=99.55%) and OPTION-5 (47.7 vs 31.8, p&lt;0.001, k=65, I&lt;sup&gt;2&lt;/sup&gt;=99.39%). In univariable meta-regression, longer consultation duration and female patient percentage (only for OPTION-12) were associated with higher scores. However, multivariable meta-regression revealed that clinical setting was the sole independent predictor for OPTION-12 (p=0.007), whereas consultation duration remained the primary independent predictor for OPTION-5 (p=0.003).&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;Since the 2015 previous review, little overall improvement has been observed. This limited progress raises important questions about how we interpret changes in observed SDM. Specifically, it remains unclear what degree of change in OPTION-12 scores reflects a meaningful improvement. Our multivariable findings provide a more nuanced perspective: while consultation duration remains the primary independent predictor for patient involvement when measured with OPTION-5, clinical setting emerges as a more critical independent driver for OPTION-12. These results suggest that the influence of time is not uniform across assessment tools and that structural barriers in different clinical environments ","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"300-314"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147762016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reconsidering the role of tramadol in chronic pain management.","authors":"Antonio Alcántara Montero","doi":"10.1136/bmjebm-2025-114276","DOIUrl":"10.1136/bmjebm-2025-114276","url":null,"abstract":"","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"355-356"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145399836","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mapping drug-class associations with sudden cardiac arrest in the outpatient setting: an exploratory population-based case-control analysis.","authors":"Safoura Sheikh Rezaei, Martina Mittlböck, Michael Schwameis, Thomas Bergmair, Michael Holzer, Mario Krammel, Magdalena Sophie Boegl, Anne Elisabeth Merrelaar, Nina Buchtele, Alexandra Julia Lipa, Daniel Kogler, Harald Herkner","doi":"10.1136/bmjebm-2025-113993","DOIUrl":"10.1136/bmjebm-2025-113993","url":null,"abstract":"<p><strong>Objectives: </strong>To map the risk profile of drug classes at the population level by identifying associations between newly prescribed reimbursed drugs and sudden cardiac arrest (SCA), thereby providing a systematic basis for assessing individual drug effects.</p><p><strong>Design: </strong>Population-based, matched case-control study.</p><p><strong>Setting: </strong>Greater Vienna area, Austria, using linked data from the Austrian Healthcare Reimbursement Database and the Vienna Ambulance Service between 2013 and 2020.</p><p><strong>Participants: </strong>A total of 31 330 insured individuals were included from an estimated 14 448 612 person-years at risk. Cases (n=6266) were patients with SCA, each matched 1:4 to controls (n=25 064) without SCA by age, sex and event date. The estimated absolute case risk equals 43.4 per 100 000 person-years.</p><p><strong>Interventions: </strong>Newly prescribed reimbursed drugs within 4 weeks before the SCA event, classified according to four-digit Anatomical Therapeutic Chemical (ATC) codes.</p><p><strong>Main analyses: </strong>Associations between newly prescribed drug classes and SCA, estimated using conditional logistic regression and expressed as ORs with 95% CIs, adjusted for comorbidities and concomitant medications prescribed 120-29 days before the event.</p><p><strong>Results: </strong>Among 322 relevant ATC drug classes, 245 were newly prescribed within 28 days prior to the SCA event. Of these, 57 (23%) were significantly associated with SCA. Eight drug classes demonstrated a markedly elevated risk (OR≥10), including oripavine derivatives (OR 64, 95% CI 8 to 486), other cardiac preparations (OR 22, 95% CI 2 to 204), plain antiandrogens (OR 18, 95% CI 1.2 to 275) and phenylpiperidine derivatives (OR 16, 95% CI 7 to 38). The most frequently prescribed associated drug classes were penicillins with beta-lactamase inhibitors (179 cases; OR 2.17, 95% CI 1.8 to 2.7), pyrazolones (167 cases; OR 2.14, 95% CI 1.7 to 2.7) and adrenergics combined with anticholinergics (130 cases; OR 2.94, 95% CI 2.2 to 3.9).</p><p><strong>Conclusions: </strong>Numerous outpatient-prescribed drug classes were associated with SCA. This exploratory, population-based analysis provides a systematic map of potential safety signals to inform more detailed pharmaco-epidemiological investigations, in which confounding and causality can be examined more rigorously.</p>","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"315-323"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146008885","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating data extraction error by a large language model from randomised controlled trials: a large-scale empirical study.","authors":"Shiqi Fan, Ming Chen, Suhail A Doi, Zhangnan Ye, Zhen Peng, Yuan Tian, Chengzhi Zhang, Luis Furuya-Kanamori, Lifeng Lin, Evan Mayo-Wilson, Mohammad Hassan Murad, Xiuhong Meng, Chang Xu","doi":"10.1136/bmjebm-2025-114044","DOIUrl":"10.1136/bmjebm-2025-114044","url":null,"abstract":"<p><strong>Objective: </strong>To examine the potential errors of a general large language model (LLM) (ie, Claude 3.5 Sonnet) on data extraction from randomised controlled trials (RCTs).</p><p><strong>Design and setting: </strong>An empirical study comparing Claude 3.5 Sonnet extractions against a human-performed verification dataset. The extraction tasks for Claude 3.5 Sonnet were based solely on original RCT portable document format (PDF) files. For PDFs that could not be directly extracted by Claude 3.5 Sonnet, optical character recognition was employed to convert them into text format before extraction.</p><p><strong>Participants: </strong>A random sample of 664 trials was selected from a well-established trial bank and a final data pool was established based on rigorous manual cross-checking as a reference standard.</p><p><strong>Data sources: </strong>PubMed, EMBASE, Scopus, Web of Science (all databases) and the Cochrane Central Register of Controlled Trials (CENTRAL) up to February 2023.</p><p><strong>Eligibility criteria for selecting studies: </strong>RCTs on children involving medication and adverse events.</p><p><strong>Main outcome measures: </strong>Claude 3.5 Sonnet was applied to extract the basic information (eg, trial design, population information and source of funding) and adverse outcomes (ie, name of adverse events, number of events). Claude 3.5 Sonnet outputs were compared against the final data pool and all errors were recorded. Results are presented as error rates and with 95% CI, estimated using a generalised linear mixed model.</p><p><strong>Results: </strong>For the 664 trials, a total of 23 069 data cells were extracted via Claude 3.5 Sonnet, with 10 624 for basic information and 12 445 for adverse outcomes. The overall error rate for data extraction was 6.6% (95% CI 5.4% to 8.2%), with 5.7% (95% CI 5.2% to 6.1%) in basic information and 7.6% (95% CI 4.9% to 11.8%) in adverse outcomes. When stratified the 1542 total errors by error types, misallocation (assigning data to incorrect fields; 57.1%, 881/1542) and missed or omitted data (incomplete extraction of available data; 23.2%, 357/1542) accounted for the two most frequent errors, with misallocation occurring more in basic information (53.3%, 470/881), while missed or omitted data occurred more in adverse outcomes (96.1%, 343/357). Post hoc analysis examining the association between trial reporting quality (assessed using Consolidated Standards of Reporting Trials (CONSORT) 2025 and LLM data extraction error rates indicated that higher CONSORT adherence was associated with lower extraction error rates.</p><p><strong>Conclusions: </strong>The data extraction error of Claude was relatively low, but it alerts LLM applications in evidence synthesis. Detailed checking for LLM outputs should be the primary consideration for evidence synthesisers.</p>","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"324-331"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147761982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correspondence on 'Effectiveness and clinical relevance of kinesio taping in musculoskeletal disorders: an overview of systematic reviews and evidence mapping' by Mo <i>et al</i>.","authors":"Jean-Pascal Grenier","doi":"10.1136/bmjebm-2026-114689","DOIUrl":"10.1136/bmjebm-2026-114689","url":null,"abstract":"","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"353-354"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147761985","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reporting GUideline for Intervention DEscription in Rehabilitation (GUIDE-Rehab): a tool to open the 'black box' of rehabilitation complex interventions.","authors":"Stefano Negrini, Chiara Arienti, Susan Armijo-Olivo, Pierre Côté, Allen Walter Heinemann, Carlotte Kiekens, Dinesh Kumbhare, William Levack, Thorsten Meyer-Feil, John Whyte","doi":"10.1136/bmjebm-2025-113997","DOIUrl":"10.1136/bmjebm-2025-113997","url":null,"abstract":"<p><p>In 2023, the World Health Assembly adopted a resolution to strengthen rehabilitation within health systems, calling for rehabilitation research. Within health, the term rehabilitation has multiple meanings, including a core strategy, a sector, a service and an intervention. The latter has been defined as complex and characterised as a 'black box', similar to complex interventions in other fields. The existing reporting guidelines are not sufficiently effective in describing interventions within the rehabilitation field. We developed the GUideline for Intervention DEscription in Rehabilitation (GUIDE-Rehab) to address these challenges.According to the Enhancing the QUAlity and Transparency Of health Research Network, we followed a Delphi process with multiple Consensus Meetings and piloting and used ACcurate COnsensus Reporting Document for reporting. The background research involved 21 papers. We based GUIDE-Rehab on the Rehabilitation Treatment Specification System, developed over 15 years of research to improve rehabilitation description; the definition of rehabilitation for research purposes; and the Template for Intervention Description and Replication reporting guideline. 68 representatives from global rehabilitation stakeholders (scientific societies, journals, evidence and methods groups), including individuals with lived experience of disability, from 26 countries across all continents and economies, participated. The piloting involved 17 chief editors, 7 research groups and participants from 10 scientific meetings.The complete version comprises 16 items, while the version for uncontrolled studies includes 13. The short version (10 items for text, 6 for appendix) helps reduce the manuscripts' length. The GUIDE-Rehab graphical illustration (nine items) facilitates the intervention description. GUIDE-Rehab will assist in the reporting of interventions in rehabilitation to enhance clinical research and support clinical implementation.</p>","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"342-349"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145751735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accurate is not enough: select formats for communicating probabilities to achieve specific outcomes.","authors":"Brian J Zikmund-Fisher, Jessica S Ancker","doi":"10.1136/bmjebm-2025-113738","DOIUrl":"10.1136/bmjebm-2025-113738","url":null,"abstract":"","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"287-289"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12320222/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144538411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impact of prompt engineering on large language models for risk of bias assessment: a comparative study.","authors":"Yu-Tao Xiong, Yuan-Chen Wang, Ya-Nan Sun, Zhengzhe Zhan, Wei Liu, Hao-Nan Liu, Ruo-Yu Meng, Wei Tang, Chang Liu","doi":"10.1136/bmjebm-2025-114055","DOIUrl":"10.1136/bmjebm-2025-114055","url":null,"abstract":"<p><strong>Objectives: </strong>To evaluate the performance of large language models (LLMs) in risk of bias assessment and to examine whether prompt engineering improves their accuracy and alignment with expert reasoning.</p><p><strong>Methods: </strong>We analysed 158 randomised controlled trials from 10 dental systematic reviews and their risk of bias assessments were reviewed and revised to serve as the reference standard. Two LLMs (DeepSeek-V3 and GPT-5) were evaluated under four prompting strategies, including direct command, command with reference, constrained output and formula-constrained output. The direct command served as the blank control group, simulating the approach commonly used by clinicians, whereas the other three groups employed different prompt engineering. The performance of LLMs across the seven domains of RoB-1 was evaluated using accuracy and agreement. The reasoning process of the LLMs was expressed in the form of syllogisms and its similarity to expert reasoning was assessed using MMD².</p><p><strong>Results: </strong>LLMs showed limited capability in risk of bias assessment under the blank control condition, with mean accuracies of 0.72 for DeepSeek-V3 and 0.65 for GPT-5. With formula-constrained prompting, the performance of both LLMs improved significantly, and the overall accuracy increased to 0.85 for both DeepSeek-V3 and GPT-5 (both vs the blank control group, p<0.001). Agreement metrics showed a similar pattern, with higher agreement under formula-constrained prompting than under the other prompting strategies (p<0.001 for both models). In addition, the syllogistic output format provided a clear representation of the reasoning process underlying risk of bias assessment. Compared with constrained output, formula-constrained prompting also produced reasoning that was more closely aligned with the reference answers, as indicated by lower MMD² values (DeepSeek-V3: 0.0765 vs 0.1239; GPT-5: 0.0548 vs 0.1068).</p><p><strong>Conclusion: </strong>Prompt engineering substantially improved the performance of LLMs in risk of bias assessment. Although LLMs cannot currently replace human reviewers, they may serve as efficient and transparent tools to support this process.</p>","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"332-341"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147962206","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring the application of GRADE in formulating WHO public health guidelines: a scoping evidence review.","authors":"Xiuxia Li, Tao Nian, Wendi Liu, Xue Shang, Xinxin Deng, Kangle Guo, Nan Chen, Yan Wang, Yaolong Chen, Kehu Yang","doi":"10.1136/bmjebm-2024-113487","DOIUrl":"10.1136/bmjebm-2024-113487","url":null,"abstract":"<p><strong>Objectives: </strong>To systematically review the utilisation of the Grading of Recommendations Assessment, Development and Evaluations (GRADE) framework within the context of WHO public health guidelines (PHGs), identifying key features, areas of concentration and potential deficiencies.</p><p><strong>Design and setting: </strong>From 2007 to February 2024, a comprehensive search of the WHO website was conducted to identify PHGs that have incorporated the GRADE methodology.</p><p><strong>Participants: </strong>The study focused on the PHGs identified through the above search.</p><p><strong>Interventions: </strong>Data extraction and analysis were independently conducted by researchers using Microsoft Excel 2019.</p><p><strong>Main outcome measures: </strong>For each PHG, key recorded characteristics included publication details, thematic areas, the strength of recommendations, the certainty of evidence, and characteristics of the GRADE downgrading/upgrading domains.</p><p><strong>Results: </strong>Out of 228 PHGs examined, 9234 (90.55%) outcome indicators used the GRADE rating system, predominantly in the area of sexual and reproductive health (50%). Only 31.62% of the outcomes reported moderate and high certainty in evidence. The main clustering results were dominated by adverse events. Among the 4013 recommendations, 2067 (51.51%) were strong, while 1477 (36.81%) were weak/conditional. It is noteworthy that 46.83% of the strong recommendations were based on low or very low confidence in evidence. Among these strong recommendations, 119 met the criteria of five paradigmatic situations where it was necessary to issue a strong recommendation despite low or very low confidence in the effect estimates. Among the 13 230 instances of downgrading, 41.09% were due to the risk of bias and 35.90% were due to imprecision. Only 0.25% outcomes were upgraded by the magnitude of effect size and 0.03% by dose-response gradient.</p><p><strong>Conclusion: </strong>The GRADE approach is widely used in the development of PHGs. More than half of the recommendations in PHGs are based on low or very low-quality evidence, primarily due to risks of bias and imprecision. Additionally, strong recommendations based on low-confidence or very low-confidence estimates are frequently made. Therefore, it is necessary to enhance guideline developers' understanding of the GRADE methodology and to further investigate and clarify discordant recommendations in PHGs.</p>","PeriodicalId":9059,"journal":{"name":"BMJ Evidence-Based Medicine","volume":" ","pages":"290-299"},"PeriodicalIF":10.8,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146008851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}