Journal of Clinical Epidemiology最新文献

{"title":"Journal of clinical epidemiology The Importance of Reporting CT Scan Intervals in Real-World Oncology Studies: A Simulation Analysis of Afatinib in Advanced-Stage Non-Small Cell Lung Cancer.","authors":"Sasiwimon Iwsakul, Sarayut Lucien Geater","doi":"10.1016/j.jclinepi.2025.111841","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111841","url":null,"abstract":"<p><strong>Objectives: </strong>In real-world oncology studies, progression-free survival (PFS) is often used as a surrogate endpoint. However, the interval of the CT scans, which might impact PFS, is usually not mentioned in publications based on real-world data. This study aims to determine the relationship between CT scan interval and PFS captured across different assessment frequencies (captured PFS or cPFS). The secondary objective is to identify appropriate correction factors for median PFS (mPFS) in real-world data, necessary for comparisons with clinical trials.</p><p><strong>Study design and setting: </strong>A simulation was conducted to create theoretical PFS data. Time-to-event analyses, including Kaplan-Meier survival curves, Cox regression models, parametric survival models, and multilevel survival regression models, were used to assess the effect of various CT scan intervals on cPFS durations. The Laplace regression model was employed to compute fiftieth-percentile ratios.</p><p><strong>Results: </strong>The mPFS (95% CI) captured by cPFS6, cPFS12, cPFS16, and cPFS20 were 11.56 (10.78, 12.71), 12.42 (11.17, 13.80), 12.98 (11.70, 13.96), and 13.80 (12.71, 14.49) months, respectively. There was a significant difference in mPFS durations captured at different CT scan intervals. The correction factors for converting mPFS from cPFS12, cPFS16, and cPFS20 to cPFS6 were 1.06, 1.09, and 1.14, respectively.</p><p><strong>Conclusions: </strong>Longer CT intervals resulted in longer mPFS, leading to a systematic overestimation of about 1-2 months.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111841"},"PeriodicalIF":7.3,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144152849","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Network meta-regression including baseline risk analysis and interactive visualisations as implemented by the MetaInsight web application.","authors":"Tom Morris, Janion Nevill, Clareece Nevill, Naomi Bradbury, Suzanne Freeman, Nicola Cooper, Alex Sutton","doi":"10.1016/j.jclinepi.2025.111839","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111839","url":null,"abstract":"<p><strong>Background and objectives: </strong>The MetaInsight web application (https://apps.crsu.org.uk/MetaInsight) allows users to carry out network meta-analyses (NMA) via a point-and-click interface, without the need for statistical programming. Network meta-regression (NMR) is an extension of NMA that adds covariates to the model and is used to investigate heterogeneity and inconsistency within the network. Specifically, NMR allows users to explore the interaction between treatment and study-level covariates. The aim of this paper is to describe the implementation and application of NMR in MetaInsight, for a single covariate, which may be a study-level variable or baseline risk with the uncertainty in the latter correctly accounted for.</p><p><strong>Methods: </strong>NMR has been added to MetaInsight using the R packages gemtc and bnma. The type of regression coefficients fitted can be set to shared, exchangeable or unrelated relating to whether the same or different relationships are assumed between the covariate and each treatment. A graph has been added to show the distribution of covariate values, and a novel visualisation has been developed to show which studies contribute to which comparisons, for multiple comparisons simultaneously.</p><p><strong>Results: </strong>The new functionality is described and illustrated with an example, and screenshots of the app are included.</p><p><strong>Conclusion: </strong>This extensive update of the app greatly facilitates such complex analyses making them freely accessible to researchers from a wide range of backgrounds. This in turn should improve the reporting and reliability of published NMA which ultimately should positively impact clinical decision making.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111839"},"PeriodicalIF":7.3,"publicationDate":"2025-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144144265","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Patient Preference Trials in Oncology: A Scoping Review and Recommendations on Reporting.","authors":"D Kamp, R Kessels, N Haj Mohammad, R S J M Schmitz, J IntHout, H W M van Laarhoven, A M May, R H H Groenwold","doi":"10.1016/j.jclinepi.2025.111837","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111837","url":null,"abstract":"<p><strong>Objective: </strong>The feasibility, generalisability and validity of randomised controlled trials (RCTs) may be compromised when patients have preferences for specific trial arms. The patient preference trial (PPT) design, which assigns at least a subset of the patients based on their preference rather than randomisation, has been proposed as an alternative. The aim of this review was to provide an overview of the application of the PPT design in oncological research, with a particular focus on PPTs designed to assess the effectiveness of an intervention, and provide recommendations on reporting.</p><p><strong>Study design and setting: </strong>We performed a scoping review including all prospective oncological studies where trial arm allocation for part or all of the patients was based on patients' preferences. Retrieved information included research objectives, motivation for the design choice, baseline covariate adjustments, and sample size calculations.</p><p><strong>Results: </strong>We identified 44 PPTs, of which 34 (77%) aimed to investigate intervention effectiveness and were reviewed in more detail. Of these 34 effectiveness PPTs, 24 (71%) studies were completed trials, while 10 were protocols. The most frequently indicated rationale behind opting for the PPT design was the perceived infeasibility of an RCT (17 (50%) studies). Baseline covariate adjustment was performed in 18 (53%) studies. Fourteen out of 24 completed trials reported an unequal allocation ratio across trial arms not anticipated during the design.</p><p><strong>Conclusion: </strong>This review identified several challenges for PPTs in oncology. Future PPTs should take bias due to confounding and unequal group sizes arising from the preferential allocation into account and report on these issues, as these aspects are crucial for the validity and statistical power of the study.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111837"},"PeriodicalIF":7.3,"publicationDate":"2025-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144144268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Systematic reviews of observational studies frequently conclude based on meta-analyses of biased results: Standards must be improved.","authors":"Mical Paul, Judith Olchowski, Leonard Leibovici","doi":"10.1016/j.jclinepi.2025.111840","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111840","url":null,"abstract":"<p><strong>Background: </strong>Systematic reviews of observational studies are common, yet the methodology of these systematic reviews and meta-analyses based on observational data are not standardized.</p><p><strong>Methods: </strong>We reviewed systematic reviews including observational studies published in JAMA, Lancet, BMJ or Annals of Internal Medicine journals in 2024. We extracted information on the methodology and reporting of these reviews. We propose minimal standards for conduct and reporting of systematic reviews including observational studies.</p><p><strong>Results: </strong>We included 63 systematic reviews that compiled observational studies. Nearly all (51/63, 80.9%) performed meta-analyses of crude, unadjusted, results from observational studies. Only 22/63 (34.9%) addressed anywhere in the article or supplement adjusted association estimates from observational studies. Adjusted and unadjusted meta-analyses were presented in 9/63 (14.3%) reviews. The study designs included in the review were not presented in the abstract or in the methods section in about a third of the reviews. These systematic reviews frequently concluded strongly on effects that cannot be assessed in observational data. We analysed the methodology used in the systematic reviews addressing adjusted results from observational studies and propose standards for analysis and reporting of systematic reviews of observational studies.</p><p><strong>Conclusions: </strong>Use of crude data is common in systematic reviews including observational studies and ignores bias by indication, immortal time bias and other biases. These reviews are misleading and better standards are needed.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111840"},"PeriodicalIF":7.3,"publicationDate":"2025-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144144271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"At the CORE of measurement - What do you want to measure? And how do you want to measure it? A COSMIN Perspective.","authors":"L B Mokkink, S Herbelet, C B Terwee, M Boers","doi":"10.1016/j.jclinepi.2025.111836","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111836","url":null,"abstract":"<p><p>Any measurement starts with determining what to measure (i.e. which outcome) and how to measure it (i.e. which outcome measurement instrument). This article is the first from a series of COSMIN Perspective in which we will explain measurements and how to evaluate measurement quality. In this first Perspective we explain the rationale for defining an outcome and operationalizing it into an outcome measurement instrument. An outcome should first be defined based on theory, and next be operationalized into one or more observable items, questions, tasks or parameters. But a measurement instrument is more than its items or machine. Any type of measurement instrument consist of five components: (1) equipment, (2) preparatory actions, (3) unprocessed data collection or collection of the biological sample(s), (4) processing and storage, and (5) assignment of the score or determination of the value of the biological sample. Thinking in terms of the components of measurement instruments can enhance the design and understanding of studies on measurement properties.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111836"},"PeriodicalIF":7.3,"publicationDate":"2025-05-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144129584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Maternal early warning scores shown to be methodologically weak and at high risk of bias.","authors":"Mae Chester Jones, Shona Kirtley, Fema Er, Vidoushee Jogarah, Yu Qiao, Ruth Tunn, Naomi Vides, Peter J Watkinson, Marian Knight, Stephen Gerry, Gary S Collins","doi":"10.1016/j.jclinepi.2025.111833","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111833","url":null,"abstract":"<p><strong>Objective: </strong>To systematically review and critically appraise the methodology of developing Modified Obstetric Early Warning Scores (MOEWSs).</p><p><strong>Study design and setting: </strong>We searched Medline, CINAHL, EMBASE and the Web of Science for MOEWS studies published between 01 January 2000 and 31 December 2023. Eligible studies included models predicting maternal death, ICU admission, and/or a composite of two or more maternal morbidities occurring in a hospital setting in women of any gestational age and up to one week after the end of pregnancy. Models were critically appraised using the Prediction Model Risk of Bias Assessment Tool (PROBAST) and adherence to the Transparent Reporting of Prediction Models (TRIPOD).</p><p><strong>Results: </strong>20 studies were included: five (25%) were model development studies, five (25%) were model development and validation, and ten (50%) were validation only. Four development studies used statistical methods, and the remaining six studies used clinical consensus (i.e., expert opinion). The four data-driven model development studies did not address key statistical challenges such as repeated measures or missing data, assess the performance adequately or dataset characterises clearly. All but one study (95%) were rated at high risk of bias due to data sources, poor reporting and analysis limitations. The fifteen validation studies were poorly reported and eleven (73%) were at high risk of bias. None of the data-driven models were independently validated, a key step towards implementation.</p><p><strong>Conclusions: </strong>There is a lack of MOEWS developed using methods that followed recommended statistical guidelines. Substantial problems with the methodological quality of included studies both in development and validation, along with high risk of bias indicating published scores could perform poorly or be potentially harmful if used in clinical practice. Future work should address handling missing data and repeated measures and consider how a MOEWS will perform in different populations and key subgroups.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111833"},"PeriodicalIF":7.3,"publicationDate":"2025-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144121510","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Endorsement of Reporting Guidelines and Trial Registration Remains Inconsistent Across Top Endocrinology Journals.","authors":"Chance Bratten, Hassan Khan, Diana King, Rohaan Muhammad, Christian Hemmerich, Caleb A Smith, Danya Nees, Griffin Hughes, Matt Vassar","doi":"10.1016/j.jclinepi.2025.111835","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111835","url":null,"abstract":"<p><strong>Background: </strong>Clinicians rely on evidence-based research for clinical practice to ensure safe, efficacious patient care. Reporting guidelines and clinical trial registration improve the quality of this research by increasing transparency and reducing the risk of biases. The extent of which endocrinology journals adopt the use of these tools is unclear. Therefore, the primary outcome of this study is to assess the recommendation and requirement of reporting guidelines and the secondary outcome is to assess clinical trial registration in the top endocrinology journals.</p><p><strong>Methods: </strong>The top 100 journals in the \"Endocrinology, Obesity, and Metabolism\" subcategory were identified using the 2021 Scopus CiteScore tool. The \"instructions to authors\" of each journal was analyzed for statements regarding select reporting guidelines outlined by the Enhancing the Quality and Transparency of Health Research (EQUATOR) Network as well as clinical trial registration. Statements were recorded as \"Not Mentioned,\" \"Recommended\", \"Does Not Require\", or \"Required\". To prevent unfair assessment, each journal was contacted to confirm the article types that are accepted.</p><p><strong>Results: </strong>Of 100 journals examined, CONSORT was the most commonly mentioned guideline with 46 journals recommending adherence and 36 journals requiring adherence. PRISMA was recommended by 61 journals and required by 19 journals. Finally, 77 journals required clinical trial registration.</p><p><strong>Conclusion: </strong>Our study reveals a lack of consistent endorsement of reporting guidelines in top endocrinology journals. This may undermine transparency and introduce bias. The main limitation of this study is the narrow scope of the study leading to a low generalizability. We suggest that journal editors in this field enforce validated reporting guidelines more strictly in order to improve the quality of published research.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111835"},"PeriodicalIF":7.3,"publicationDate":"2025-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144121506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Plug-and-play use of tree-based methods: Consequences for clinical prediction modelling.","authors":"Lotta M Meijerink, Ewoud Schuit, Karel G M Moons, Artuur M Leeuwenberg","doi":"10.1016/j.jclinepi.2025.111834","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111834","url":null,"abstract":"<p><strong>Objective: </strong>Tree-based models such as Random Forest and XGBoost are increasingly being used for clinical prediction, but certain aspects of their behavior are often overlooked. This article aims to illustrate these aspects and discuss the implications of plug-and-play use of tree-based models for clinical prediction. We focus on their ability to learn smooth, monotonic (i.e., consistent predictor effect where an increase in predictor leads to an increase in predicted risk), and additive predictor-outcome associations (i.e., each predictor independently and additively contributes to the outcome), and how they behave when making predictions outside the range of observed data (extrapolation).</p><p><strong>Study design and setting: </strong>We illustrated the behavior of plug-and-play use of tree-based models in a simulation study where we sampled predictors from standard normal distributions and binary outcomes determined by the logistic function of the predictors, and translate this into potential clinical implications in a real-world clinical example of post-radiotherapy toxicity prediction setting. To show the generalizability of our findings we also assessed the model's behavior in a publicly available dataset of head and neck cancer patients. For each analysis we visualized the learned predictor-outcome associations across different sample sizes.</p><p><strong>Results: </strong>In the simulation study, the models show stepwise fluctuations in their learned continuous predictor-outcome associations, caused by the inherent categorization of continuous predictors in a decision tree. Even with a large data size, the associations were not smooth or monotonic. Furthermore, because tree-based models can only split orthogonally to the axes, they struggle to learn an additive effect. Additionally, tree-based models extrapolate in a somewhat unintuitive way, by predicting a constant value beyond the observed data, regardless of further increases in predictor values. Using the clinical example and case study, we highlight that the learned associations are biologically implausible and may lead to issues regarding generalizability and trustworthiness.</p><p><strong>Conclusion: </strong>Using tree-based models in a plug-and-play manner for clinical prediction may result in undesirable predictor-outcome associations. Therefore, we recommend carefully taking their behavior into account during modeling decisions and evaluations. Further research is needed to explore the potential value of recent developments in decision tree literature, such as using constraints to incorporate prior knowledge and using soft split decision trees.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111834"},"PeriodicalIF":7.3,"publicationDate":"2025-05-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144121512","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"What makes a good guideline? - A systematic review and analysis of 120 clinical practice guidelines using the AGREE II tool.","authors":"Marina L Fotteler, Thomas D Kocar, Jana Willems, Sebastian Voigt-Radloff, Christoph Leinert, Dhayana Dallmeier, Clara Hertneck, Michael Denkinger","doi":"10.1016/j.jclinepi.2025.111830","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111830","url":null,"abstract":"<p><strong>Introduction: </strong>Clinical practice guidelines (CPGs) translate evidence into actionable recommendations to enhance care quality, improve clinical outcomes, reduce treatment variations, and make healthcare delivery more cost-effective. CPGs primarily aid healthcare practitioners but are also used by patients, policymakers, and organizations. The study aim was to assess CPG quality using the AGREE II instrument and to identify AGREE II items and domains that influence the overall assessment.</p><p><strong>Methods: </strong>Medline and eight CPG databases were searched for guidelines applicable to older patients (≥60 years, frail, or with dementia/delirium) in acute orthopedic/traumatological settings, published in English or German since January 1st, 2016, and employing an evidence appraisal. Titles, abstracts, and full texts were independently screened by two reviewers using Covidence. AGREE II assessments were conducted across all 23 items in six domains by three reviewers with different professional backgrounds. Data were analyzed using descriptive statistics, Mann-Whitney U test, Pearson's r correlation matrix, variance inflation factor (VIF), univariable and multivariable regression (non-negative least squares), and intraclass correlation coefficient (ICC). Significance was set at p<0.05.</p><p><strong>Results: </strong>A total of 120 CPGs have been appraised, reaching a mean overall rating of 4.35 (±1.13). Most guidelines received an overall rating of five (n=40, 33.33%), one guideline received an overall rating of one (0.8%). Using a standardized evidence rating framework (e.g. GRADE) is significantly associated with a better overall rating (p<0.001). The multivariable analysis showed that items 9, 12, and 15 had the highest influence on the overall AGREE II rating. Domain 6 (editorial independence) did not have an influence on the overall rating in a multivariable analysis.</p><p><strong>Conclusion: </strong>Methodological rigor, particularly the use of a standardized evidence rating framework, is essential for a good overall AGREE II rating and thus for high-quality CPGs. The results from this analysis can assist different stakeholders who also conduct AGREE II appraisals, develop CPGs, or are charged with implementing CPG recommendations.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111830"},"PeriodicalIF":7.3,"publicationDate":"2025-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144081833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identifying predictors of early trial termination: A meta-epidemiologic study utilising elements of the research ethics committee evaluation.","authors":"Marieke S Jansen, Lodewijk A Pet, Jeroen T Buijs, Bob Siegerink, Rolf H H Groenwold, Olaf M Dekkers","doi":"10.1016/j.jclinepi.2025.111832","DOIUrl":"https://doi.org/10.1016/j.jclinepi.2025.111832","url":null,"abstract":"<p><strong>Objectives: </strong>Early trial termination remains frequent. Research ethics committees (RECs) could play a role in reducing the probability of early termination. Their review process provides a window for both identifying trials at high risk of terminating prematurely and imposing preventive measures, such as design modifications. This study aimed to explore whether characteristics of ethics review, alongside trial characteristics, are related to subsequent early trial termination.</p><p><strong>Study design and setting: </strong>This meta-epidemiologic cohort study assessed 198 clinical trials approved by a Dutch REC between 2015 - 2018. Data from archived trial protocols, related study documents and correspondence between the REC and investigators were analysed to identify predictors of early termination during ethics review.</p><p><strong>Results: </strong>Of the 198 trials, 69 (34.8%) terminated early, most often due to recruitment failure (n = 31, 35.2%). Several characteristics were associated with early termination, such as multicentre design (vs. single centre, RR: 1.89, 95% CI: 1.24-3.14), number of comments raised during ethics review (per comment, RR: 1.02, 95% CI: 1.00-1.05), and specific comments regarding privacy and confidentiality (RR: 1.21, 95% CI: 1.05-1.41) and participant information sheets (RR: 1.05, 95% CI: 1.02-1.08). Investigator sponsorship, longer review durations, and comments raised regarding privacy and confidentiality and subject selection were associated with an increased likelihood of recruitment failure, specifically.</p><p><strong>Conclusion: </strong>This exploratory study showed that various characteristics of ethics review have the potential to predict early trial termination. Further studies are needed to validate and expand upon these findings.</p>","PeriodicalId":51079,"journal":{"name":"Journal of Clinical Epidemiology","volume":" ","pages":"111832"},"PeriodicalIF":7.3,"publicationDate":"2025-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144081855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}