BMC Medical Research Methodology最新文献

{"title":"Validating methods for inferring co-occurring diseases: a flexible framework for simulating synthetic data.","authors":"Hannah Marchi, Sophie Schmiegel, Tamara Schamberger, Christiane Fuchs","doi":"10.1186/s12874-026-02966-2","DOIUrl":"https://doi.org/10.1186/s12874-026-02966-2","url":null,"abstract":"<p><strong>Background: </strong>The validation of methods is an integral part of statistical research, defining conditions under which methods yield reliable results. Empirical validation requires a solid data basis to control and manage relevant characteristics like sample size, dimensionality, and underlying dependency structures. Real-world data often fails to meet these requirements, particularly in medical contexts where privacy regulations restrict availability. For this reason, synthetic data is an effective alternative for method validation. However, generating synthetic data is demanding when it must precisely mirror complex dependence structures while simultaneously controlling specific target characteristics.</p><p><strong>Methods: </strong>We address the medical context of co-occurring diseases, where symptoms may overlap or conflict. We propose a four-step framework to generate synthetic data for the simulation-based validation of statistical methods. The framework involves: (I) generating patient covariates; (II) connecting this information to predictors for single or joint disease occurrence; (III) transforming predictors into disease probabilities or scores; and (IV) converting these into disease occurrences. Each step offers several alternatives for modeling the overall dependence structure. We apply our framework to a case study of pain-causing diseases which share certain similarities in their clinical presentations, and which can occur either individually or jointly. By employing five combinations of methodological alternatives, we evaluate the approaches' ability to achieve target characteristics and demonstrate their specific strengths and weaknesses.</p><p><strong>Results: </strong>Matching the data-generating process with the estimation method allows for the successful recovery of input information, such as coefficients and correlations. Target properties like disease prevalence and associations are achieved to varying degrees depending on the methods used.</p><p><strong>Conclusions: </strong>While the proposed theory-driven framework is broadly applicable beyond the specific medical use case, it relies on careful, domain-informed parameter curation to generate meaningful synthetic datasets. Its flexible, adjustable input settings enable researchers to tailor data generation to their precise methodological requirements, providing a controlled basis for simulation-based validation without implying direct clinical inference.</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":"26 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13488252/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148788168","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":"Causal effect estimation for ordinal outcomes: a performance comparison of G-computation, IPW, and TMLE-based methods.","authors":"Aeran Park, Hyonggin An","doi":"10.1186/s12874-026-02951-9","DOIUrl":"10.1186/s12874-026-02951-9","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Ordinal outcomes are frequently used in clinical and observational studies. However, because ordinal variables do not have clearly defined intervals between categories, causal effect estimation for ordinal outcomes is more challenging than that for binary or continuous outcomes. Therefore, accurate causal inference for ordinal outcomes is important, as inadequate covariate adjustment may lead to biased estimates due to confounding. In this study, we defined the average treatment effect (ATE) for ordinal outcomes and compared the performance of various causal inference methods for estimating the ATE.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;A total of nine methods were compared, including the adjusted estimator, parametric- and Super Learner-based G-computation, inverse probability weighting (IPW), and both threshold-based and continuous-score Targeted Maximum Likelihood Estimation (TMLE). Various simulation scenarios were constructed according to whether the outcome and treatment models were correctly specified. The methods were evaluated in terms of bias, empirical standard error (ESE), and root mean squared error (RMSE). In the real-data analysis, the Primary Biliary Cirrhosis (PBC) dataset was used to estimate the average treatment effect of D-penicillamine treatment, and the results were compared across methods using confidence intervals. Confidence intervals were constructed using model-based, bootstrap-based, or influence function-based variance estimation according to the characteristics of each estimator.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Simulation results showed that, under the specific data-generating mechanisms considered in this study, TMLE-based methods generally exhibited the most stable overall performance. In particular, threshold-based TMLE maintained relatively low bias and RMSE under single-model misspecification scenarios. In contrast, G-computation- and IPW-based methods were sensitive to misspecification of the outcome and treatment models, respectively, and IPW-based methods exhibited substantial variability. When both models were simultaneously misspecified, the performance of most methods deteriorated. The Super Learner library, consisting of a generalized linear model and a random forest, improved the performance of several outcome model-based methods but did not fully mitigate the variability observed in IPW-based methods. For TMLE-based approaches, the Super Learner-based and parametric implementations showed comparable performance. In the real-data analysis, most methods yielded negative treatment effect estimates, suggesting a potential reduction in disease severity associated with treatment. However, because most 95% confidence intervals included zero, evidence for a statistically significant treatment effect was limited. Methods based on the Super Learner library generally produced narrower confidence intervals, suggesting more stable estimation performance. Nevertheless, differences in treatm","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":"26 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13678972/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149223190","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":"Applications of causal and structural equation modeling in epidemiology: a systematic and critical review.","authors":"Scholastique Midokpè Merveille Essetcheou, Houétchénou Gislain Fortuné Dovonou, Souand Peace Gloria Tahi, Sèton Calmette Ariane Houetohossou, Valère Kolawolé Salako, Marcel Tadogbè Donou Hounsode, Romain Glèlè Kakaï","doi":"10.1186/s12874-026-02944-8","DOIUrl":"10.1186/s12874-026-02944-8","url":null,"abstract":"<p><strong>Background: </strong>Structural equation modeling (SEM) and causal modeling (CM) are powerful statistical approaches for identifying complex interrelationships among variables. However, their application in epidemiology remains limited and under-documented, especially in infectious disease research, which requires integrated analytical frameworks for effective control.</p><p><strong>Methods: </strong>To examine how SEM and CM have been applied, their methodological characteristics, and reporting practices, a systematic and critical review was conducted following PRISMA guidelines. The search covered studies published between 1987 and 2025 across PubMed, Scopus, Web of Science, ScienceDirect, SpringerLink, Google Scholar, and the Directory of Open Access Journals. After rigorous screening, 458 articles were thoroughly evaluated.</p><p><strong>Results: </strong>Most studies focused on neuropsychiatric (32.1%) and chronic (30.1%) conditions, with few addressing infectious diseases (24.0%), primarily malaria, tuberculosis, and HIV, particularly in low-income countries where context-specific evidence is urgently needed to inform targeted interventions. SEM studies predominantly used maximum likelihood estimation (57.7%) and large samples ([Formula: see text] observations in 85%) with CB-SEM remaining the dominant approach across all sample size categories. In contrast, CM studies showed substantial variability in sample sizes across approaches, ranging from fewer than 100 to over 200 observations (coefficient of variation [Formula: see text]), with no consistent sample size threshold across methods. Methodological reporting was often incomplete, notably regarding study design (17.4%), measurement validity (15.8%), and model fit criteria (5.2%), reducing transparency and reproducibility.</p><p><strong>Conclusion: </strong>Overall, broader application of SEM and CM to infectious diseases, combined with improved methodological transparency, could substantially strengthen causal inference and guide evidence-based disease control strategies. Moreover, integrating longitudinal study designs would further enhance the robustness and interpretability of causal findings.</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":" ","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13628753/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148547618","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":"A systematic review and meta synthesis of barriers and facilitators to Black women's participation in healthcare research.","authors":"S S Hall, G Clancy, S Patel, N Thorpe, B Majmudar, C Merriman, S Agyapong, J Usman, C Henshall","doi":"10.1186/s12874-026-02953-7","DOIUrl":"10.1186/s12874-026-02953-7","url":null,"abstract":"<p><strong>Background: </strong>Black women are under-represented in clinical research, contributing to persistent health inequities and undermining the validity and generalisability of research findings. Understanding facilitators and barriers to their research engagement is essential to addressing these challenges. This study aimed to identify and synthesise evidence on barriers and facilitators which influence Black women's participation in healthcare research, to inform inclusive recruitment strategies and equitable research designs.</p><p><strong>Methods: </strong>This systematic review was conducted in accordance with PRISMA guidelines and registered with PROSPERO (CRD42024587308). Findings were narratively synthesised using thematic analysis and discussed iteratively through Patient and Public Involvement. Five databases were searched (CINAHL, PsycInfo, Embase, MIDIRS, MEDLINE) for studies published since 2010. Papers in any language were screened for inclusion. Patient and Public Involvement members informed review objectives and data interpretation. Primary research studies reporting on Black women's experiences, attitudes, and facilitators/barriers to participating in any type of healthcare research. Studies were included if ≥ 50% (to ensure primary representation) of the sample recruited were Black women.</p><p><strong>Results: </strong>Seventy-two studies were included. Most were from America and spanned research areas relating to oncology, Human Immunodeficiency Virus (HIV), dementia, maternity, and general health. Five overarching themes identified how Black women's participation in healthcare research is shaped by a complexity of factors. These were mistrust, interpersonal experiences, altruism, self-interest, and low health/research literacy and awareness. Patient and Public Involvement consultations concurred that the findings aligned with a \"bridge and barriers\" analogy, illustrating how various factors can facilitate or prevent Black women from engaging with research opportunities. This provided a clear framework to guide culturally sensitive implementation for researchers, healthcare professionals, and policymakers.</p><p><strong>Conclusions: </strong>Inclusive strategies for recruiting Black women into research must focus on co-constructing a bridge between communities and research systems. This requires structural and systemic change in research delivery environments and ongoing collaborations with all stakeholders including Black women, healthcare system providers, and research teams.</p><p><strong>Trial registration: </strong>PROSPERO (CRD42024587308).</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":" ","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13673856/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148547665","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":"Robust power and sample size calculations in quasi-likelihood models: methods and practice.","authors":"Shijie Yuan, Amy Cochran, Paul Rathouz","doi":"10.1186/s12874-026-02940-y","DOIUrl":"10.1186/s12874-026-02940-y","url":null,"abstract":"<p><strong>Background: </strong>Accurate power and sample size calculations are essential in study planning, yet they are often difficult to carry out for quasi-likelihood (QL) models. Traditional power and sample size (PSS) approaches often rely on restrictive distributional assumptions, limiting their applicability when responses have non-standard distributions, variance functions are misspecified, or when covariates exhibit complex dependence structures.</p><p><strong>Methods: </strong>We examine whether two effect size measures-2 Standard Deviations in the Linear Predictor (2SLiP) and Pseudo-Partial [Formula: see text] (P2R2)-originally developed for Wald tests involving generalized linear models, are effective at power and sample size calculations in the QL framework. Through extensive simulations across diverse outcome types, link functions, and variance structures, we assess their performance under Wald tests and explore whether they remain useful for score tests. To illustrate practical utility, we apply these effect size measures to survey data on frontline health care workers to quantify the association between perceived personal protective equipment (PPE) adequacy and burnout risk during the COVID-19 pandemic, adjusting for covariates.</p><p><strong>Results: </strong>We show that the two generalized linear model (GLM)-based effect sizes are fundamentally moment-based objects, and therefore extend directly to QL models. Across all simulation settings, both measures remained accurate, with sample size and power estimates within 3% and 2% of the target, respectively. In the case study, the estimated effect sizes for perceived PPE adequacy (2SLiP = 0.096 and P2R2 = 0.020) correspond to small but meaningful associations with burnout risk after adjustment for demographic and occupational covariates; both yielded accurate recommendations for sample size.</p><p><strong>Conclusions: </strong>Both 2SLiP and P2R2 offer robust alternatives for power and sample size calculation in QL settings. By requiring minimal distributional assumptions, these measures enhance the flexibility and reliability of power and sample size calculations for realistic study designs commonly encountered in medical and public health research.</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":" ","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13670577/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148547755","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":"Bayesian semi-parametric inference for joint modelling of childhood growth and appetite phenotypes.","authors":"Andrea Cremaschi, Beatrice Franzolini, Maria De Iorio, Mary Chong, Jia Ying Toh, Navin Michael, Varsha Gupta, Fabian Yap, Yung Seng Lee, Johan Erikkson, Anna Fogel","doi":"10.1186/s12874-026-02950-w","DOIUrl":"10.1186/s12874-026-02950-w","url":null,"abstract":"<p><strong>Purpose: </strong>Early eating behaviours have been associated with later weight outcomes, yet their longitudinal development and relationship with growth trajectories remain insufficiently understood. This study aims to investigate how appetite-related behaviours in early childhood co-evolve with growth patterns and contribute to obesity risk.</p><p><strong>Methods: </strong>We develop a Bayesian semi-parametric joint modelling framework to analyse repeated measures of growth indicators, such as body mass index, alongside questionnaire-based eating behaviour scores collected at multiple time points in children from the Singaporean GUSTO cohort (\"Growing Up in Singapore Towards Healthy Outcomes\"). The approach extends established models for ordinal questionnaire data to accommodate longitudinal observations and covariate effects, while growth trajectories are flexibly represented using spline-based regressions. Subject-specific demographic and clinical covariates are incorporated into both components of the model, allowing their effects on growth and eating behaviour to be assessed simultaneously. The two components are linked through a Bayesian nonparametric prior, specifically a Normalised Generalised Gamma Process, that enables data-driven identification of subgroups of children with similar developmental profiles.</p><p><strong>Results: </strong>The proposed framework captures the dynamic interplay between appetite phenotypes and growth trajectories over time, allowing the identification of clinically meaningful clusters characterised by distinct behavioural and growth patterns.</p><p><strong>Conclusion: </strong>The proposed integrated modelling strategy provides a nuanced understanding of how eating behaviours and growth co-develop in early life, offering new insights into mechanisms underlying childhood obesity risk and supporting the design of targeted early interventions.</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":" ","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13670690/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148535289","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":"Estimating treatment effects from non-overlapping cohorts with application to antimicrobial resistance.","authors":"Avi Baraz, Daniel Nevo, Amos Cahan, Tal Brosh-Nissimov, Michal Chowers, Uri Obolski","doi":"10.1186/s12874-026-02947-5","DOIUrl":"10.1186/s12874-026-02947-5","url":null,"abstract":"<p><strong>Background: </strong>The positivity assumption, also known as overlap, is one of the key assumptions for causal inference. However, in certain observational studies it does not hold. This work is motivated by such a scenario, where the goal was to compare the effect of ceftriaxone versus cefuroxime treatment on subsequent 3rd generation cephalosporin resistance within one year post-treatment in Gram-negative bacteria among hospitalized patients. Our data consisted of electronic medical records from 24,969 patients hospitalized between 2015-2025 in two hospitals with distinct policies: one prescribed cefuroxime, while the other prescribed ceftriaxone. These prescription policies created two non-overlapping patient cohorts, as each hospital treated patients with either cefuroxime or ceftriaxone. We developed and applied a statistical framework that leverages additional data, for which positivity holds, and clinically driven assumptions, to identify the average treatment effect among these non-overlapping cohorts, without removing any observations.</p><p><strong>Methods: </strong>Our framework is based on a logistic regression model for the conditional expected bacterial resistance outcome, and it can be applied to any generalized linear model. The model is identifiable under a clinically driven assumption that the effect modification of the hospital on cefazolin equals that of cefuroxime and ceftriaxone, when compared to penicillins. We rely on cefazolin and penicillins being prescribed in both hospitals, and on penicillins inducing negligible resistance to 3rd generation cephalosporins. We also implemented inverse probability of censoring weighting for censored outcomes, and utilized aminoglycoside resistance as a negative control outcome to assess sufficient confounding adjustment.</p><p><strong>Results: </strong>The estimated average treatment effect, on the risk difference scale, of ceftriaxone compared to cefuroxime was 5.83% (95% CI: -2.06%, 13.39%). The conditional odds ratio for ceftriaxone compared to cefuroxime, adjusted for confounders, was 1.31 (95% CI: 0.91, 1.89). No evidence of unmeasured confounding was found using the negative control outcome falsification test.</p><p><strong>Conclusions: </strong>We develop and apply a statistical framework applicable in settings of non-overlapping cohorts that violate the positivity assumption. Our framework allows identification of the average treatment effect thanks to parametric assumptions about the conditional expected outcome, additional data and clinically driven assumptions. The empirical findings, though not statistically significant, may point to a modest resistance increase with ceftriaxone compared to cefuroxime use; and suggest further research about the preferential use of cefuroxime is needed.</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":" ","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13670598/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148535303","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":"The Semmelweis Framework: structuring evidence foundations and expert opinion in clinical practice guidelines.","authors":"Blin Nagavci, Margaret Renn Andrews, Jens Vikse, Paul Studenic, Tanja Stamm, Botond Lakatos","doi":"10.1186/s12874-026-02938-6","DOIUrl":"10.1186/s12874-026-02938-6","url":null,"abstract":"<p><strong>Background: </strong>For many guideline-developing organisations, expert opinion remains essential when evidence is scarce. However, its unstandardised use remains a significant gap in guideline methodology. This article presents the development of the Semmelweis Framework, a methodological tool for structuring expert opinion in clinical practice guidelines when evidence is limited, indirect, or absent.</p><p><strong>Methods: </strong>The Semmelweis Framework was developed through a stepwise methodological process. First, a meta-epidemiological study mapped expert opinion use across 2,296 recommendations from 136 organisations. Second, a methodological review of 98 guideline manuals synthesised the functional definition of expert opinion that informed the Framework. Building on this definition, the Framework was refined through independent multi-reviewer application to 30 published recommendations, with calibration meetings, qualitative comparison of classifications, and parallel development of a Preconditions Checklist.</p><p><strong>Results: </strong>The Semmelweis Framework offers an approach for structuring the role of expert opinion in clinical guidelines by classifying recommendations in four classes according to their evidentiary foundation. These classes form a continuum: from recommendations based on direct comparative evidence to those relying entirely on expert opinion when no relevant group-level evidence exists. Class I is grounded in direct comparative evidence. Class II addresses indirect evidence and requires explicit labelling of indirectness while using expert opinion to support applicability. Class III relies on non-comparative evidence and is heavily influenced by expert opinion, whereas Class IV is based entirely on expert opinion, operationalised as structured expert judgement, when no group-level evidence exists. Across all classes, fixed verbs and notations are proposed to standardise communication. A core component of the Framework is the specifically developed Preconditions Checklist, designed to ensure integrity and transparency when recommendations rely substantially on expert opinion. Under the Semmelweis Framework, \"expert opinion\" should not be used as a generic recommendation label; instead, its roles are redistributed across recommendation classes, each with explicit foundations and labelling.</p><p><strong>Conclusions: </strong>The Semmelweis Framework reframes the role of expert opinion and offers an approach for structuring its use in clinical guidelines. By introducing safeguards and new recommendation classes according to their evidentiary foundations, the Framework addresses a long-standing methodological gap.</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":" ","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13540927/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148497142","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":"A stakeholder-inclusive conceptual framework for modeling routinely collected health data for therapeutic decision-making illustrated by means of multistate models.","authors":"Michelle Pfaffenlehner, Andrea Dreßing, Dietrich Knoerzer, Markus Wagner, Peter Heuschmann, André Scherag, Harald Binder, Nadine Binder","doi":"10.1186/s12874-026-02946-6","DOIUrl":"10.1186/s12874-026-02946-6","url":null,"abstract":"<p><strong>Background: </strong>Routinely collected health data are increasingly used to generate real-world evidence for therapeutic decision-making. Their use, however, depends on the expectations of multiple stakeholders. Clinicians require clinically interpretable analyses, pharmaceutical stakeholders need robust evidence on effectiveness and safety, patient advocacy groups emphasize transparency, privacy, and meaningful outcome measures, and statisticians focus on bias control, reproducibility, and methodological rigor. Without explicit consideration of these perspectives, analyses risk being fragmented, misaligned with end-user needs, or lacking transparency. Aligning these perspectives early in the design of routine data analyses therefore remains a central challenge.</p><p><strong>Methods: </strong>We developed a stakeholder-inclusive conceptual framework for modeling routine health data, through expert panel discussions, an interdisciplinary workshop and targeted literature examples. The synthesis focused on four stakeholder perspectives: clinicians, pharmaceutical industry, patient advocates, and statisticians. To illustrate how stakeholder priorities can be translated into analytical strategies, we reviewed selected applications of multistate models (MSMs) in routine health data settings.</p><p><strong>Results: </strong>The conceptual framework links stakeholder-specific priorities, methodological requirements and identifies shared needs for analyses that are clinically meaningful, transparent, reproducible, and able to represent patient pathways, intermediate events, treatment trajectories, disease progression, safety outcomes, and patient-reported measures. While the framework is intended to be applicable across various analytical approaches MSMs are used here to illustrate how these diverse requirements can be operationalized in practice. They can capture longitudinal health processes, competing events, recurrent or intermediate states, and state-specific outcomes while retaining an interpretable graphical structure, and the reviewed examples show their applicability across different research questions using routine health data. Beyond specific methodological choices, clinical research relies fundamentally on statistical expertise. The framework also highlights that the statistician's role varies with the complexity of the research question, ranging from consultation on standard analyses to adaptation or development of advanced methods.</p><p><strong>Conclusions: </strong>The stakeholder-inclusive framework provides methodological guidance for designing analyses of routine health data that are clinically meaningful, scientifically rigorous, and socially acceptable. By aligning the research question with the intended perspective from the beginning, it supports more robust and transparent evidence generation, with multistate models serving as a flexible tool to operationalize this integration.</p>","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":"26 1","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13366620/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148435238","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":"Adaptive Gaussian process search for simulation-based sample size estimation in clinical prediction models: validation of the pmsims R package.","authors":"Oyebayo Ridwan Olaniran, Diana Shamsutdinova, Sarah Markham, Felix Zimmer, Daniel Stahl, Gordon Forbes, Ewan Carr","doi":"10.1186/s12874-026-02935-9","DOIUrl":"10.1186/s12874-026-02935-9","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Determining an adequate sample size is essential for developing reliable and generalisable clinical prediction models, yet practical guidance on selecting appropriate sample size calculation methods remains limited. Existing analytical and simulation-based tools impose restrictive assumptions and focus on mean-based criteria. We present and validate pmsims, an R package that uses Gaussian process (GP) surrogate modelling to provide a flexible and efficient simulation-based framework for sample size determination, validated here across binary, continuous, and survival prediction modelling contexts.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;We conducted a comprehensive simulation study with two aims. Aim 1 compared three search engines implemented in pmsims, a GP surrogate-based adaptive procedure (gp), a deterministic bisection method (bisection), and a hybrid GP-bisection approach (gp-bs), across binary, continuous, and survival outcomes. Scenarios varied outcome prevalence or event rate, predictor dimensionality ([Formula: see text]), target performance metric (discrimination and calibration slope), aggregation criterion (mean vs 80% assurance), and total simulation budget ([Formula: see text]). Each scenario was replicated 100 times; estimator stability was assessed via the coefficient of variation (CV). Aim 2 benchmarked the best-performing pmsims engine against pmsampsize (analytical) and samplesizedev (simulation-based) across a wider range of realistic scenarios, evaluating recommended sample sizes, computational time, and achieved model performance on independent validation datasets of 30,000 observations.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;The GP-based search engine consistently yielded the most stable sample size estimates (lowest CV) across all outcome types, ranking highest in 9/12 outcome-aggregation metric configurations. Its advantage was most pronounced in low-signal, high-dimensional settings, and was accentuated with [Formula: see text] replications per evaluation and a budget of [Formula: see text], after which gains were minimal. In benchmark comparisons, pmsims (mean) achieved performance deviations within [Formula: see text] of the prespecified target across binary, continuous, and survival outcomes, comparable to samplesizedev and substantially outperforming pmsampsize in high-discrimination settings (deviations up to [Formula: see text]).&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;The pmsims package, using the GP-based search engine with [Formula: see text] replications per evaluation and a budget of [Formula: see text], provides a computationally efficient and flexible framework for principled sample size planning in clinical prediction modelling. It reliably achieves performance targets across the range of standard-model scenarios evaluated while requiring fewer model evaluations than non-adaptive simulation approaches, offering a compelling alternative to both analytical formulae and exhaustive simulation-bas","PeriodicalId":9114,"journal":{"name":"BMC Medical Research Methodology","volume":" ","pages":""},"PeriodicalIF":3.7,"publicationDate":"2026-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13640277/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148429770","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}