Clinical Trials

BackgroundIncomplete dissemination of clinical trial results remains an important challenge for research transparency and evidence synthesis. Although prior studies have quantified the overall extent of non-dissemination, less is known about whether trial characteristics observable at registration are associated with subsequent dissemination within sponsor portfolios. Methods and findingsWe conducted a retrospective cohort study of 17,537 completed interventional clinical trials registered on ClinicalTrials.gov between 2007 and 2024 across the 20 largest global pharmaceutical companies. We developed the Operational Complexity Index (OCI), a composite measure derived from planned enrollment, facility count, and geographic scope, and examined its association with trial dissemination using multivariable logistic regression and time-to-event analyses. Higher OCI was associated with greater odds of dissemination (adjusted odds ratio [aOR] = 2.40, 95% CI 2.23-2.60; p < 0.001), with dissemination increasing from 47% in the lowest OCI decile to 95% in the highest. Higher operational complexity was also associated with earlier dissemination; over a 1,095-day horizon, high-OCI trials were disseminated a mean of 310.88 days earlier than low-OCI trials (RMST difference, 310.88 days; 95% CI 300.59-320.96; p < 0.001). This pattern was observed across sponsors, clinical phases, and therapeutic areas. In predictive analyses using registration-time variables, the structural model achieved a cross- validated AUC of 0.816 and a holdout AUC of 0.814, whereas the full model, including sponsor identity, achieved a cross-validated AUC of 0.858 and a holdout AUC of 0.857. Using benchmark phase-based costing assumptions, the 5,019 non-disseminated trials corresponded to an estimated US$10.94-15.26 billion in sunk research investment. ConclusionsAmong trials conducted by the 20 largest pharmaceutical sponsors, greater operational complexity at registration was associated with a higher likelihood of dissemination and earlier dissemination. These findings suggest that aggregate sponsor-level transparency metrics may mask important heterogeneity within sponsor portfolios. Future work should assess whether registration-time trial characteristics can help identify trial subgroups at higher risk of non-dissemination. AUTHOR SUMMARYO_ST_ABSWhy was this study done?C_ST_ABSO_LIIncomplete dissemination of clinical trial results reduces the completeness of the medical evidence base and the public value of research participation. C_LIO_LIPrevious studies have described overall rates of trial non-dissemination, but less is known about whether dissemination varies systematically across different types of trials within sponsor portfolios. C_LIO_LIWe examined whether trial characteristics available at registration were associated with later dissemination of results among large pharmaceutical sponsors. C_LI What did the researchers do and find?O_LIWe analyzed 17,537 completed interventional clinical trials sponsored by the 20 largest pharmaceutical companies and registered on ClinicalTrials.gov between 2007 and 2024. C_LIO_LIWe developed an Operational Complexity Index (OCI) based on planned enrollment, number of facilities, and geographic scope to measure trial operational scale at registration. C_LIO_LIHigher OCI was associated with a greater likelihood of dissemination and earlier dissemination. Dissemination ranged from 47% in the lowest OCI decile to 95% in the highest. C_LIO_LIThis pattern was observed across sponsor portfolios, clinical phases, and therapeutic areas, with an average within-sponsor dissemination gap of 40 percentage points between lower- and higher-complexity trials. C_LIO_LIIn manual validation of 344 sampled trials, the automated dissemination-classification pipeline achieved 92.1% accuracy. C_LIO_LIUsing benchmark phase-based costing assumptions, the 5,019 non-disseminated trials corresponded to an estimated US$10.9-15.3 billion in sunk research investment. C_LI What do these findings mean?O_LIDissemination was not uniform across trial types within sponsor portfolios; trials with lower operational complexity were less likely to be disseminated than trials with higher operational complexity. C_LIO_LIAggregate sponsor-level transparency measures may therefore miss important differences within portfolios. C_LIO_LIRegistration-time trial characteristics showed predictive signal for non-dissemination, but whether such information could support monitoring strategies would require prospective validation. C_LIO_LIMore complete dissemination of trial results would strengthen the scientific record and improve the public value of clinical research. C_LI

Background Simon two-stage designs for binary endpoints and their time-to-event analogues, including the Kwak and Jung method, rely on a fixed null benchmark. Their Type I error control is valid only when that benchmark is correctly specified. In practice, historical benchmarks are often inconsistent due to small samples, population heterogeneity, changing eligibility criteria, and evolving standards of care. Even modest misspecifications can substantially inflate the Type I error rate, leading to costly advancement of ineffective treatments. Methods We propose the Interval-Null Robust (INR) two-stage design framework that accounts for uncertainty in the historical null benchmark. We define the null hypothesis as a plausible range of clinically uninteresting values: p[isin][p0L, p0U] for binary endpoints and {lambda}[isin][{lambda}0L, {lambda}0U] (or equivalent survival probabilities) for time-to-event endpoints. Type I error is controlled uniformly over the full null interval: sup{theta}[isin]{theta}0 Pr{theta}(Go) [&le;] . Under the monotonicity of the Go probability, the supremum occurs at the least favorable null configuration - p0U and {lambda}0L - but the design is not reduced to a point-null formulation. The interval defines the uncertainty set for error control and is used in selecting among feasible designs through robust criteria such as worst-case regret or minimal average expected sample size. Results Across representative planning scenarios for both endpoint types, classic designs calibrated to a single benchmark exhibit substantial Type I error inflation when the true null parameter exceeds the assumed planning value. INR designs maintain the nominal Type I error rate across the full null interval, directly addressing this vulnerability to benchmark misspecification. The robustness-efficiency trade-off can be managed through design constraints and robust optimization criteria while preserving uniform Type I error control. Conclusions INR two-stage designs offer a transparent framework for addressing historical control uncertainty in single-arm Phase II trials. By replacing reliance on a fixed benchmark assumption with a more realistic interval of clinically plausible null values, INR design reduces the risk of false-positive Go-decisions caused by benchmark misspecification. INR applies to both binary and time-to-event endpoints and is implemented in the open-source INRDesign R package and accompanying interactive Shiny app.

Target trial emulation (TTE) offers a principled way to estimate treatment effects using real-world observational data, but analyses of time-varying treatment strategies remain vulnerable to immortal time bias. The clone-censor-weight (CCW) approach is increasingly used to address this problem, yet key aspects of its causal interpretation and implementation remain unclear. In this work, we emulate a target trial using electronic health records (EHRs) to compare completion of a 3-dose 9-valent human papillomavirus vaccination (HPV) series within 12 months versus remaining partially vaccinated among vaccine initiators. We link CCW to the classic potential outcome framework in causal inference, evaluate the role of different weighting mechanisms, and account for within-subject correlation induced by cloning using cluster-robust variance estimation. Our study provides practical guidance for applying CCW in real-world comparative effectiveness studies to address immortal time bias and supports more rigorous and interpretable treatment effect estimation in TTE.

Randomized controlled trials (RCTs) provide high internal validity but often rely on restrictive eligibility criteria that limit generalizability and complicate real-world trial emulation. We propose AERO (AI Agent for Adaptive Eligibility Refinement and Optimization), an agentic framework that systematically adapts clinical trial eligibility criteria for application to electronic health record data. AERO integrates external clinical knowledge sources and large language model-based reasoning to classify criteria as strict inclusion, safety exclusion, confounder, or operational artifact. We evaluated AERO by emulating the WARCEF trial using Mayo Clinic Platform data restricted to the pre-trial completion period. Emulation with optimized criteria yielded a hazard ratio of 1.561 (p = 0.0605), consistent with the original neutral trial finding (HR = 1.01, p = 0.91). An ablation analysis demonstrated that eligibility handling decisions materially influence observed treatment effects. These results highlight the importance of systematic, knowledge-informed eligibility refinement in real-world evidence generation.

Machine learning models for causal inference aim to adjust for confounding factors that are associated with both an exposure and an outcome, creating a spurious biased association. But, these methods are rarely empirically evaluated to assess their success in mitigating such bias. Recent advances in knowledge representation, including both foundation models and knowledge graphs, could enrich these models, but rigorous evaluations are needed in order to assess their potential. Here, we ask whether enriching existing causal inference models with knowledge representations from foundation models can improve confounding control. Rather than using semi-simulated data to address this question, we focus on examples of real confounding: we emulate target randomized active comparator trials that are subject to confounding by indication. Our results can guide researchers aiming to develop or apply methods for discovering causal effects from observational data.

Large language models are increasingly used for clinical guidance while their parent companies introduce advertising. We tested whether pharmaceutical ads embedded in the prompts of 12 models from OpenAI, Anthropic, and Google shift drug recommendations across 258,660 API calls and four experiments probing distinct epistemic conditions. When two drugs were both guideline-appropriate, advertising shifted selection of the advertised drug by +12.7 percentage points (P < 0.001), with some model-scenario pairs shifting from 0% to 100%. Google models were the most susceptible (+29.8 pp), followed by OpenAI (+10.9 pp), while Anthropic models showed minimal change (+2.0 pp). When the advertised product lacked evidence or was clinically suboptimal, models resisted. This reveals a structured vulnerability: advertising does not override medical knowledge but fills the space where clinical evidence is underdetermined. An open-response sub-analysis (2,340 calls across three representative models) confirmed that advertising restructures free-text clinical reasoning: models echoed ad claims at 2.7 times the baseline rate while maintaining high stated confidence and rarely disclosing the ad. Susceptibility was provider-dependent (Google: +29.8 pp; OpenAI: +10.9 pp; Anthropic: +2.0 pp). Because this bias operates within clinically correct answers, it is invisible to accuracy-based evaluation, identifying a class of AI safety vulnerability that standard testing cannot detect.

Background: The global SARS-CoV-2 pandemic disrupted healthcare systems worldwide, raising concerns about its impact on clinical research. Early reports suggested reductions in participant enrollment, interruptions to ongoing trials, and challenges to protocol adherence, yet the magnitude and duration of these operational disruptions remain unclear. Methods: We conducted a registry-based analysis comparing clinical trials during the COVID-19 pandemic (December 2019 to November 2022) with a matched pre-pandemic cohort (December 2016 to November 2019). Studies were included if they reported any modifications to trial status, enrollment, or protocols during the study periods. Key variables included trial stoppage, enrollment changes, and adoption of remote or hybrid procedures. Results: The global SARS-CoV-2 pandemic resulted in widespread disruptions to trial operations with 13,323 clinical trials terminated, suspended or withdrawn over the course of the pandemic, a 38% increase compared to the 9,665 trials that stopped in the 3 years prior to the pandemic. Registries indicated a sharp decline in new participant enrollment across geographic regions and therapeutic areas, with partial recovery in later months. Review findings highlighted barriers including patient inaccessibility, staff redeployment, and supply chain interruptions. Conclusions: The pandemic caused system-wide operational shocks that compromised trial timelines and may have downstream methodological consequences. Recovery in enrollment does not imply restoration of pre-pandemic protocol fidelity or outcome ascertainment. Standardized reporting of disruptions, proactive contingency planning, and resilient trial designs are needed to maintain data integrity during large-scale disruptions and to support reliable evidence generation.

In February 2026, the FDA announced that a single pivotal phase 3 (P3) trial would become the new default standard for drug approval - a regulatory direction that had been legally enabled since the FDA Modernization Act of 1997. This announcement has strategic, scientific, and economic implications for drug developers, contract research organizations (CROs), and biotech investors. We argue that the expansion of this framework, originally reserved for various niche submissions, represents a paradigm change, dramatically increasing the value of rigorous early phase (P1 and P2) trial design, requiring sponsors to establish both statistical efficacy signals and mechanistic biological understanding before entering phase 3. Using a CNS indication cost model, we show that single P3 approval can reduce total development expenditure from approximately $447 million over 14 years to $297 million over 12 years - a savings of $150 million and providing two years of additional commercial runway for a modeled CNS drug. Case examples including lecanemab, omaveloxolone, and tofersen illustrate how biomarker-informed early phase strategies can establish the confirmatory evidence necessary for single-trial approval. We provide practical guidance for maximizing the value of P1 and P2 under this evolving framework.

Traditional audit methods that rely on written records often miss the nuances of clinical reasoning that influence patient care. Ambient artificial intelligence captures spoken clinical encounters, allowing the analysis of real clinician-patient dialogue at scale. In a study of 124 urology consultations, a transcript-centered audit identified inter-physician variation and expert disagreement that conventional review missed. We explore the epistemic gains of this approach, its nonverbal blind spots, behavioral effects, technical vulnerabilities, and the EU AI Acts regulatory landscape.

Background: Adolescent substance-use rehabilitation is a care-continuum problem spanning detection, engagement, active treatment, relapse prevention, aftercare, family support, and equity-oriented implementation. Existing reviews are often modality-specific and do not show how evidence aligns with substances, populations, outcomes, stages of care, or policy needs. Objectives: To map and synthesise the 2015-2025 adolescent and transitional-age youth SUD rehabilitation literature across intervention domains, stages, substances, outcomes, equity/disadvantage, geography, and economics, and to perform meta-analysis only where pooling was clinically defensible. Methods: PubMed, Scopus, and Web of Science records were harmonised to 2015-2025 and deduplicated. Two reviewer roles applied a predefined charting codebook for substance focus, technique family, rehabilitation stage, equity/disadvantage flags, outcome family, and study-design signal. Evidence was synthesised across AI/digital, psychiatric/psychotherapeutic, pharmacological, family/social, behavioural, residential/continuing-care, school/community, harm-reduction, and policy domains. Random-effects meta-analysis was restricted to comparative youth OUD medication-supported trials with extractable binary outcomes. Results: The search identified 1,676 records; 554 duplicates were removed, leaving 1,122 unique records. Metadata screening retained 579 records for evidence-map charting: 112 high-confidence records and 467 conservative metadata-supported records requiring full-text verification before final selective-journal submission. The charted evidence was concentrated in active treatment (n=433) and relapse prevention (n=114); aftercare/follow-up was weak (n=8). Intervention-family signals were led by pharmacological/MOUD (n=72), psychotherapy/psychiatric care (n=65), school/community/brief interventions (n=46), residential/continuing care (n=41), family/social therapy (n=30), AI/digital/telehealth (n=25), harm-reduction/policy (n=24), and CM (n=22). The primary youth OUD retention/completion meta-analysis favoured medication-supported treatment (OR 7.67, 95% CI 3.98-14.78; I^2=0%; k=2; n=188). An exploratory favourable-outcome analysis produced a similar estimate (OR 7.94, 95% CI 4.24-14.89; I^2=0%; k=3; n=229). Conclusions: The strongest pooled quantitative claim supports medication-supported treatment for youth OUD. For non-opioid substances, digital care, family therapy, CM, residential care, aftercare, and equity-oriented implementation, the literature is clinically important but not yet consistently synthesis-ready. Future trials should evaluate complete care pathways, adopt core outcomes, report age-banded and equity subgroup effects, and include economic and implementation endpoints.

ObjectiveStructured extraction from clinical free-text depends on human annotators whose labels are susceptible to errors and knowledge-driven mistakes; exhaustive quality control is impractical at scale. We evaluate whether disagreement among multiple locally hosted large language models (LLMs) can prioritize human annotations for targeted review. MethodsMultiple LLMs independently extract the same set of structured variables annotated by a human reviewer. For each annotation, an agreement score counts the LLMs matching the human label. Using four locally hosted LLMs (Gemma 3 27B, DeepSeek-R1 70B, GPT-OSS 120B, Mistral Large 3), we evaluated this approach on 910 German-language colonoscopy reports describing endoscopic mucosal resection, with five structured variables per case (anatomical location, two diameters, resection technique, multiple polyps), yielding 4,550 annotations and a 377-case adjudication sample. A stratified sample oversampling low-agreement strata was adjudicated blinded by an experienced reviewer and analyzed with prevalence-adjusted estimates ResultsHuman error rates rose as LLM agreement fell, from 0% at scores 3-4 to 76% at score 0. The lowest-agreement stratum was only 6.5% of annotations yet concentrated an estimated 80% of errors. The multi-LLM disagreement score achieved a prevalence-adjusted AUC-ROC of 0.991 (95% CI 0.987-0.994) and AUC-PR of 0.893 (95% CI 0.851-0.929) for error detection. DiscussionMulti-LLM disagreement outperformed single models and provided graded operating points for risk-stratified review. ConclusionMulti-LLM disagreement provides a scalable quality-control signal for targeted review of the highest-yield cases. Because all models run locally, the framework is GDPR-compliant; its language- and task-agnostic design supports application across clinical domains.

Precision medicine has given rise to a spectrum of biomarker-guided trial designs, from simple enrichment and strategy designs to more complex adaptive frameworks. To address the need for user-friendly tools that span this spectrum, we developed a unified R Shiny platform that first implements three standard designs: the randomize-all design, the enrichment design, and the biomarker-strategy design, allowing researchers to perform power and sample size calculations under each framework with intuitive inputs and visual outputs. Building on this foundation, the platform further extends to support two-stage general randomized basket trial designs with interim analysis, which can be viewed as a generalization of the standard designs to multiple biomarker-defined subgroups. The tool was rigorously validated by comparison with established R pipelines and published formulas, and user testing confirmed its intuitive interface. By providing seamless integration from standard to advanced designs under a common input-output framework, our platform enables researchers to directly compare power and sample size requirements across different design choices using the same underlying assumptions. The result is a freely accessible tool offering effective visualizations for the full spectrum of biomarker-guided trial designs, available at https://ampt.obicloud.ca/. Future improvements may further expand the tools capabilities to accommodate the increasing complexity of trial designs needed by the research community.

BackgroundArtificial intelligence (AI) has the potential to improve the efficiency of evidence synthesis and reduce human error. However, robust methods for evaluating rapidly evolving AI tools within the practical workflows of evidence synthesis remain underdeveloped. This protocol describes a study design for assessing the effectiveness, efficiency, and usability of AI tools in comparison to traditional human-only workflows in the context of Cochrane systematic reviews. MethodsMembers of the Cochrane Evaluation of (Semi-) Automated Review (CESAR) Methods Project developed an adaptive platform study-within-a-review (SWAR) design, modeled after clinical platform trials. This design employs a master protocol to concurrently evaluate multiple AI tools (interventions) against a standard human-only process (control) across three key review tasks: title and abstract screening, full-text screening, and data extraction. The adaptive framework allows for the addition or removal of AI tools based on interim performance analyses without necessitating a restart of the study. Performance will be assessed using metrics such as accuracy (sensitivity, specificity, precision), efficiency (time on task), response stability, impact of errors, and usability, in alignment with Responsible use of AI in evidence SynthEsis (RAISE) principles. ResultsThe study will generate comparative data about the performance and usability of specific AI tools employed in a semi- or fully automated manner relative to standard human effort. The protocol provides a flexible framework for the assessment of AI tools in evidence synthesis, addressing the limitations of static, one-time evaluations. DiscussionThis study protocol presents a novel methodological approach to addressing the challenges of evaluating AI tools for evidence syntheses. By validating entire workflows rather than individual technologies, the findings will establish an evidence base for determining the viability of integrating AI into evidence-synthesis workflows. The adaptive design of this study is flexible and can be adopted by other investigators, ensuring that the evaluation framework remains relevant as new tools emerge.

Long COVID affects millions of people worldwide, yet no disease-modifying treatment has been approved, and existing interventions have shown only modest and inconsistent benefits. A key reason for this limited progress is that current computational drug repurposing pipelines do not match well with the clinical reality of Long COVID. These patients often have persistent, multi-systemic symptoms and may already be taking multiple medications, making treatment safety a primary concern. However, most repurposing workflows still treat safety as a downstream filter and rely on disease-associated targets rather than causal drivers. They also assume that the findings of one analysis would generalize across the diverse presentations of Long COVID. We introduce SPLIT, a safety-first repurposing framework that addresses these limitations. SPLIT prioritizes safety at the start of the candidate evaluation, integrates complementary causal inference strategies to identify likely driver genes, and uses a counterfactual substitution design to compare drugs within specific cohort contexts. When applied to cognitive and respiratory Long COVID cohorts, SPLIT revealed three main findings. First, drugs with similar predicted efficacy could have very different predicted safety profiles. Second, the drugs flagged as unfavorable were often different between the two cohorts, showing that drug prioritization is phenotype-specific. Third, SPLIT flagged 18 drugs currently under active investigation in Long COVID trials as having unfavorable predicted profiles. SPLIT provides a practical framework to identify safer, more context-appropriate candidates earlier in the process, supporting more targeted and better-tolerated treatment strategies for Long COVID.

BackgroundSustained retention in care supports continuous access to antiretroviral therapy, routine clinical monitoring, and long-term viral suppression. ObjectiveTo compare the effectiveness of interventions for improving retention in care among people living with HIV (PLHIV). DesignSystematic review and network meta-analysis Data sourcesPubMed, Embase, CINAHL, PsycINFO, Web of Science, and the Cochrane Library from 1995 to December 2024. Eligibility criteriaRandomised controlled trials (RCTs) evaluating interventions to improve retention in care, viral load suppression, or quality of life (QoL) among PLHIV, compared with standard of care (SoC) or other interventions. Data extraction and synthesisPairs of reviewers independently screened studies, extracted data, and assessed risk of bias using ROBUST-RCT. We conducted a fixed-effect frequentist network meta-analysis and rated interventions categories relative to SoC based on effect estimates effects and the certainty of evidence.. Dichotomous outcomes were summarized as odds ratios (ORs) with 95% confidence intervals (CIs), and continuous outcomes as mean differences (MDs) with 95% CI. ResultsEighty-four trials enrolling 107 137 PLHIV evaluated 13 intervention categories. For retention in care, five interventions supported by moderate or high certainty evidence proved superior to SoC: multi-month dispensing (OR 2.02, 95% CI 1.32 to 3.09), task shifting (OR 1.94, 95% CI 1.42 to 2.66), differentiated service delivery (OR 1.47, 95% CI 1.22 to 1.76), behavioural counselling (OR 1.36, 95% CI 1.21 to 1.54), and supportive interventions (OR 1.31, 95% CI 1.11 to 1.55). For viral load suppression, two interventions supported by moderate or high certainty evidence proved superior to SoC: task shifting (OR 2.07, 95% CI 1.25 to 3.43) and behavioural counselling (OR 1.34, 95% CI 1.11 to 1.67). Across outcomes, no intervention demonstrated convincing superiority over other active interventions. ConclusionsAmong 13 intervention categories, only a subset provided moderate or high-certainty evidence of superiority to the standard of care, and no superiority to other interventions. Persistent evidence gaps for key populations, diverse settings, and long-term outcomes support the need for context-sensitive and patient-centred interventions. RegistrationPROSPERO CRD42024589177 Strengths and limitations of this study[tpltrtarr] This systematic review followed Cochrane methods and was reported in accordance with PRISMA-NMA guidelines. [tpltrtarr]The network meta-analysis integrated direct and indirect evidence to compare multiple intervention categories within a single framework. [tpltrtarr]Risk of bias and certainty of evidence were assessed using ROBUST-RCT and the GRADE approach for network meta-analysis, respectively. [tpltrtarr]Some networks were sparse, and limited representation of key populations and long-term follow-up constrained the strength and generalisability of inferences.

Background: Manual review of 30-day hospital readmissions can identify actionable quality and safety problems, but it is labor-intensive. We developed and evaluated an agentic AI workflow for evidence-grounded readmission review. Materials and methods: We studied adult patients with unplanned 30-day readmission after discharge from a medicine hospitalist service at a single academic health system. An AI agent using a large language model queried a database containing notes, encounters, procedures, laboratory results, and other clinical data, and completed the same structured readmission-review rubric used by physicians. In the primary comparative evaluation, 20 randomly selected readmissions from 2025 were each reviewed by two physicians and the AI system. Blinded physician evaluators rated review quality. After rubric refinement, the AI workflow was applied to 100 recent readmissions in an exploratory expanded-cohort analysis of recurring improvement opportunities. Results: In the primary comparative evaluation, the AI classified 9/20 readmissions (45%) as preventable, compared with 19/40 physician reviews (47.5%). Blinded overall quality ratings were similar for AI and physician reviews (4.35 vs. 4.20 on a 1-5 scale; mean difference 0.15, 95% CI -0.20 to 0.48; p=0.49), as were factuality/support and usefulness/actionability ratings. No AI hallucinations were identified during factuality review. Agreement on preventability and primary readmission category was low for both AI-human and human-human comparisons. The AI system cost $0.23 per chart; physician reviewers took a median of 15 minutes, corresponding to an estimated $42.43 per chart. In the exploratory expanded-cohort analysis, AI-assisted review identified recurring vulnerabilities in post-discharge follow-up plans, incomplete inpatient workups, medication-safety transitions, and indwelling-device transitions. Conclusions: Agentic AI produced readmission reviews with similar blinded quality ratings to physician reviews in this small single-center primary comparative evaluation and supported identification of recurring quality-improvement themes in the exploratory expanded-cohort analysis. Preventability judgments remained variable among both AI and physicians, underscoring the need for human oversight and prospective evaluation before operational use.

Randomized controlled trials (RCTs) play a central role in assessing the benefits and harms of interventions. Incomplete reporting in RCT publications can compromise the verifiability and usefulness of RCTs. SPIRIT and CONSORT reporting guidelines aim to improve the completeness of RCT protocols and results publications, respectively. However, many RCTs are not reported completely. Checking manuscripts automatically could help authors improve the completeness of reports prior to publication. We previously annotated SPIRIT-CONSORT-TM, a corpus of 200 articles (comprising 100 protocol-results publication pairs) using 83 checklist items drawn from SPIRIT 2013 and CONSORT 2010. We also trained machine learning models to automatically assess reporting at the item level. Each checklist item can include multiple constituent elements (i.e., specific details required for that item), and an item might be considered fully reported when all of its elements are present. However, prior work does not explicitly capture or evaluate reporting at the element level. To address this gap, we extended SPIRIT-CONSORT-TM by incorporating element-level annotations and using them to assess reporting completeness (SPIRIT-CONSORT-ELM). We formulated element-level assessment as a machine reading comprehension task, operationalized through 119 questions, where each question targets a specific reporting element within a checklist item. Using the 200 articles included in SPIRIT-CONSORT-TM, two annotators independently answered 119 questions for 50 articles (25 protocol-results pairs) and resolved any discrepancies through discussion; the remaining 150 articles (75 protocol-results pairs) were assessed by a single annotator. We then developed an automated pipeline for element-level assessment using SPIRIT-CONSORT-ELM. The pipeline first applies a PubMedBERT-based model to identify sentences containing item-level reporting information, then it uses a generative large language model (LLM; GPT-5) with chain-of-thought reasoning to answer element-level questions based on the retrieved evidence. Agreement between the two annotators was high (Gwet's AC1: 0.782) and our pipeline achieved high accuracy in identifying element-level reporting evidence (F1: 0.822, Gwet's AC1: 0.796). Ablation studies indicate that chain-of-thought reasoning and the inclusion of illustrative in-context examples modestly improve LLM performance on the machine reading comprehension task. SPIRIT-CONSORT-ELM provides a benchmark for evaluating reporting guideline completeness at the element level, enabling assessment of RCT transparency beyond the simple presence or absence of checklist items and is publicly available at https://osf.io/kznx4/. The automated pipeline establishes a robust baseline for assessing RCT reporting and demonstrates potential as a practical aid for authors, reviewers, and editors to identify and address gaps in completeness and transparency of RCT reports.

Clinical trials often enrol populations that differ from those who ultimately receive the interventions, raising concerns about external validity and health equity. Trial registries could provide an early opportunity to assess representativeness, but it is unclear whether registry data contain sufficient information to enable such assessments. This study evaluated the feasibility of using registry data to assess representativeness in Phase II and III pharmacological randomised controlled trials. A search of ClinicalTrials.gov from December 2024 to January 2025 identified trials with results posted after 1 January 2023 across cardiovascular disease (CVD) excluding stroke, diabetes mellitus, and selected mental health disorders. Of 1,328 records screened, 98 trials met inclusion criteria (51 Phase III, 47 Phase II). Reporting completeness was variable, particularly in Phase II studies. CVD and diabetes trials predominantly included middle-aged to older adults, while mental health trials recruited mainly individuals aged 36 to 50 years. Across CVD and mental health trials, participants were largely male. Reporting of BMI, contraception, and comorbidity criteria was inconsistent, though available data suggested these factors influenced sample composition. Fewer than 10% of trials reported equity-relevant characteristics beyond age and sex, and none addressed intersectionality. Assessing equity using registry data is feasible but constrained by incomplete and inconsistent reporting.

Systematic Reviews (SRs) are the gold standard for evidence synthesis, but the manual title and abstract screening of thousands of references creates a severe bottleneck. Existing automated tools have historically struggled to achieve the near-perfect recall (sensitivity) required for reliable reviews. We developed MechaScreener as a "zero-shot" automated screening tool that utilises a Large Language Model (LLM) to rank article relevance. The tool requires no initial training data or manual pre-screening, as MechaScreener directly applies user-provided question elements (PICO) or inclusion/exclusion criteria to assign an inclusion probability score (1-5) to each reference. We evaluated the tool in two phases: a development phase using five reference libraries to optimise prompts, and an independent evaluation phase using 10 diverse Cochrane review libraries (comprising both randomised controlled trials and non-RCTs) containing over 58,000 references. In the evaluation dataset, MechaScreener achieved a perfect mean recall of 1.00 (100%, pooled 95% CI: 0.98-1.00), ensuring no relevant articles were missed. Concurrently, it achieved an overall mean specificity of 0.61 (61%, pooled 95% CI: 0.59-0.60). Specificity varied: from 0.21 in broad public health topics to 0.91 in precise pharmacological interventions-reflecting the tools built-in conservatism when evaluating ambiguous abstracts. By safely eliminating over 60% of irrelevant literature during the initial screening phase without compromising recall, MechaScreener functions as a highly reliable but low-effort "first-pass" filter, allowing researchers to substantially reduce manual workloads and reallocate resources toward full-text review and data extraction.

Title and abstract screening limit the timeliness of systematic reviews used for clinical guidelines. We evaluated audited large language model (LLM) triage at Sweden's National Board of Health and Welfare. Ten LLMs from five model families were tested on 419 Cochrane reviews comprising 26,892 records, and the selected ensemble was externally validated on 133 reviews including 8,501 records matched to planned guideline topics. The same locked model pair was then used prospectively across 24 systematic reviews in two national guideline programmes. On the 419-review selection benchmark, the selected Gemini-3-flash plus GPT-5.1 ensemble achieved 98.0% (95% CI, 97.3-98.7) mean review-level sensitivity, while topic-matched validation yielded 96.7% sensitivity (95% CI, 93.7-98.9). Prospective deployment screened 74,679 records, placed 63,858 (85.5%) in the AI-excluded pool and reduced estimated first-pass screening effort from 415 to 34 person-days. Across 600 randomly sampled AI-excluded records from the migraine and dementia programmes, none was confirmed as a final false negative after post-unblinding adjudication; across the completed 680-record audit, all 38 final retained records had been AI flagged, whereas locked blinded human consensus missed seven. These findings support locked, audited LLM triage, with human oversight and programme-specific monitoring, for systematic reviews used in national guidelines.