Explore the vast range of titles available.

Randomized controlled trials (RCTs) face challenges such as limited generalizability, insufficient recruitment diversity, and high failure rates, often due to restrictive eligibility criteria and inefficient patient selection. Large language models (LLMs) have shown promise in various clinical tasks, but their potential role in RCT design remains underexplored. This study investigates the ability of LLMs, specifically GPT-4-Turbo-Preview, to assist in designing RCTs that enhance generalizability, recruitment diversity, and reduce failure rates, while maintaining clinical safety and ethical standards. We conducted a noninterventional, observational study analyzing 20 parallel-arm RCTs, comprising 10 completed and 10 registered studies published after January 2024 to mitigate pretraining biases. The LLM was tasked with generating RCT designs based on input criteria, including eligibility, recruitment strategies, interventions, and outcomes. The accuracy of LLM-generated designs was quantitatively assessed by 2 independent clinical experts by comparing them to clinically validated ground truth data from ClinicalTrials.gov. We have conducted statistical analysis using natural language processing-based methods, including Bilingual Evaluation Understudy (BLEU), Recall-Oriented Understudy for Gisting Evaluation (ROUGE)-L, and Metric for Evaluation of Translation with Explicit ORdering (METEOR), for objective scoring on corresponding LLM outputs. Qualitative assessments were performed using Likert scale ratings (1-3) for domains such as safety, clinical accuracy, objectivity or bias, pragmatism, inclusivity, and diversity. The LLM achieved an overall accuracy of 72% in replicating RCT designs. Recruitment and intervention designs demonstrated high agreement with the ground truth, achieving 88% and 93% accuracy, respectively. However, LLMs showed lower accuracy in designing eligibility criteria (55%) and outcomes measurement (53%). Natural language processing statistical analysis reported BLEU=0.04, ROUGE-L=0.20, and METEOR=0.18 on average objective scoring of LLM outputs. Qualitative evaluations showed that LLM-generated designs scored above 2 points and closely matched the original designs in scores across all domains, indicating strong clinical alignment. Specifically, both original and LLM-based designs ranked similarly high in safety, clinical accuracy, and objectivity or bias in published RCTs. Moreover, LLM-based design ranked noninferior to original designs in registered RCTs in multiple domains. In particular, LLMs enhanced diversity and pragmatism, which are key factors in improving RCT generalizability and addressing failure rates. LLMs, such as GPT-4-Turbo-Preview, have demonstrated potential in improving RCT design, particularly in recruitment and intervention planning, while enhancing generalizability and addressing diversity. However, expert oversight and regulatory measures are essential to ensure patient safety and ethical standards. The findings support further integration of LLMs into clinical trial design, although continued refinement is necessary to address limitations in eligibility and outcomes measurement.

Although trials with anti-seizure drugs have not shown anti-epileptogenic or disease-modifying activity in humans, new compounds are on the horizon that may require novel trial designs. We briefly discuss the unique challenges and the available options to identify innovative clinical trial designs that differentiate novel anti-epileptogenic and disease-modifying compounds, preferably early in phase II, from current anti-seizure drugs. The most important challenges of clinical testing of agents for epilepsy prevention include having sufficient preclinical evidence for a suitable agent to proceed with a human trial of an anti-epileptogenic drug, and to demonstrate the feasibility of doing such a trial. Major challenges in trial design to assess agents for disease modification include the choice of suitable study parameters, the identification of a high-risk study population, the type of control, the time and duration of treatment, and a feasible follow-up period.

Abstract Inflammatory bowel diseases (IBD) are highly heterogeneous conditions, varying in clinical manifestations, disease localization, progression, and response to treatment. Failing to account for this heterogeneity can substantially diminish the power of clinical trials and reduce the likelihood of detecting a true effect. In this review, we explore the transformative potential of Bayesian statistics in IBD clinical research, highlighting its ability to provide deeper insights, refine trial design, and facilitate more informed medical decision-making. We explain how Bayesian methods are best incorporated into innovative IBD clinical trial designs, such as single-arm trials utilizing historical data, master protocols, and adaptive trials. In adaptive designs, Bayesian techniques enable dynamic adjustments to sample sizes based on interim data, helping to maintain adequate power while optimizing resource allocation. For network meta-analysis, Bayesian statistics enhance the estimation of treatment effects in complex or sparse data situations by integrating prior knowledge and effectively managing hierarchical models. These methods are also applied in pharmacokinetic decision-making to address inter-patient variability in IBD, offering more accurate predictions of drug concentrations and target attainment at the outset of treatment. A checklist is added for non-specialist readers on how to approach reading an article that employs Bayesian methods, as part of a Users’ Guide to the Literature. Lay Summary This review highlights the potential of Bayesian statistics to improve IBD clinical research by refining trial designs and enabling more meaningful inferences. By using some examples, we explain how prior knowledge, such as historical or placebo data, is incorporated into Bayesian analysis to inform real-time decisions and increase the likelihood of detecting true treatment effects.

A practical impediment in adaptive clinical trials is that outcomes must be observed soon enough to apply decision rules to choose treatments for new patients. For example, if outcomes take up to six weeks to evaluate and the accrual rate is one patient per week, on average three new patients will be accrued while waiting to evaluate the outcomes of the previous three patients. The question is how to treat the new patients. This logistical problem persists throughout the trial. Various ad hoc practical solutions are used, none entirely satisfactory. We focus on this problem in phase I–II clinical trials that use binary toxicity and efficacy, defined in terms of event times, to choose doses adaptively for successive cohorts. We propose a general approach to this problem that treats late-onset outcomes as missing data, uses data augmentation to impute missing outcomes from posterior predictive distributions computed from partial follow-up times and complete outcome data, and applies the design’s decision rules using the completed data. We illustrate the method with two cancer trials conducted using a phase I–II design based on efficacy–toxicity trade-offs, including a computer stimulation study. Supplementary materials for this article are available online.

Background Awareness of model-based designs for dose-finding studies such as the Continual Reassessment Method (CRM) is now becoming more commonplace amongst clinicians, statisticians and trial management staff. In some settings toxicities can occur a long time after treatment has finished, resulting in extremely long, interrupted, CRM design trials. The Time-to-Event CRM (TiTE-CRM), a modification to the original CRM, accounts for the timing of late-onset toxicities and results in shorter trial duration. In this article, we discuss how to design and deliver a trial using this method, from the grant application stage through to dissemination, using two radiotherapy trials as examples. Methods The TiTE-CRM encapsulates the dose-toxicity relationship with a statistical model. The model incorporates observed toxicities and uses a weight to account for the proportion of completed follow-up of participants without toxicity. This model uses all available data to determine the next participant’s dose and subsequently declare the maximum tolerated dose. We focus on two trials designed by the authors to illustrate practical issues when designing, setting up, and running such studies. Results In setting up a TiTE-CRM trial, model parameters need to be defined and the time element involved might cause complications, therefore looking at operating characteristics through simulations is essential. At the grant application stage, we suggest resources to fund statisticians’ time before funding is awarded and make recommendations for the level of detail to include in funding applications. While running the trial, close contact of all involved staff is required as a dose decision is made each time a participant is recruited. We suggest ways of capturing data in a timely manner and give example code in R for design and delivery of the trial. Finally, we touch upon dissemination issues while the trial is running and upon completion. Conclusion Model-based designs can be complex. We hope this paper will help clinical trial teams to demystify the conduct of TiTE-CRM trials and be a starting point for using this methodology in practice.

Osteoporosis has a significant economy, society, and health burden. The recent advancement of available therapies caused ethical questions regarding the use of placebo-controlled studies in osteoporosis. Some specialists believe it is not ethically to subject the study participants to any additional risk when there is already verified effective therapy for the disease and established therapies can significantly decrease the likelihood of osteoporotic fracture.  Accordingly, researchers have expressed ethical concerns over placebo-controlled trials. Here, we have briefly addressed ethical and methodological aspects regarding conducting placebo-control trials as well as potential alternatives.

Machine learning (ML) and artificial intelligence (AI) applications have increased across different stages of clinical research. Their use in clinical trials (CTs) has been discussed but not quantified. A scoping review was conducted by searching PubMed, Embase (Ovid), and Scopus for CTs or protocols. The goal was to understand the extent of ML and AI applications in the design, conduct, and analysis of CTs. Screening was performed on Covidence, with GPT model support. After title/abstract and full-text screening, 108 records were included; in some studies, AI/ML was applied across multiple stages. For the design, 20 studies involved advanced methods, six applied them to stratification, four to treatment selection during randomization, six to participant selection, two for outcome assessment, and two for site selection. Seven studies involved them in the collection and analysis of data from wearable devices, and one for monitoring. More commonly, AI/ML has been used at the analysis stage of 93 CTs; however, limitations in reporting trial objectives make it difficult to distinguish the purpose between primary and exploratory analyses. This research identifies a serious mismatch between the potential and actual applications of ML in CTs. Considering the potential benefits of ML in CTs, such underuse could hinder the evolution of CTs toward faster and more efficient approaches.

Biomarker-guided treatment is a rapidly developing area of medicine, where treatment choice is personalised according to one or more of an individual’s biomarker measurements. A number of biomarker-guided trial designs have been proposed in the past decade, including both adaptive and non-adaptive trial designs which test the effectiveness of a biomarker-guided approach to treatment with the aim of improving patient health. A better understanding of them is needed as challenges occur both in terms of trial design and analysis. We have undertaken a comprehensive literature review based on an in-depth search strategy with a view to providing the research community with clarity in definition, methodology and terminology of the various biomarker-guided trial designs (both adaptive and non-adaptive designs) from a total of 211 included papers. In the present paper, we focus on non-adaptive biomarker-guided trial designs for which we have identified five distinct main types mentioned in 100 papers. We have graphically displayed each non-adaptive trial design and provided an in-depth overview of their key characteristics. Substantial variability has been observed in terms of how trial designs are described and particularly in the terminology used by different authors. Our comprehensive review provides guidance for those designing biomarker-guided trials.

To propose a tool to assist trialists in making design decisions that are consistent with their trial's stated purpose. Randomized trials have been broadly categorized as either having a pragmatic or explanatory attitude. Pragmatic trials seek to answer the question, “Does this intervention work under usual conditions?,” whereas explanatory trials are focused on the question, “Can this intervention work under ideal conditions?” Design decisions make a trial more (or less) pragmatic or explanatory, but no tool currently exists to help researchers make the best decisions possible in accordance with their trial's primary goal. During the course of two international meetings, participants with experience in clinical care, research commissioning, health care financing, trial methodology, and reporting defined and refined aspects of trial design that distinguish pragmatic attitudes from explanatory. We have developed a tool (called PRECIS) with 10 key domains and which identifies criteria to help researchers determine how pragmatic or explanatory their trial is. The assessment is summarized graphically. We believe that PRECIS is a useful first step toward a tool that can help trialists to ensure that their design decisions are consistent with the stated purpose of the trial.

This is the introductory paper in a series of eight papers. In this series, we integrate the theoretical design options with the practice of conducting pragmatic trials. For most new market-approved treatments, the clinical evidence is insufficient to fully guide physicians and policy makers in choosing the optimal treatment for their patients. Pragmatic trials can fill this gap, by providing evidence on the relative effectiveness of a treatment strategy in routine clinical practice, already in an early phase of development, while maintaining the strength of randomized controlled trials. Selecting the setting, study population, mode of intervention, comparator, and outcome are crucial in designing pragmatic trials. In combination with monitoring and data collection that does not change routine care, this will enable appropriate generalization to the target patient group in clinical practice. To benefit from the full potential of pragmatic trials, there is a need for guidance and tools in designing these studies while ensuring operational feasibility. This paper introduces the concept of pragmatic trial design. The complex interplay between pragmatic design options, feasibility, stakeholder acceptability, validity, precision, and generalizability will be clarified. In this way, balanced design choices can be made in pragmatic trials with an optimal chance of success in practice.