Network meta-analysis aims to integrate direct and indirect evidence, make a comprehensive comparison and in-depth analysis of three or more interventions and treatments, compare and rank the advantages and disadvantages of different treatment measures, so as to provide strong evidence for decision-making. However, there may be some bias in the process of making network meta-analysis, analyzing data and interpreting results. Therefore, accurate assessment and proper handling of the risks of bias in network meta-analysis is conductive to improve the quality of decision-making and promote the achievement of good clinical outcomes. At present, the number of published network meta-analysis has increased significantly globally, but the quality remains to be improved, and there is a lack of assessment tools for risks of bias in network meta-analysis. In 2025, Canadian scholar Carole Lunny and colleagues developed The Risk of Bias in Network Meta-Analysis (RoB NMA) tool for evaluating the risk of bias in network meta-analysis and published it in the BMJ, which is important to reduce the bias in network meta-analysis and promote optimal clinical decision-making. This study will interpret it with examples, aiming to help researchers better understand and apply this evaluation tool.
ObjectiveTo analyze the research hotspots and development trends of core outcome set (COS) from 2015 to 2024, providing a reference for future research in this field. MethodsWe retrieved literature on COS research from the Web of Science Core Collection and CNKI spanning January 1, 2015 to December 31, 2024. We extracted and organized data on the number of publications, journals, citation frequency, and keywords using Excel 2021. We performed keyword clustering analysis using VOSviewer 1.6.13 and generated strategic coordinate maps using Bibliometrix 3.13 in R 4.3.1. ResultsWe included a total of 1 288 studies, comprising 1 085 English publications and 203 Chinese publications. From 2015 to 2024, the number of COS publications showed a steady increase. English journals covered a wide range of fields, while Chinese journals were mainly focused on traditional Chinese medicine. High-impact articles primarily focused on COS methodology. Chinese literature mainly concentrated on the application of COS in traditional Chinese medicine, while English literature focused on child health, Delphi surveys, quality of life, and pain. The results of the strategic coordinate map showed that research on acupuncture core outcome indicators, qualitative studies of surgical COS, and Delphi-based COS for quality of life in patients with rheumatoid diseases were relatively weak, with significant room for improvement. ConclusionOver the past decade, COS research has shown a steady growth trend and has gradually become an important tool for improving the standardization and scientific rigor of clinical research. As COS research continues to expand, there is increasing overlap in the scope and findings of different studies. Future research could incorporate umbrella and basket study designs to optimize resource utilization and promote the application of COS in clinical practice.
ObjectiveThis study investigates the adherence to ethical principles in doctoral dissertations focused on human as the research subject, aiming to provide a foundation for enhancing ethical awareness among medical doctoral candidates. MethodsUtilizing the Chinese database of doctoral dissertations, a total of 1 733 relevant papers published in 2021 were collected. The study compared ethical considerations among double first-class universities, other high-ranking institutions, different university types, various disciplines, diverse training orientations, enrollment types, and medical doctoral dissertations from different regions. ResultsIn 2021, among Chinese medical doctoral dissertations involving human as the research subject, 73.34% mentioned ethical considerations, and 86.27% mentioned informed consent. Dissertations reporting ethical approval descriptions, approval numbers, ethical approvals, and informed consent constituted only 2.19%. Notably, 12.52% of medical doctoral dissertations failed to incorporate ethical considerations and informed consent details in their content. ConclusionThe ethical awareness of medical doctoral candidates in China and the reporting of ethical information in their dissertations require urgent enhancement and improvement.
Studies on chatbot health advice (CHA) driven by large language models are rapidly increasing, yet their reporting is marked by significant heterogeneity and incompleteness, which severely limits the scientific credibility and reproducibility of their findings. To promote the effective dissemination and application of the newly released chatbot assessment reporting tool (CHART) statement, this paper provides a systematic interpretation and example-based analysis of the guideline. This paper dissects the 12 main items and 39 sub-items of the CHART checklist on an item-by-item basis, systematically elaborating on the methodological rationale behind each reporting requirement. A particular focus is placed on key requirements tailored to the unique characteristics of generative AI, such as the transparent disclosure of prompt engineering, query strategies, and dialogue safety. To bridge the gap between theory and practice, a high-quality, published CHA study is used as an exemplar to demonstrate the practical application of each reporting item. This interpretation report aims to provide a clear and practical handbook for researchers, journal reviewers, and editors, with the goal of fostering standardized, high-quality development in the field of CHA research and promoting the safe and effective application of AI in healthcare.
Living systematic reviews (LSR) represent an evolving methodology for systematic review that is continuously updated to incorporate new evidence in a timely manner, ensuring that healthcare professionals and policymaker shave access to the most last information to make optimal decisions. The global publication of LSR has been a rapid increase. But the quality of reporting remains to be enhanced. In 2024, the PRISMA-LSR working group, in conjunction with the characteristics of LSR to form the reporting standards for living systematic reviews, which plays a significant role in promoting the transparent, complete, and accurate reporting of LSR. It has been published in the BMJ journal. This article interpreted PRISMA-LSR with representative examples, aiming to provide a reference for the standardization of LSR by domestic scholars.
Mendelian randomization (MR) studies use genetic variants as instrumental variables to explore the effects of exposures on health outcomes. STROBE-MR (strengthening the reporting of observational studies in epidemiology using Mendelian randomization) assists authors in reporting their MR studies clearly and transparently, and helpfully to improve the quality of MR. This paper interpreted the STROBE-MR, aiming to help Chinese scholars better understand, disseminate, and apply it.
Systematic reviews and meta-analyses are essential methods in evidence-based medicine for integrating research evidence and guiding clinical decision-making. However, with the rapid expansion of medical research data, traditional approaches face significant challenges in terms of efficiency, accuracy, and reliability. In recent years, the rapid advancement of artificial intelligence (AI) technologies, particularly in natural language processing (NLP), machine learning (ML), and large language models (LLMs), has provided robust support for automating and intelligentizing systematic reviews and meta-analyses. This paper systematically reviews the progress of AI applications in these fields, tracing the evolution from traditional tools to intelligent platforms, and analyzes the functional characteristics, application scenarios, and limitations of existing AI-driven tools. Furthermore, it explores the challenges posed by AI in terms of adaptation to the medical field, multimodal data processing, and ethical transparency, while offering potential solutions and optimization strategies. Looking ahead, with the continuous optimization of technology, enhanced data sharing, and the establishment of industry standards, AI is expected to significantly improve the efficiency and quality of systematic reviews and meta-analyses, driving the transition from "tool-driven" to "intelligent collaboration." The deep integration of AI not only injects innovative momentum into evidence-based medicine but also reshapes its methodological foundation, laying a solid basis for a more intelligent, equitable, and efficient future.
The burgeoning application of large language models (LLM) in healthcare demonstrates immense potential, yet simultaneously poses new challenges to the standardization of research reporting. To enhance the transparency and reliability of medical LLM research, an international expert group published the TRIPOD-LLM reporting guideline in Nature Medicine in January 2024. As an extension of the TRIPOD+AI guideline, TRIPOD-LLM provides detailed reporting items specifically tailored to the unique characteristics of LLMs, including general foundational models (e.g., GPT-4) and domain-specific fine-tuned models (e.g., Med-PaLM 2). It addresses critical aspects such as prompt engineering, inference parameters, generative evaluation, and fairness considerations. Notably, the guideline introduces an innovative modular design and a "living guideline" mechanism. This paper provides a systematic, item-by-item interpretation and example-based analysis of the TRIPOD-LLM guideline. It is intended to serve as a clear and practical handbook for researchers in this field, as well as for journal reviewers and editors responsible for assessing the quality of such studies, thereby fostering the high-quality development of medical LLM research in China.
Accurately assessing the risk of bias is a critical challenge in network meta-analysis (NMA). By integrating direct and indirect evidence, NMA enables the comparison of multiple interventions, but its outcomes are often influenced by bias risks, particularly the propagation of bias within complex evidence networks. This paper systematically reviews commonly used bias risk assessment tools in NMA, highlighting their applications, limitations, and challenges across interventional trials, observational studies, diagnostic tests, and animal experiments. Addressing the issues of tool misapplication, mixed usage, and the lack of comprehensive tools for overall bias assessment in NMA, we propose strategies such as simplifying tool operation, enhancing usability, and standardizing evaluation processes. Furthermore, advancements in artificial intelligence (AI) and large language models (LLMs) offer promising opportunities to streamline bias risk assessments and reduce human interference. The development of specialized tools and the integration of intelligent technologies will enhance the rigor and reliability of NMA studies, providing robust evidence to support medical research and clinical decision-making.
Systematic reviews and meta-analyses have become the cornerstone methodologies for integrating multi-source research data and enhancing the quality of evidence. Traditional meta-analyses often demonstrate limitations when handling multiple treatment options. Network meta-analysis (NMA) overcomes these limitations by constructing a network of evidence that encompasses various treatment options, allowing for the simultaneous comparison of both direct and indirect evidence across multiple treatment plans. This provides more comprehensive and precise support for clinical decision-making. This article comprehensively reviews the statistical principles of NMA, its three fundamental assumptions, and the statistical inference framework. It also critically analyzes the mainstream NMA software and packages currently available, such as R (including gemtc, netmeta, rjags, pcnetmeta), Stata (mvmeta, network), WinBUGS, SAS, ADDIS, and various online applications, highlighting their strengths, weaknesses, and suitable scenarios. This analysis provides researchers with a scientific and unified framework for conducting clinical studies and policy-making.
