Systematic reviews (SRs) serve as a core methodology in evidence-based medicine (EBM), providing critical evidence for clinical practice and health decision-making. However, the manual screening of titles and abstracts in SRs is labor-intensive and time-consuming, becoming a major bottleneck in research efficiency. Recent advancements in artificial intelligence (AI), particularly large language models (LLMs), have introduced new opportunities and transformations in this field. This article provided an overview of the current status of intelligent screening for titles and abstracts in systematic reviews, with a focus on the application and effectiveness of LLMs. It aims to provide recommendations for users and developers, facilitating the better integration of automation algorithms into the SR process.
Systematic reviews can provide important evidence support for clinical practice and health decision-making. In this process, literature screening and data extraction are extensively time-consuming procedures. Natural language processing (NLP), as one of the research directions of computer science and artificial intelligence, can accelerate the process of literature screening and data extraction in systematic reviews. This paper introduced the requirements of systematic reviews for rapid literature screening and data extraction, the development of NLP and types of machine learning; and systematically collated the NLP tools for the title and abstract screening, full-text screening and data extraction in systematic reviews; and discussed the problems in the application of NLP tools in the field of systematic reviews and proposed a prospect for its future development.
ObjectiveTo construct the comprehensive evaluation indicator system of Chinese patent medicines for ischemic stroke, to determine the weight of indicators, and to provide references for the comprehensive evaluation of the efficacy, safety, and economy of Chinese patent medicines.MethodsTwo rounds of expert consultation by Delphi method were applied to establish the comprehensive evaluation indicator system of Chinese patent medicines for ischemic stroke, and the weight of each indicator was determined by the analytic hierarchy process method.ResultsQuestionnaire recovery rates of 2 rounds were 92.59% and 96.00%, the expert authority coefficient was greater than 0.7, and the coordination coefficients of experts in the total index were 0.224 and 0.370 (P<0.001). A three-level comprehensive evaluation indicator system for Chinese patent medicines for ischemic stroke was established and the three first-level indicators included efficacy, safety, and economy. And there were 15 second-level indicators, and 33 third-level indicators. Through the analytic hierarchy process method, the weights of each first-level indicator were 0.626 4, 0.301 2, and 0.072 4, respectively.ConclusionThe comprehensive evaluation indicator system contains efficacy, safety and economy, and provides a basis for a comprehensive evaluation of Chinese patent medicines for ischemic stroke. The indicator system is of great significance for the design of outcomes for clinical trials of ischemic stroke, the conduction of systematic reviews, and the development of clinical practice guidelines for ischemic stroke patients when selecting study outcomes.
Traditional Chinese medicine (TCM) is a treasure of the Chinese nation. Presence of clinical effects represents a fundamental issue for TCM development. Nevertheless, the complexities of TCM interventions often result in presented effects deviating from expected ones, a phenomenon so called as "effect off-target"; this issue has become a major challenge for the development and use of TCM interventions. In continuing efforts, we have proposed an innovative evidence-based medicine model for studying the effects of TCM interventions, termed "systems evidence-based medicine (sysEBM)". Essentially, the sysEBM model integrates clinical and non-clinical evaluation to develop a systematic pathway for studying effects of TCM interventions, and the methodological steps typically include the development of PICO framework for a putative effect, exploration of the effect and confirmation of the effect by using animal models, observational studies and clinical trials. As an additional step, multidisciplinary technologies including pharmaceutical, pharmacological, information and biological technologies will be used to provide multidimensional analyses of potential action networks and mechanisms of TCM interventions. Building on this concept, we have developed a sysEBM model ("6R" model) for acupuncture and marketed Chinese patent medicines by integrating real-world evidence, clinical trials, evidence syntheses, and rapid recommendation methodologies, as well as information technology and biomedical technologies. We also applied this model for developing TCM interventions for maternal health, critical care, and knee osteoarthritis.