25 June 2024 in Journal of Medical Internet Research
Large language models (LLMs) such as ChatGPT have become widely applied in the field of medical research. In the process of conducting systematic reviews, similar tools can be used to expedite various steps, including defining clinical questions, performing the literature search, document screening, information extraction, and language refinement, thereby conserving resources and enhancing efficiency. However, when using LLMs, attention should be paid to transparent reporting, distinguishing between genuine and false content, and avoiding academic misconduct. In this viewpoint, we highlight the potential roles of LLMs in the creation of systematic reviews and meta-analyses, elucidating their advantages, limitations, and future research directions, aiming to provide insights and guidance for authors planning systematic reviews and meta-analyses.
1 August 2023 in Psychiatry Research
Sertraline has been shown to be effective in the treatment of depression. However, the relationship between the dosage of sertraline and its efficacy and safety are unclear. We identified RCTs that compared sertraline with placebo for the treatment of depression, conducted conventional meta-analyses on the efficacy and safety of sertraline, and assessed the nonlinear dose-response relationship between sertraline dosage and the changes in HAM-D and CGI-S scores, dropout from care for any reason or due to adverse effects, and the rate of adverse effects, using a 1-stage restricted cubic spline regression model. Twenty-one RCTs involving 4,235 patients were included. The pooled mean differences (MD) in the change in HAM-D total score [MD=-2.34, 95% confidence interval (CI) -2.93, -1.76], CGI-S score and MADRS score, but also the dropout rate for adverse effects, and rate of adverse effects were higher in sertraline group. The therapeutic response of sertraline for treating depression increased with the dosage. Meanwhile, the risk of total adverse reactions slightly decreased between 50 and 150 mg, and increased at doses above 150 mg. The dose-dependence of both efficacy and safety need to be considered when choosing the optimal dosage of sertraline.
14 August 2025 in JMIR Research Protocols
Background: The burden of multimorbidity is increasing globally, which complicates the use of guidelines in clinical practice and health care: practitioners may need to increasingly refer to multiple guidelines with potentially conflicting recommendations. Objective: We aim to develop a guideline-based decision support framework for the management of patients with multimorbidity to help clinicians efficiently evaluate, select, and adapt recommendations focusing on the different comorbidities and aspects of multimorbidity. Methods: We will conduct the project using the following steps: (1) needs assessment (searching published literature and documents on guideline use in multimorbidity care through the study initiators, and assessing the necessity of developing a comprehensive decision-making framework focusing on multimorbidity in a broad sense), (2) establishing international working groups (a coordination team, an evidence support group, and a consensus group) by leveraging existing participants’ networks and inviting experts with relevant academic publications or activities, (3) conducting literature reviews of multimorbidity guidelines and original qualitative research involving interest-holders in multimorbidity care and/or guideline development to formulate an initial draft framework, (4) a consensus process including an expert survey and a consensus meeting, (5) formulating and releasing the final framework, and (6) testing the framework (collecting feedback through educating health professionals in different settings and applying the framework in practice to evaluate and improve it). We plan to complete the project within 3 years. Results: The project has started in March 2024 and is due to conclude in June 2026. As of May 2025, we have finished the literature reviews and qualitative studies and are currently conducting the first round of the expert survey. Conclusions: This framework will help clinicians from all levels of health care institutions to make decisions in the management of patients with multimorbidity based on the latest available evidence, and to reduce potential health risks to their patients. One limitation of this framework is that such a broad framework may not fully fit all disease combinations or realistic situations. To reduce the degree of inapplicability, after completion of the framework, we will continue to monitor its use with regular updates as needed. International Registered Report Identifier (IRRID): DERR1-10.2196/63339
20 February 2024 in JMIR Preprints
UNSTRUCTURED Large language models (LLMs) such as ChatGPT have become widely applied in the field of medical research. In the process of conducting systematic reviews, similar tools can be used to expedite various steps, including defining clinical questions, performing the literature search, document screening, information extraction, and language refinement, thereby conserving resources and enhancing efficiency. However, when using LLMs, attention should be paid to transparent reporting, distinguishing between genuine and false content, and avoiding academic misconduct. In this viewpoint, we highlight the potential roles of LLMs in the creation of systematic reviews and meta-analyses, elucidating their advantages, limitations, and future research directions, aiming to provide insights and guidance for authors planning systematic reviews and meta-analyses.
17 June 2024 in JMIR Preprints
BACKGROUND The burden of multimorbidity is increasing globally, which complicates the use of guidelines in clinical practice and health care: practitioners may need to increasingly refer to multiple guidelines with potentially conflicting recommendations. OBJECTIVE We aim to develop a guideline-based decision support framework for the management of patients with multimorbidity to help clinicians efficiently evaluate, select, and adapt recommendations focusing on the different comorbidities and aspects of multimorbidity. METHODS We will conduct the project using the following steps: (1) needs assessment (searching published literature and documents on guideline use in multimorbidity care through the study initiators, and assessing the necessity of developing a comprehensive decision-making framework focusing on multimorbidity in a broad sense), (2) establishing international working groups (a coordination team, an evidence support group, and a consensus group) by leveraging existing participants’ networks and inviting experts with relevant academic publications or activities, (3) conducting literature reviews of multimorbidity guidelines and original qualitative research involving interest-holders in multimorbidity care and/or guideline development to formulate an initial draft framework, (4) a consensus process including an expert survey and a consensus meeting, (5) formulating and releasing the final framework, and (6) testing the framework (collecting feedback through educating health professionals in different settings and applying the framework in practice to evaluate and improve it). We plan to complete the project within 3 years. RESULTS The project has started in March 2024 and is due to conclude in June 2026. As of May 2025, we have finished the literature reviews and qualitative studies and are currently conducting the first round of the expert survey. CONCLUSIONS This framework will help clinicians from all levels of health care institutions to make decisions in the management of patients with multimorbidity based on the latest available evidence, and to reduce potential health risks to their patients. One limitation of this framework is that such a broad framework may not fully fit all disease combinations or realistic situations. To reduce the degree of inapplicability, after completion of the framework, we will continue to monitor its use with regular updates as needed. INTERNATIONAL REGISTERED REPORT DERR1-10.2196/63339