LDLiang Du

Liang Du

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Liang Du has a total of 195 co-authors. Here are the top co-authors by number of shared publications:

  • Jin Huang

    Health Management Center, General Practice Medical Center, Innovation Institute for Integration of Medicine and Engineering, West China Hospital, Sichuan University, Chengdu, China

    12 shared publications
  • Youlin Long

    Chinese Evidence-Based Medicine Center, West China Hospital, Sichuan University, Chengdu, People’s Republic of China

    10 shared publications
  • Qiong Guo

    Innovation Institute for Integration of Medicine and Engineering, West China Hospital, Sichuan University, Chengdu, People’s Republic of China

    8 shared publications
  • Yifei Lin

    Department of Urology, West China Hospital, Sichuan University, Chengdu, People’s Republic of China

    7 shared publications
  • Liu Yang

    General Practice Medical Center, and Chinese Evidence-Based Medicine Center, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, China

    6 shared publications
  • Yifan Cheng

    General Practice Medical Center, and Chinese Evidence-Based Medicine Center, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, China

    6 shared publications

Latest Publications

  • Open Access

    Potential and limitations of a large language model in the statistical review of comparative categorical data: an exploratory study of structured prompt-guided approach

    by Min Dong, Xuemei Liu, Yunmei Luo, Yuanxiang Zeng, Fang Fei, Liang Du

    Background: Statistical review is essential for research quality and integrity, yet traditional manual review is inefficient. Large language models (LLMs) offer potential support but are unreliable when used without guidance for precise calculations and raise concerns about accountability. This study evaluated whether a structured, rule-based prompt can reliably constrain an LLM to perform statistical review of comparative categorical data, and characterized both its feasibility and its inherent risks from an accountability perspective. Methods: This study employed a two-stage design based on the DeepSeekV3.2. In the first stage, a structured prompt was developed through dozens of "test-fail-iterate" cycles using 20 published medical articles. The prompt assigned the LLM the role of a "statistics expert" and provided a closed set of computational rules and a "recognize data-select calculation formula-calculate" workflow for analyzing categorical data, including Pearson's Chi-square test, continuity correction, and McNemar's tests. In the second stage, the performance of the final prompt was evaluated on a test set of 20 independent manuscripts. The model's output was compared against the results calculated by a senior statistician (the gold standard). The primary outcome measures were the performance in statistical method selection and numerical computation, including accuracy, sensitivity (recall), specificity, positive predictive value, negative predictive value, F1 score, and Cohen's Kappa. Secondary measures included reproducibility and efficiency. Results: The test set consisted of 15 manuscripts with independent samples and 5 with paired samples. In the assessment of the appropriateness of statistical method selection for 148 analysis items, the model achieved an accuracy of 99.3% (147/148), a sensitivity of 96.2% (25/26) (F1=98.0%, κ=0.976). For the test of computational consistency in 97 independent sample tests, the accuracy for χ 2 value consistency was 94.8% (92/97) (F1=89.3%, κ=0.859), and for P -value consistency, it was 96.9% (94/97) (F1=90.9%, κ=0.891). In the paired-sample analysis, the model's methods and results were in perfect agreement with the manual review, and prompt optimization eliminated discrepancies in degrees-of-freedom calculation rules. Efficiency analysis showed no statistically significant difference in time consumption between the model (407 s) and manual review (374 s) ( P =0.601). In reproducibility tests, the intraclass correlation coefficients for both χ 2 values and P -values exceeded 0.91. However, qualitative analysis revealed 3 typical failure modes in the task workflow: (1) Instability: The model's failure to produce identical outputs across repeated runs, manifesting as inconsistent data extraction or the failure to process all designated tasks (scope neglect). (2) Performance degradation/"lazy" behavior: A decline in execution quality on long or complex tasks, often characterized by the model abandoning its reasoning process to copy author-provided values without verification. (3) Anchoring effect: The model's tendency to over-rely on author-provided statistical values (the "anchor"), causing its verification process to be unduly influenced. Conclusions: A structured, rule-based prompt can guide the DeepSeek to achieve high accuracy in standardized statistical review tasks, but its reliability is contingent on operational stability. Inherent failure modes, including performance instability and a strong anchoring effect on author-provided data, persist and can lead to significant errors, particularly when source data are flawed. These findings suggest that the the DeepSeek is not suitable for autonomous auditing. Their most appropriate application is as assistive tools within a human-in-the-loop framework, where rigorous human supervision is essential for risk mitigation and to maintain ultimate accountability.

  • Open Access

    The Trial Bank initiative to promote trustworthy living evidence for medication safety (PROMISE) of children: design and pilot validation

    by Chang Xu, Suhail A. Doi, Sunita Vohra, Luis Furuya-Kanamori, Yingxia Wang, Lijun Tang, Yuan Tian, Shiqi Fan, Zhen Peng, Evan Mayo-Wilson, et al.

    While medications are essential for preventing and treating disease, they can also cause harm. Evidence synthesis has been widely adopted for evaluating ha


  • Open Access

    Incidence and outcome characteristics of adverse event in surgery: an assessment based on systematic reviews of barbed suture

    by Xianlin Gu, Qiong Guo, Qingwen Zhao, Guiyu Jiang, Yanjiao Shen, Youlin Long, Xin Chen, Yifei Lin, Jin Huang, Liang Du

    Background: Systematic reviews (SRs) have affirmed the efficacy of barbed sutures (BS), but raise safety concerns, particularly in adverse events (AEs). We conducted a cross-sectional study to comprehensively collate SRs on BS, summarize the incidence of AEs, and analyze the characteristics of the reported indicators. Methods: A thorough search was conducted in PubMed, Embase, China National Knowledge Infrastructure, Wanfang Database, China Biology Medicine disc and VIP Database from inception to April 14, 2025, and the methodological quality was evaluated using AMSTAR-2. AE characteristics of the SR were evaluated utilizing a self-developed item evaluation form. Results: A total of 52 SRs were included, encompassing 15 surgical procedures and 32 AEs. In terms of methodological quality, 1 SR was rated as high, 5 as moderate, 10 as low, and 36 as critically low. Notably, 92.0% of the SRs did not provide definitions for AEs, 76.0% did not consider severity grading, and none of the SRs assessed the adequacy of sample size. The use of BS significantly reduced the occurrence of needle pricks and biliary fistulas, in addition to increase the rates of suture breakage. Statistically insignificant results included wound dehiscence, blister, exudation, peri − incisional erythema, aseptic redness, hematoma, bleeding, urinary fistulae, stitch abscess, cellulitis, anastomotic stenosis and leak. Inconsistencies were observed in ecchymosis, vaginal cuff dehiscence, postoperative leakage rate after radical prostatectomy, infection, anastomotic bleeding, and overall postoperative complications across different SRs. Conclusion: Evidence from SRs has demonstrated advantages as well as certain disadvantages of BSs — such as an increase in the likelihood of suture breakage and wound-related complications. Current evidence remains inadequate and inconsistent, due to lack of definitions for AEs, severity grading, sample size assessments, and details on suture brand and location of use.


  • Open Access

    Transparency of clinical practice guidelines: A mixed methods research

    by Xinyi Wang, Youlin Long, Tengyue Hu, Zixin Yang, Liqin Liu, Liu Yang, Yifan Cheng, Ran Gu, Yanjiao Shen, Nan Yang, et al.


  • Open Access

    Trends and determinants of nurses’ mental health following the COVID-19 pandemic in China: a longitudinal, comparative study over a two-year period

    by Yan Liu, Yifei Lin, Wenyao Cui, Xianlin Gu, Youlin Long, Wenjie Liu, Ga Liao, Liang Du, Jin Huang

    Introduction: The COVID-19 pandemic has placed nurses at the forefront of healthcare, exposing them to various mental health challenges such as depression, anxiety, and post-traumatic stress disorder (PTSD). However the long-term effects and risk factors of the COVID-19 pandemic on nurses’ mental health are unknown. The objective of our study is to investigate the enduring impact of the COVID-19 pandemic on the psychological distress and PTSD of nurses, while also identifying the factors that influence these outcomesMethods: To investigate this question, we conducted a self-reported questionnaire survey of 8785 registered nurses recruited using snowball sampling methodology from 22 provinces in China, at five time points (T0-T4). At all times, we took measures of general health, while at the last four times we also measured PTSD. And we used logistic regression analysis to explore their impacts.Results: The incidence of whole levels psychological distress among nurses was 27.7% at T0, peaking at 57.6% at T4. For severe psychological distress, it began at 5.5% at T0 and rose to 9.5% at T4. PTSD rates among nurses were recorded at 7.8% in T1, reaching a maximum of 14.7% in T4. They all got progressively worse in the volatility. In all times, coping strategies are independent factors for various levels of psychological distress and PTSD, while social support is a key factor for severe psychological distress and PTSD.Conclusions: Our findings indicated a gradual deterioration in the mental health of nurses following the COVID-19 outbreak. In all instances, coping strategies exhibit an independent correlation with all grades of psychological distress and post-traumatic stress disorder (PTSD), whereas social support emerges as an independent protective factor mitigating the risk of severe psychological distress and PTSD.


  • Open Access

    Discrepancies between self-reported medication in adherence and indirect measurement adherence among patients undergoing antiretroviral therapy: a systematic review

    by Rujun Liao, Zihuan Tang, Na Zhang, Lin Hu, Zongqi Chang, Jiayi Ren, Xuefei Bai, Jinhong Shi, Sisi Fan, Rong Pei, et al.

    Background: Given the critical importance of medication adherence in HIV/AIDS treatment, this study aims to compare medication adherence measured by self-report (SR) and indirect measurement among antiretroviral therapy (ART) patients, exploring the differences of adherence results measured by different tools. Methods: We systematically searched PubMed, Embase, and the Cochrane Library to identify all relevant literature published up to November 22, 2023, without language restrictions, reporting adherence to ART measured by both SR and indirect measurement methods, while also analyzing individual and group adherence separately. Discrepancies between SR and indirect measurement results were assessed using the Mann–Whitney U test or Wilcoxon signed-rank test, with correlations evaluated using the Pearson correlation coefficient. Following one-to-one comparisons, meta-epidemiological one-step analysis was conducted, and network meta-analysis techniques were applied to compare results obtained through specific adherence assessment tools reported in the identified articles. Results: The analysis encompassed 65 original studies involving 13,667 HIV/AIDS patients, leading to 112 one-to-one comparisons between SR and indirect measurement tools. Statistically significant differences were observed between SR and indirect measurement tools regarding both individual and group adherence (P < 0.05), with Pearson correlation coefficients of 0.843 for individual adherence and 0.684 for group adherence. During meta-epidemiological one-step analysis, SR-measured adherence was determined to be 3.94% (95% CI: -4.48–13.44%) higher for individual adherence and 16.14% (95% CI: 0.81–18.84%) higher for group adherence compared to indirectly measured results. Subgroup analysis indicated that factors such as the year of reporting and geographic region appeared to influence the discrepancies between SR and indirect measurements. Furthermore, network meta-analysis revealed that for both individual and group adherence, the results obtained from most SR and indirect measurement tools were higher than those from electronic monitoring devices, with some demonstrating statistical significance (P < 0.05). Conclusions: The findings underscored the complexity of accurately measuring medication adherence among ART patients. Significant variability was observed across studies, with self-report methods showing a significant tendency towards overestimation. Year of reporting, geographic region, and adherence measurement tools appeared to influence the differences between SR and indirect measurements. Future research should focus on developing and validating integrated adherence measurements that can combine SR data with indirect measures to achieve a more comprehensive understanding of adherence behaviors. Graphical


  • Open Access

    Association between TB delay and TB treatment outcomes in HIV-TB co-infected patients: a study based on the multilevel propensity score method

    by Rujun Liao, Lin Hu, Jie Yu, Ying Chen, Miaoshuang Chen, Jingmin Yan, Xin Li, Xinyue Han, Chunnong Jike, Gang Yu, et al.

    Background: HIV-tuberculosis (HIV-TB) co-infection is a significant public health concern worldwide. TB delay, consisting of patient delay, diagnostic delay, treatment delay, increases the risk of adverse anti-TB treatment (ATT) outcomes. Except for individual level variables, differences in regional levels have been shown to impact the ATT outcomes. However, few studies appropriately considered possible individual and regional level confounding variables. In this study, we aimed to assess the association of TB delay on treatment outcomes in HIV-TB co-infected patients in Liangshan Yi Autonomous Prefecture (Liangshan Prefecture) of China, using a causal inference framework while taking into account individual and regional level factors. Methods: We conducted a study to analyze data from 2068 patients with HIV-TB co-infection in Liangshan Prefecture from 2019 to 2022. To address potential confounding bias, we used a causal directed acyclic graph (DAG) to select appropriate confounding variables. Further, we controlled for these confounders through multilevel propensity score and inverse probability weighting (IPW). Results: The successful rate of ATT for patients with HIV-TB co-infection in Liangshan Prefecture was 91.2%. Total delay (OR = 1.411, 95% CI: 1.015, 1.962), diagnostic delay (OR = 1.778, 95% CI: 1.261, 2.508), treatment delay (OR = 1.749, 95% CI: 1.146, 2.668) and health system delay (OR = 1.480 95% CI: (1.035, 2.118) were identified as risk factors for successful ATT outcome. Sensitivity analysis demonstrated the robustness of these findings. Conclusions: HIV-TB co-infection prevention and control policy in Liangshan Prefecture should prioritize early treatment for diagnosed HIV-TB co-infected patients. It is urgent to improve the health system in Liangshan Prefecture to reduce delays in diagnosis and treatment.


  • Open Access

    Characterization and trajectories of hematological parameters prior to severe COVID-19 based on a large-scale prospective health checkup cohort in western China: a longitudinal study of 13-year follow-up

    by Yifei Lin, Yong Yang, Nanyan Xiang, Le Wang, Tao Zheng, Xuejun Zhuo, Rui Shi, Xiaoyi Su, Yan Liu, Ga Liao, et al.

    Background: The relaxation of the “zero-COVID” policy on Dec. 7, 2022, in China posed a major public health threat recently. Complete blood count test was discovered to have complicated relationships with COVID-19 after the infection, while very few studies could track long-term monitoring of the health status and identify the characterization of hematological parameters prior to COVID-19. Methods: Based on a 13-year longitudinal prospective health checkup cohort of ~ 480,000 participants in West China Hospital, the largest medical center in western China, we documented 998 participants with a laboratory-confirmed diagnosis of COVID-19 during the 1 month after the policy. We performed a time-to-event analysis to explore the associations of severe COVID-19 patients diagnosed, with 34 different hematological parameters at the baseline level prior to COVID-19, including the whole and the subtypes of white and red blood cells. Results: A total of 998 participants with a positive SARS-CoV-2 test were documented in the cohort, 42 of which were severe cases. For white blood cell-related parameters, a higher level of basophil percentage (HR = 6.164, 95% CI = 2.066–18.393, P = 0.001) and monocyte percentage (HR = 1.283, 95% CI = 1.046–1.573, P = 0.017) were found associated with the severe COVID-19. For lymphocyte-related parameters, a lower level of lymphocyte count (HR = 0.571, 95% CI = 0.341–0.955, P = 0.033), and a higher CD4/CD8 ratio (HR = 2.473, 95% CI = 1.009–6.059, P = 0.048) were found related to the risk of severe COVID-19. We also observed that abnormality of red cell distribution width (RDW), mean corpuscular hemoglobin concentration (MCHC), and hemoglobin might also be involved in the development of severe COVID-19. The different trajectory patterns of RDW-SD and white blood cell count, including lymphocyte and neutrophil, prior to the infection were also discovered to have significant associations with the risk of severe COVID-19 (all P < 0.05). Conclusions: Our findings might help decision-makers and clinicians to classify different risk groups of population due to outbreaks including COVID-19. They could not only optimize the allocation of medical resources, but also help them be more proactive instead of reactive to long COVID-19 or even other outbreaks in the future.


  • Open Access

    Assessing the application of barbed sutures in comparison to conventional sutures for surgical applications: a global systematic review and meta-analysis of preclinical animal studies

    by Nanyan Mb Xiang, Yifei Lin, Xiaoyi Mb Su, Zifan Mb Hu, Jinyu Mb Zhou, Yi Mb Wu, Liang Du, Jin Huang

    Background: Following an initiative published by Lancet in 2002 and an IDEAL-D framework, the value of preclinical animal studies has garnered increasing attention in recent research. Numerous preclinical animal experiments tried to generate evidence to guide the development of barbed sutures. However, discernible drawbacks and incongruities in outcomes have emerged between clinical and preclinical animal studies. Therefore, this meta-analysis aimed to review the preclinical animal experiments comparing barbed sutures with conventional sutures. The authors hope to facilitate clinical translation of barbed sutures by evaluating effectiveness, safety, and physical properties/reliability. Materials and methods: A systematic search of PubMed, EMBASE, Cochrane Library, and ClinicalTrials.gov was conducted to identify controlled preclinical animal experiments comparing barbed sutures with conventional sutures. The risk of bias was assessed using SYRCLE. GRADE approach was used to evaluate evidence quality. Revman was applied to analyze all the data. Subgroup, sensitivity, and meta-regression analyses were also performed. Results: A total of 62 articles were eligible with low to moderate quality, including 2158 samples from 10 different animal species across 27 surgical procedures. Barbed suture exhibited a significant reduction in suture time, limited change in Cross-Sectional Area (CSA), and decreased instances of tissue disruption (all P<0.05). Subgroup analyses, considering both clinical and research significance, indicated that barbed sutures might cause more specific adverse events and demonstrate suboptimal performance of physical properties/reliability. Meta-regression suggested that heterogeneity resulted from variations in studies and animal models. Conclusion: Although barbed suture demonstrated superiority in numerous surgeries for time efficiency, its safety and physical properties/reliability might be influenced by diverse preclinical models, sutures’ brand, surgeries, and anatomical sites. Further evaluation, based on standardized and well-designed animal experiments, is essential for medical device development and applications in human beings.


  • Effectiveness and safety of knotless barbed sutures in cosmetic surgery: A systematic review and meta-analysis

    by Xiaoyi Su, Yifei Lin, Yi Wu, Kun Feng, Nanyan Xiang, Zifan Hu, Jinyu Zhou, Qiong Guo, Zhenglong Chen, Ga Liao, et al.