背景 虚拟标准化患者作为医学教育中的新型教学工具, 已广泛用于提升学生的临床问诊能力。随着生成式人工智能的快速发展, 基于大语言模型（LLMs）构建的VSP系统成为研究热点。然而, 目前尚缺乏对不同LLM在模拟患者角色方面表现的系统比较。目的 比较ChatGPT-4o与DeepSeek两种主流LLM在VSP模拟中的适用性, 评估其在病史采集、语言自然度、线索引导能力及教学辅助效果等方面的表现差异。方法 采用类实验研究,参与者为某医学院校临床医学专业本科四年级学生, 所有参与者均已修完《诊断学》课程, 具备基础问诊技能, 研究对象共60人, 按学号尾数单双分为两组, 分别与ChatGPT-4o或DeepSeek驱动的VSP系统进行交互。进行模拟急性阑尾炎问诊, 并在完成病史采集后提交诊断判断与体验问卷。结果 ChatGPT-4o在结构化信息整合、线索引导及技术稳定性方面更为优越, 而DeepSeek则在语言亲和力与情感回应方面表现更具人文关怀色彩。结论 不同LLM在VSP中的优势方向不同, 可根据教学目标进行有针对性地系统选择与设计。未来研究可进一步拓展至不同病种、交互方式及评估维度,以全面评估LLM驱动VSP在医学教育场景下的适应性与教学成效。

Background Virtual standardized patients（VSPs）have emerged as a novel tool in medical education, widely adopted to enhance students’ clinical interview skills．With the rapid development of generative artificial intelligence, VSP systems powered by large language models（LLMs）have become a new focus of research．However, few studies have systematically compared the performance of different LLMs in simulating patient roles．Objective This study aims to compare the applicability of two mainstream LLMs, ChatGPT-4o and DeepSeek, in VSP-based medical interview simulations, focusing on their differences in history-taking performance,linguistic naturalness, clue guidance,and educational support．Methods A quasi-experimental study was conducted involving 60 fourth-year clinical medicine undergraduates from a medical school．All participants had completed a diagnostics course and possessed basic interviewing skills．Students were assigned to either the ChatGPT-4o or DeepSeek group based on the parity of their student ID numbers．Each participant conducted a text-based simulated interview with a VSP presenting with acute appendicitis, then submitted both a preliminary diagnosis and a structured satisfaction questionnaire．Results ChatGPT-4o demonstrated superior performance in structured information integration, clue-based prompting, and system stability．In contrast, DeepSeek showed more natural language affinity and emotional responsiveness,reflecting stronger humanistic communication traits．The two models displayed divergent strengths within the VSP framework, suggesting that system selection and integration should be tailored to specific teaching objectives．Conclusions Future research should expand the scope to include diverse disease scenarios, interaction modalities, and evaluation dimensions, to comprehensively assess the educational utility and adaptability of LLM-driven VSP systems in medical training．

受试者招募工作关乎临床研究质量与进度。无法按计划招募到合适的受试者,一直是研究者发起的临床研究（IIT）开展过程中面临的主要挑战之一。本文分析影响IIT项目受试者招募进度的常见因素,并借鉴国内外经验,从提高受试者认知度与信任度、拓宽招募渠道、加强人文关怀、建立多中心伦理协作审查机制等方面探讨推进受试者招募的具体措施,以期为IIT研究者及科研管理部门提供参考。

Recruitment of subjects is crucial to the quality and progress of clinical research．However,the inability to recruit suitable subjects according to the plan has been one of the major challenges faced by investigators in the process of conducting investigator-initiated trial（IIT）．This article analyzes the common factors that affect the recruitment progress of IIT projects,draws on domestic and international experiences,and explores specific measures to promote subject recruitment,including improving subject awareness and trust,expanding recruitment channels,enhancing humanistic care and establishing a multi-center ethical collaboration review mechanism,in order to provide reference for IIT researchers and research management departments．

目的 探究Harris-Benedict(HB)公式用于估算机械通气的危重症患者能量代谢的准确性,以及不同BMI分组对其影响。方法 使用间接能量测定法测量患者的静息能量(ICREE),通过HB公式计算其静息能量代谢估算值(HBREE)。将80例患者按BMI分为4组,并通过配对样本t检验对ICREE与HBREE进行比较,Pearson分析用于分析ICREE与HBREE的相关性。结果 共纳入80例机械通气的危重症患者。除肥胖组外的其余各组病人,ICREE均高于HBREE(均P＜0.01),HB公式的准确率为23.75%。ICREE与HBREE相关性差(r=0.331,P＜0.01)。当各组使用校正系数对HB公式进行调整后,ICREE与HBREE差异无统计学意义,准确率提高至38.8%。结论 使用HB公式不能很好地反应危重症患者的实际能量代谢。BMI可能是影响HB公式准确性的重要因素。依据不同BMI分组,使用相应校正系数可提高HB公式的准确性。

Objective To explore the accuracy of the Harris-Benedict (HB) formula used to estimate the energy metabolism in critically ill patients undergoing mechanical ventilation and the effects of different BMI groups on it. Methods Indirect calorimetry was used to measure the resting energy of the patient,and the estimated resting energy metabolism was calculated by the HB formula. 80 patients were divided into four groups according to BMI. ICREE and HBREE were compared by paired sample t test. Pearson analysis was used to analyze the correlation between ICREE and HBREE. Results This study included 80 critically ill patients undergoing mechanical ventilation.Except for the obese group,ICREEE was higher than HBREE in all patients and the remaining groups of patients. The accuracy rate of the HB formula was 23.75%. The correlation between ICREE and HBREE is poor(r=0.331,P＜0.01). There was no statistical difference between ICREEE and HBREE and the accuracy rate increased to 38.8% after the Harris-Benedict equation was adjusted by using the correction factor. Conclusion Using the HB formula can not reflect the actual energy metabolism of critically ill patients well. BMI may be an important factor affecting the accuracy of HB formula. The accuracy of the HB equation can be improved by using different correction factors according to different BMI groupings.

目的 调查盐酸氨溴索对放射性肺损伤中转化生长因子β1(TGF-β1)以及肿瘤坏死因子α(TNF-a)水平的影响。方法 选取共98例在放射治疗局部晚期肺癌患者,随机分为治疗组和对照组。自放疗开始予治疗组中患者盐酸氨溴索口服,剂量60 mg,每天三次,持续应用3个月。然后对两组患者血浆中TGF-β1和TNF-α的水平进行分析。临床症状和病情变化情况采用高分辨率计算机断层扫描进行检测。结果 对照组中TGF-β1水平显著升高(11.8±5.5 ng/mL),而在盐酸氨溴索治疗组中,增加不显著(5.5±2.6 ng/mL,P＜0.001)。同样,对照组中TNF-α的水平也较治疗组中升高,(对照组:5.1±1.3,治疗组:2.6±0.8 ng/mL,P＜0.001)。结论 盐酸氨溴索能有效降低放疗后血浆TGF-β1及TNF-α水平,降低早期出现的放射性肺炎和晚期出现的肺纤维化发生机率,提高治疗效果及患者生活质量。

Objective The aim is to investigate the effect of ambroxol on radiation lung injury and the expression of transforming growth factor β1(TGF-β1),and tumor necrosis factor α(TNF-α)in plasma. Methods Ninety-eight patients with locally advanced lung cancer in radiotherapy were randomized into treatment and control groups.Patients in the treatment group took ambroxol orally at a dosage of 60 mg,three times per day for 3 months from the beginning of radiotherapy.The expression of TGF-β1 and TNF-αin plasma was analyzed.The clinical symptoms and lung diffusing capacity were monitored using high resolving power computed tomography. Results The level of TGF-β1 in the control group was increased(11.8 ± 5.5 ng/mL),whereas in ambroxol-treated patients,the increase was not significant(5.5 ± 2.6 ng/mL,P＜0.001). Radiotherapy-induced elevation of TNF-α levels,seen in control patients,was also abolished after treatment with ambroxol(5.1 ± 1.3 vs 2.6 ± 0.8 ng/mL,P＜0.001). Conclusion Ambroxol can obviously decrease the plasma TGF-β1 and TNF-α levels after radiotherapy,and decrease the chances of early radiation pneumonitis and late pulmonary fibrosis,and improve treatment effect and quality of life of patients.