首页 期刊 论文 学科刊群 资讯 加盟 关于
EN
image
Article(id=1322924541126779839, tenantId=1146029695717560320, journalId=1220038315760373764, issueId=1322924415817757501, articleNumber=null, orderNo=null, doi=10.12102/j.issn.1009-6493.2026.17.016, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1766419200000, receivedDateStr=2025-12-23, revisedDate=1783958400000, revisedDateStr=2026-07-14, acceptedDate=null, acceptedDateStr=null, onlineDate=1793339389353, onlineDateStr=2026-10-30, pubDate=1788969600000, pubDateStr=2026-09-10, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1793339389353, onlineIssueDateStr=2026-10-30, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1793339389353, creator=13701087609, updateTime=1793339389353, updator=13701087609, issue=Issue{id=1322924415817757501, tenantId=1146029695717560320, journalId=1220038315760373764, year='2026', volume='40', issue='17', pageStart='2961', pageEnd='3152', issueExtLink='null', onlineDate='null', pubDate='1788969600000', pubDateStr='2026-09-10', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1793339359478, creator='13701087609', updateTime=1793343089343, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext={EN=IssueExt(id=1322940060072542250, tenantId=1146029695717560320, journalId=1220038315760373764, issueId=1322924415817757501, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1322940060072542251, tenantId=1146029695717560320, journalId=1220038315760373764, issueId=1322924415817757501, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null, downloadFileDto=null}, startPage=3068, endPage=3073, ext={EN=ArticleExt(id=1322924541453935553, articleId=1322924541126779839, tenantId=1146029695717560320, journalId=1220038315760373764, language=EN, title=Methodological study on topic modeling based on hospital needlestick injury reports: comparison of LDA and BERTopic in risk identification and management decision support, columnId=1220689484526572183, journalTitle=Chinese Nursing Research, columnName=Original Articles, runingTitle=null, highlight=null, articleAbstract=
Objective

Taking the hospital needlestick injury report as the application scenario,to construct a comprehensive assessment framework of "theme quality-interpretability-action transformation-technical characteristics" to compare the potential Dirichlet scores,the differences between the two topic modeling methods,LDA and the Deep Learning Topic Framework Based on Pre-trained Language Models(BERTopic),in risk identification and management decision support,and input the theme-driven intervention suggestions.

Methods

A total of 160 needlestick injury reports were collected from a tertiary hospital between January 2018 and September 2025.After preprocessing including Chinese word segmentation,stop-word removal,and synonym merging,LDA and BERTopic models were trained separately.Model comparison was conducted using topic coherence,expert blind evaluation,and technical indicators such as training time,parameter sensitivity,and degree of automation.The resulting topics were then mapped onto a tiered intervention framework to formulate actionable countermeasures.

Results

Both models identified key risk scenarios including blood specimen collection, needle withdrawal/needle handling,and complex procedures in special departments.LDA generated five macro-level topics,suitable for rapid overview and routine monitoring.BERTopic generated four semantically coherent topics and one noise cluster, which more clearly delineated actionable scenarios such as blood collection chain risks among intern nurses,systemic failure in post-needle processing,and exposure in high-infection-risk cases.The topic coherence scores for BERTopic and LDA were 0.587 and 0.452, respectively.The overall expert rating scores were 4.555 and 3.944,respectively.Training times were 182 seconds and 45 seconds,respectively.

Conclusion

BERTopic demonstrates superior advantages in semantic depth and action translatability, making it suitable for root cause analysis and targeted improvement initiatives.LDA is better suited for rapid daily screening and trend monitoring.This study recommends a tiered application strategy of "routine LDA monitoring plus periodic BERTopic deep dives" in hospitals.The topic results can be transformed into structured interventions and evaluation indicators targeting populations,workflows,and devices,thereby forming a closed-loop management cycle of risk identification-intervention-review.

, authors=Tingfen CAO, Chao YAN, Tao ZHONG, Ting JIANG, authorsList=Tingfen CAO, Chao YAN, Tao ZHONG, Ting JIANG, authorCompany=null, correspAuthors=Chao YAN, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1322924541877560259, articleId=1322924541126779839, tenantId=1146029695717560320, journalId=1220038315760373764, language=CN, title=基于医院针刺伤报告的主题建模方法学研究：LDA与BERTopic用于风险识别与管理决策支持的比较, columnId=1220689484681761433, journalTitle=护理研究, columnName=科研论著, runingTitle=null, highlight=null, articleAbstract=
目的

以医院针刺伤报告为应用场景，构建“主题质量-可解释性-行动转化-技术特性”的综合评估框架，比较潜在狄利克雷分配（LDA）与基于预训练语言模型的深度学习主题框架（BERTopic）两种主题建模方法在风险识别与管理决策支持中的差异，并输出主题驱动的干预建议。

方法

收集某三级甲等医院2018年1月—2025年9月医院感染管理信息系统中针刺伤报告160份，完成中文分词、去停用词与同义归并等预处理后，分别训练LDA与BERTopic模型。采用主题一致性、专家盲评及训练时间、参数敏感性、自动化程度等技术指标进行对比，并将主题结果映射至分层干预框架形成可操作对策。

结果

两种模型均识别出血标本采集、针头处理过程、特殊科室复杂操作等针刺伤的关键风险环节。LDA生成5个宏观主题，适用于快速概览与常规监测；BERTopic生成4个语义聚合主题及1个噪声类，能够更清晰地刻画实习护士采血链风险、针头后处理系统失效、高感染风险病例暴露等可直接转化为管理行动的场景。BERTopic与LDA主题一致性得分分别为0.587、0.452分，专家总体评分分别为4.555、3.944分，训练时间分别为182、45 s。

结论

BERTopic在语义深度与行动转化方面更具优势，适合根因分析与专项改进；LDA更适合日常快速筛查与趋势监测。建议医院采用“日常LDA监测+阶段性BERTopic深挖”的分层应用策略，将主题结果转化为面向人群、流程与器具的结构化干预与评估指标，形成风险识别-干预-复盘的闭环管理。

, authors=曹廷芬, 晏超*, *, 钟涛, 蒋婷, authorsList=曹廷芬, 晏超, 钟涛, 蒋婷, authorCompany=null, correspAuthors=晏超, authorNote=

曹廷芬，副主任护师，本科

, correspAuthorsNote=
晏超，E-mail：
YAN Chao, E-mail:
, copyrightStatement=除非特别声明，本刊刊出的所有文章不代表本刊编委会的观点。, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=y8c5NQvlfjFsosWluPDPHQ==, magXml=zuwH0Z2wS0OuVjKO7Qi8zQ==, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=G5pqi1nXpvQ5dUXexWokXw==, mapNumber=null, fund=null)}, authors=[Author(id=1322924542242464714, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322924542330545100, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542242464714, language=EN, stringName=Tingfen CAO, firstName=Tingfen, middleName=null, lastName=CAO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924542435402701, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542242464714, language=CN, stringName=曹廷芬, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1贵州航天医院（护理部），贵州 563000, bio={"content":"

曹廷芬，副主任护师，本科

"}, bioImg=null, bioContent=

曹廷芬，副主任护师，本科

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])]), Author(id=1322924542682866639, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=muyanchao0520@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1322924542779335633, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542682866639, language=EN, stringName=Chao YAN, firstName=Chao, middleName=null, lastName=YAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924543077131218, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542682866639, language=CN, stringName=晏超, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, *, address=1贵州航天医院（护理部），贵州 563000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])]), Author(id=1322924543156822996, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322924544851321816, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924543156822996, language=EN, stringName=Tao ZHONG, firstName=Tao, middleName=null, lastName=ZHONG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924544931013593, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924543156822996, language=CN, stringName=钟涛, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1贵州航天医院（护理部），贵州 563000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])]), Author(id=1322924545014899675, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322924545102980061, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924545014899675, language=EN, stringName=Ting JIANG, firstName=Ting, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924545170088926, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924545014899675, language=CN, stringName=蒋婷, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1贵州航天医院（护理部），贵州 563000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])])], keywords=[Keyword(id=1322924545295918047, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, orderNo=1, keyword=needlestick injury), Keyword(id=1322924545379804128, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, orderNo=2, keyword=topic modeling), Keyword(id=1322924545451107297, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, orderNo=3, keyword=latent Dirichlet allocation(LDA)), Keyword(id=1322924545518216162, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, orderNo=4, keyword=BERTopic), Keyword(id=1322924545593713635, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, orderNo=5, keyword=medical safety), Keyword(id=1322924545660822500, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, orderNo=6, keyword=incident reports), Keyword(id=1322924545732125669, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, orderNo=7, keyword=methodology), Keyword(id=1322924545807623142, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, orderNo=1, keyword=针刺伤), Keyword(id=1322924545874732007, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, orderNo=2, keyword=主题建模), Keyword(id=1322924545937646568, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, orderNo=3, keyword=狄利克雷分配（LDA）), Keyword(id=1322924546004755433, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, orderNo=4, keyword=基于预训练语言模型的深度学习主题框架（BERTopic）), Keyword(id=1322924546084447210, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, orderNo=5, keyword=医疗安全), Keyword(id=1322924546159944683, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, orderNo=6, keyword=主题报告), Keyword(id=1322924546227053548, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, orderNo=7, keyword=方法学)], refs=[Reference(id=1322924550358443003, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2025, volume=23, issue=3, pageStart=133, pageEnd=137, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=Mccabe F J, Dunne N, Farrington S K, journalName=The Surgeon, refType=null, unstructuredReference=Mccabe F J,Dunne N,Farrington S K,et al.Needlestick injury incidence and reporting in Irish surgical trainees[J].The Surgeon,2025,23(3):133-137., articleTitle=Needlestick injury incidence and reporting in Irish surgical trainees, refAbstract=null), Reference(id=1322924550480077820, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2022, volume=21, issue=1, pageStart=298, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=Zhang L H, Li Q, Guan L, journalName=BMC Nursing, refType=null, unstructuredReference=Zhang L H,Li Q,Guan L,et al.Prevalence and influence factors of occupational exposure to blood and body fluids in registered Chinese nurses:a national cross-sectional study[J].BMC Nursing,2022,21(1):298., articleTitle=Prevalence and influence factors of occupational exposure to blood and body fluids in registered Chinese nurses:a national cross-sectional study, refAbstract=null), Reference(id=1322924550576546813, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2021, volume=53, issue=3, pageStart=296, pageEnd=305, url=null, language=null, rfNumber=3, rfOrder=2, authorNames=Homs-Romero E, Romero-Collado A, Verdú J, journalName=Journal of Nursing Scholarship, refType=null, unstructuredReference=Homs-Romero E,Romero-Collado A,Verdú J,et al.Validity of chronic venous disease diagnoses and epidemiology using validated electronic health records from primary care:a real-world data analysis[J].Journal of Nursing Scholarship,2021,53(3):296-305., articleTitle=Validity of chronic venous disease diagnoses and epidemiology using validated electronic health records from primary care:a real-world data analysis, refAbstract=null), Reference(id=1322924550765290494, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2017, volume=141, issue=9, pageStart=1267, pageEnd=1275, url=null, language=null, rfNumber=4, rfOrder=3, authorNames=Aeffner F, Wilson K, Martin N T, journalName=Archives of Pathology & Laboratory Medicine, refType=null, unstructuredReference=Aeffner F,Wilson K,Martin N T,et al.The gold standard paradox in digital image analysis:manual versus automated scoring as ground truth[J].Archives of Pathology & Laboratory Medicine,2017,141(9):1267-1275., articleTitle=The gold standard paradox in digital image analysis:manual versus automated scoring as ground truth, refAbstract=null), Reference(id=1322924550857565183, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2022, volume=22, issue=1, pageStart=287, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=4, authorNames=Liu F, Panagiotakos D, journalName=BMC Medical Research Methodology, refType=null, unstructuredReference=Liu F,Panagiotakos D.Real-world data:a brief review of the methods,applications,challenges and opportunities[J].BMC Medical Research Methodology,2022,22(1):287., articleTitle=Real-world data:a brief review of the methods,applications,challenges and opportunities, refAbstract=null), Reference(id=1322924551021143040, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2025, volume=7, issue=null, pageStart=1610228, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=5, authorNames=Veugen T, Dunning V, Marcus M, journalName=Frontiers in Digital Health, refType=null, unstructuredReference=Veugen T,Dunning V,Marcus M,et al.Secure latent dirichlet allocation[J].Frontiers in Digital Health,2025,7:1610228., articleTitle=Secure latent dirichlet allocation, refAbstract=null), Reference(id=1322924551130193920, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2023, volume=33, issue=11, pageStart=8014, pageEnd=8016, url=null, language=null, rfNumber=7, rfOrder=6, authorNames=Tejani A S, journalName=European Radiology, refType=null, unstructuredReference=Tejani A S.To BERT or not to BERT:advancing non-invasive prediction of tumor biomarkers using transformer-based natural language processing(NLP)[J].European Radiology,2023,33(11):8014-8016., articleTitle=To BERT or not to BERT:advancing non-invasive prediction of tumor biomarkers using transformer-based natural language processing(NLP), refAbstract=null), Reference(id=1322924551226662913, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2025, volume=15, issue=null, pageStart=2703, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=7, authorNames=Yang B B, Lu W, Xuan Y, journalName=Scientific Reports, refType=null, unstructuredReference=Yang B B,Lu W,Xuan Y,et al.The influences of social support expressed from doctors and disclosed from peers on patient decision-making:an analysis from the online health community[J].Scientific Reports,2025,15:2703., articleTitle=The influences of social support expressed from doctors and disclosed from peers on patient decision-making:an analysis from the online health community, refAbstract=null), Reference(id=1322924551465738242, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2020, volume=36, issue=4, pageStart=1234, pageEnd=1240, url=null, language=null, rfNumber=9, rfOrder=8, authorNames=Lee J, Yoon W, Kim S, journalName=Bioinformatics, refType=null, unstructuredReference=Lee J,Yoon W,Kim S,et al.BioBERT:a pre-trained biomedical language representation model for biomedical text mining[J].Bioinformatics,2020,36(4):1234-1240., articleTitle=BioBERT:a pre-trained biomedical language representation model for biomedical text mining, refAbstract=null), Reference(id=1322924551541235715, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2023, volume=28, issue=5, pageStart=1178, pageEnd=1206, url=null, language=null, rfNumber=10, rfOrder=9, authorNames=Wilcox K T, Jacobucci R, Zhang Z Y, journalName=Psychological Methods, refType=null, unstructuredReference=Wilcox K T,Jacobucci R,Zhang Z Y,et al.Supervised latent Dirichlet allocation with covariates:a Bayesian structural and measurement model of text and covariates[J].Psychological Methods,2023,28(5):1178-1206., articleTitle=Supervised latent Dirichlet allocation with covariates:a Bayesian structural and measurement model of text and covariates, refAbstract=null), Reference(id=1322924551612538884, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2024, volume=6, issue=3, pageStart=e222, pageEnd=e229, url=null, language=null, rfNumber=11, rfOrder=10, authorNames=Miao B Y, Sushil M, Xu A, journalName=The Lancet Digital Health, refType=null, unstructuredReference=Miao B Y,Sushil M,Xu A,et al.Characterisation of digital therapeutic clinical trials:a systematic review with natural language processing[J].The Lancet Digital Health,2024,6(3):e222-e229., articleTitle=Characterisation of digital therapeutic clinical trials:a systematic review with natural language processing, refAbstract=null), Reference(id=1322924551734173701, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2022, volume=24, issue=11, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=11, authorNames=Hagg L J, Merkouris S S, O'Dea G A, journalName=Journal of Medical Internet Research, refType=null, unstructuredReference=Hagg L J,Merkouris S S,O'Dea G A,et al.Examining analytic practices in latent dirichlet allocation within psychological science:scoping review[J].Journal of Medical Internet Research,2022,24(11):e33166., articleTitle=Examining analytic practices in latent dirichlet allocation within psychological science:scoping review, refAbstract=null), Reference(id=1322924551897751558, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2024, volume=26, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=12, authorNames=Ke Y H, Yang R, Liu N, journalName=Journal of Medical Internet Research, refType=null, unstructuredReference=Ke Y H,Yang R,Liu N.Comparing open-access database and traditional intensive care studies using machine learning:bibliometric analysis study[J].Journal of Medical Internet Research,2024,26:e48330., articleTitle=Comparing open-access database and traditional intensive care studies using machine learning:bibliometric analysis study, refAbstract=null), Reference(id=1322924551960666119, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, doi=null, pmid=null, pmcid=null, year=2025, volume=250, issue=null, pageStart=10389, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=13, authorNames=Ma L, Chen R, Ge W G, journalName=Experimental Biology and Medicine, refType=null, unstructuredReference=Ma L,Chen R,Ge W G,et al.AI-powered topic modeling:comparing LDA and BERTopic in analyzing opioid-related cardiovascular risks in women[J].Experimental Biology and Medicine,2025,250:10389., articleTitle=AI-powered topic modeling:comparing LDA and BERTopic in analyzing opioid-related cardiovascular risks in women, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])], figs=[ArticleFig(id=1322924546352882669, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, label=Table 1, caption=

Basic characteristics of needlestick injuries happened in the hospital（n=160）

, figureFileSmall=null, figureFileBig=null, tableContent=
项目类别人数构成比（%）
针刺伤报告人员实习护士9861.25
本院执业护士4930.63
新入职护士95.62
进修护士42.50
操作环节拔针2716.88
采集血标本2415.00
分离针头1610.00
针头放入锐器盒159.37
回套针帽116.87
清理医疗废物106.25
静脉穿刺85.00
其他4930.63
污染源类型血液14288.75
体液31.87
其他159.38
), ArticleFig(id=1322924546445157358, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, label=表1, caption=

医院针刺伤发生的基本特征（n=160）

, figureFileSmall=null, figureFileBig=null, tableContent=
项目类别人数构成比（%）
针刺伤报告人员实习护士9861.25
本院执业护士4930.63
新入职护士95.62
进修护士42.50
操作环节拔针2716.88
采集血标本2415.00
分离针头1610.00
针头放入锐器盒159.37
回套针帽116.87
清理医疗废物106.25
静脉穿刺85.00
其他4930.63
污染源类型血液14288.75
体液31.87
其他159.38
), ArticleFig(id=1322924546516460527, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, label=Table 2, caption=

Topic distribution and keywords of LDA

, figureFileSmall=null, figureFileBig=null, tableContent=
主题权重（%）核心关键词风险特点
基础护理操作风险28.1拔针、头皮钢针、实习护士、血液、换输液瓶基础操作频次高，人员经验不足
血标本采集风险25.6采集血标本、非安全型静脉采血针、血液、实习护士、乙型肝炎血源性病原体暴露风险突出
针头处理过程风险22.5分离针头、回套针帽、非安全型一次性注射器针头、血液违反安全操作规程
废物处置风险15.0清理废物、头皮钢针、将针头放入锐器盒、其他风险意识不足，防护缺失
手术相关操作风险8.8手术缝针、传递锐器、麻醉科、本院执业护士、血液复杂操作环境风险
), ArticleFig(id=1322924546587763696, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, label=表2, caption=

LDA主题分布及关键词

, figureFileSmall=null, figureFileBig=null, tableContent=
主题权重（%）核心关键词风险特点
基础护理操作风险28.1拔针、头皮钢针、实习护士、血液、换输液瓶基础操作频次高，人员经验不足
血标本采集风险25.6采集血标本、非安全型静脉采血针、血液、实习护士、乙型肝炎血源性病原体暴露风险突出
针头处理过程风险22.5分离针头、回套针帽、非安全型一次性注射器针头、血液违反安全操作规程
废物处置风险15.0清理废物、头皮钢针、将针头放入锐器盒、其他风险意识不足，防护缺失
手术相关操作风险8.8手术缝针、传递锐器、麻醉科、本院执业护士、血液复杂操作环境风险
), ArticleFig(id=1322924546667455473, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, label=Table 3, caption=

Semantic characteristics of topics based on the BERTopic

, figureFileSmall=null, figureFileBig=null, tableContent=
主题文档（个）核心关键词语义特征
实习护士血标本采集链风险36采集血标本、非安全型静脉采血针、实习护士、血液、乙型肝炎实习护士采血全流程风险
针头安全处理系统失效42回套针帽、分离针头、非安全型一次性注射器针头、拔针、血液针头后处理环节系统性失效
高感染风险病例暴露28乙型肝炎、梅毒、HIV、血液、头皮钢针、拔针已知血源性疾病严重暴露
特殊科室复杂操作风险24手术缝针、传递锐器、麻醉科、静脉穿刺、重症医学科特殊科室专业环境风险
噪声点30信息不完整或描述模糊
), ArticleFig(id=1322924546738758642, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, label=表3, caption=

BERTopic主题语义特征

, figureFileSmall=null, figureFileBig=null, tableContent=
主题文档（个）核心关键词语义特征
实习护士血标本采集链风险36采集血标本、非安全型静脉采血针、实习护士、血液、乙型肝炎实习护士采血全流程风险
针头安全处理系统失效42回套针帽、分离针头、非安全型一次性注射器针头、拔针、血液针头后处理环节系统性失效
高感染风险病例暴露28乙型肝炎、梅毒、HIV、血液、头皮钢针、拔针已知血源性疾病严重暴露
特殊科室复杂操作风险24手术缝针、传递锐器、麻醉科、静脉穿刺、重症医学科特殊科室专业环境风险
噪声点30信息不完整或描述模糊
), ArticleFig(id=1322924546805867507, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, label=Table 4, caption=

Objective indicators of topic quality

, figureFileSmall=null, figureFileBig=null, tableContent=
评估指标LDABERTopic优势模型
主题一致性得分（分）0.4520.587BERTopic
主题数量（个）54LDA
噪声识别能力无有BERTopic
), ArticleFig(id=1322924546889753588, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, label=表4, caption=

主题质量客观指标

, figureFileSmall=null, figureFileBig=null, tableContent=
评估指标LDABERTopic优势模型
主题一致性得分（分）0.4520.587BERTopic
主题数量（个）54LDA
噪声识别能力无有BERTopic
), ArticleFig(id=1322924547019777013, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, label=Table 5, caption=

Expert evaluation results(5-point scale) 单位：分

, figureFileSmall=null, figureFileBig=null, tableContent=
评估维度LDABERTopic差异
主题临床意义明确性3.8334.833-1.000
关键词与主题的关联度4.1674.500-0.333
风险防控指导价值3.8334.333-0.500
总体评分3.9444.555-0.611
), ArticleFig(id=1322924547678282742, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, label=表5, caption=

专家评估结果（5分制）

, figureFileSmall=null, figureFileBig=null, tableContent=
评估维度LDABERTopic差异
主题临床意义明确性3.8334.833-1.000
关键词与主题的关联度4.1674.500-0.333
风险防控指导价值3.8334.333-0.500
总体评分3.9444.555-0.611
), ArticleFig(id=1322924549376975863, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, label=Table 6, caption=

Technical characteristics

, figureFileSmall=null, figureFileBig=null, tableContent=
技术指标LDABERTopic适用场景
训练时间45 s182 sLDA：快速分析；BERTopic：深度分析
参数敏感性高中等BERTopic更稳定
自动化程度低高BERTopic更智能
结果可解释性需人工解读自动语义主题BERTopic更优秀
), ArticleFig(id=1322924549506999288, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, label=表6, caption=

技术特性

, figureFileSmall=null, figureFileBig=null, tableContent=
技术指标LDABERTopic适用场景
训练时间45 s182 sLDA：快速分析；BERTopic：深度分析
参数敏感性高中等BERTopic更稳定
自动化程度低高BERTopic更智能
结果可解释性需人工解读自动语义主题BERTopic更优秀
), ArticleFig(id=1322924549653799929, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=EN, label=Table 7, caption=

Topic-driven mapping of risk countermeasures and recommended monitoring indicators

, figureFileSmall=null, figureFileBig=null, tableContent=
主题/场景（BERTopic）可操作风险点（人员-环节-器具-后果）优先级对策（层级控制）建议监测指标（示例）
实习护士采血链风险实习护士；采集血标本/拔针；非安全型采血针；血源性暴露与职业感染风险①工程：优先配置安全型采血针与“就近锐器盒”；②管理：带教-准入-陪同制度；③培训：情景化模拟+胜任力考核；④防护：采血场景规范佩戴与双人核对安全型采血针使用率；实习护士采血考核通过率；采血相关针刺伤发生率
针头后处理系统失效回套针帽、分离针头、针头放入锐器盒不规范；流程与布局导致“手持暴露”时间延长①工程：一次性安全注射器/带护套针；锐器盒固定在操作点；②管理：禁止回套与稽核；③流程：拔针-弃置一体化；④行为：高风险环节同伴提醒回套针帽违规率；锐器盒到位率；针头处置不规范事件数
高感染风险病例暴露已知乙型肝炎/梅毒/HIV等高危污染源；暴露后处置依从性与时效性①管理：高危标识与预警；②流程：暴露后处置“绿色通道”；③培训：PEP流程演练；④防护：强化手套/面屏等PPE配置暴露后报告时效（分钟/小时）；PEP启动率；暴露后随访完成率
特殊科室复杂操作风险手术缝针/传递锐器/麻醉与ICU操作；多人员协作导致锐器交接风险①工程：无接触传递盘/磁性垫；②管理：手术间锐器清点与闭环；③流程：锐器交接口令化；④培训：团队协作情景演练手术间锐器清点合格率；交接规范执行率；相关针刺伤发生率
信息不完整/描述模糊（噪声类）关键字段缺失导致主题噪声增多，影响根因追溯与对策制定①信息化：报告单强制字段与结构化选项；②管理：报告质量质控；③培训：上报要点与示例；④反馈：数据-改进闭环通报报告完整率；“噪声类”占比；信息质控抽检合格率
), ArticleFig(id=1322924549850932218, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, language=CN, label=表7, caption=

主题驱动的风险对策映射及建议监测指标

, figureFileSmall=null, figureFileBig=null, tableContent=
主题/场景（BERTopic）可操作风险点（人员-环节-器具-后果）优先级对策（层级控制）建议监测指标（示例）
实习护士采血链风险实习护士；采集血标本/拔针；非安全型采血针；血源性暴露与职业感染风险①工程：优先配置安全型采血针与“就近锐器盒”；②管理：带教-准入-陪同制度；③培训：情景化模拟+胜任力考核；④防护：采血场景规范佩戴与双人核对安全型采血针使用率；实习护士采血考核通过率；采血相关针刺伤发生率
针头后处理系统失效回套针帽、分离针头、针头放入锐器盒不规范；流程与布局导致“手持暴露”时间延长①工程：一次性安全注射器/带护套针；锐器盒固定在操作点；②管理：禁止回套与稽核；③流程：拔针-弃置一体化；④行为：高风险环节同伴提醒回套针帽违规率；锐器盒到位率；针头处置不规范事件数
高感染风险病例暴露已知乙型肝炎/梅毒/HIV等高危污染源；暴露后处置依从性与时效性①管理：高危标识与预警；②流程：暴露后处置“绿色通道”；③培训：PEP流程演练；④防护：强化手套/面屏等PPE配置暴露后报告时效（分钟/小时）；PEP启动率；暴露后随访完成率
特殊科室复杂操作风险手术缝针/传递锐器/麻醉与ICU操作；多人员协作导致锐器交接风险①工程：无接触传递盘/磁性垫；②管理：手术间锐器清点与闭环；③流程：锐器交接口令化；④培训：团队协作情景演练手术间锐器清点合格率；交接规范执行率；相关针刺伤发生率
信息不完整/描述模糊（噪声类）关键字段缺失导致主题噪声增多，影响根因追溯与对策制定①信息化：报告单强制字段与结构化选项；②管理：报告质量质控；③培训：上报要点与示例；④反馈：数据-改进闭环通报报告完整率；“噪声类”占比；信息质控抽检合格率
)], attaches=null, journal=Journal(id=1220037649855897604, delFlag=0, nameCn=护理研究, nameEn=Chinese Nursing Research, nameHistory1=null, nameHistory2=null, issn=1009-6493, eissn=null, cn=14-1272/R, coden=null, periodic=3, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=5ND/w5RQYz//x8ucn+e6Vw==, journalPrice=null, startedYear=null, abbrevIsoEn=Chinese Nursing Research, journalRemark=null, publicationField=null, createdTime=1768809242933, updatedTime=1789628561524, createdBy=18614031015, updatedBy=13041195026, firstLetterCn=H, firstLetterEn=H, subjectCode=Medical and Pharmaceutical Sciences, subjectName=null, subjectCodeEn=Medical and Pharmaceutical Sciences, subjectNameEn=null, picCn=5ND/w5RQYz//x8ucn+e6Vw==, picEn=vXneBBf6WfTz/ls4yy3nnA==, jcr=null, cjcr=null, exts=[JournalExt(id=1307360202571928186, language=CN, name=护理研究, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1789628561871, updatedTime=1789628561871, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-sxyxqks-author&redirect_uri=https%3A%2F%2Fgj.sxyxqks.com%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=1de4c13f-4807-4ea4-8d8a-5, submissionEditorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-sxyxqks-author&redirect_uri=https%3A%2F%2Fgj.sxyxqks.com%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=321a724c-d51d-4f08-a85e-2, submissionReviewUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-sxyxqks-author&redirect_uri=https%3A%2F%2Fgj.sxyxqks.com%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=dadec7f2-d93c-4f73-8f97-e, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1307360202618065531, language=EN, name=Chinese Nursing Research, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1789628561882, updatedTime=1789628561882, createdBy=13041195026, updatedBy=13041195026, submissionGuidelinesUrl=, submissionAuthorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-sxyxqks-author&redirect_uri=https%3A%2F%2Fgj.sxyxqks.com%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=1de4c13f-4807-4ea4-8d8a-5, submissionEditorUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-sxyxqks-author&redirect_uri=https%3A%2F%2Fgj.sxyxqks.com%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=321a724c-d51d-4f08-a85e-2, submissionReviewUrl=https://journal.ids.fzyun.cn/auth/realms/journal/protocol/openid-connect/auth?client_id=journal-sxyxqks-author&redirect_uri=https%3A%2F%2Fgj.sxyxqks.com%2Foauth%2Fcallback&response_type=code&scope=phone+openid+email+profile&state=dadec7f2-d93c-4f73-8f97-e, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1220038315760373764, websiteList=[Website(id=1220043903038837076, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1220038315760373764, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/hlyj/CN, language=CN, createTime=1768810733804, createBy=18614031015, updateTime=1768811130580, updateBy=18614031015, name=护理研究-中文, tplId=1146099689490845704, title=护理研究, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1228001613021904925, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=articleTextType, value=kx, createTime=1770707999711, updateTime=1770707999711, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613005127706, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=banner, value=null, createTime=1770707999707, updateTime=1770707999707, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613038682144, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=grayFlag, value=0, createTime=1770707999715, updateTime=1770707999715, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613000933401, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=logo, value=https://castjournals.cast.org.cn/joweb/hlyj/CN/file/pic?fileId=pB29yYO+7WdTp0HLQ4KSzg==, createTime=1770707999706, updateTime=1770707999706, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613047070754, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=minRunFlag, value=0, createTime=1770707999717, updateTime=1770707999717, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613017710620, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/hlyj/CN/file/pic, createTime=1770707999710, updateTime=1770707999710, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613042876449, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=silenceFlag, value=0, createTime=1770707999716, updateTime=1770707999716, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613013516315, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1770707999709, updateTime=1770707999709, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613026099230, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=themeColor, value=null, createTime=1770707999712, updateTime=1770707999712, creator=18614031015, updator=18614031015), WebsiteProps(id=1228001613034487839, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903038837076, code=themeStyle, value=null, createTime=1770707999714, updateTime=1770707999714, creator=18614031015, updator=18614031015)]), Website(id=1220043903097557352, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1220038315760373764, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/hlyj/EN, language=EN, createTime=1768810733819, createBy=18614031015, updateTime=1768811154181, updateBy=18614031015, name=护理研究-英文, tplId=1146101810881728533, title=Chinese Nursing Research, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1220045841042166147, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=articleTextType, value=kx, createTime=1768811195860, updateTime=1768811195860, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841017000320, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=banner, value=null, createTime=1768811195854, updateTime=1768811195854, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841063137670, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=grayFlag, value=0, createTime=1768811195865, updateTime=1768811195865, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841008611711, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=logo, value=https://castjournals.cast.org.cn/joweb/hlyj/EN/file/pic?fileId=pB29yYO+7WdTp0HLQ4KSzg==, createTime=1768811195852, updateTime=1768811195852, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841075720584, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=minRunFlag, value=0, createTime=1768811195868, updateTime=1768811195868, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841033777538, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/hlyj/EN/file/pic, createTime=1768811195858, updateTime=1768811195858, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841071526279, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=silenceFlag, value=0, createTime=1768811195867, updateTime=1768811195867, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841025388929, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1768811195857, updateTime=1768811195857, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841046360452, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=themeColor, value=null, createTime=1768811195861, updateTime=1768811195861, creator=18614031015, updator=18614031015), WebsiteProps(id=1220045841054749061, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1220043903097557352, code=themeStyle, value=null, createTime=1768811195863, updateTime=1768811195863, creator=18614031015, updator=18614031015)])], journalTitle=护理研究, weixinUrl=null, journalUrl=https://www.hlyjzz.com/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Chinese Nursing Research, journalPhotoCn=5ND/w5RQYz//x8ucn+e6Vw==, journalPhotoEn=vXneBBf6WfTz/ls4yy3nnA==, journalFirstLetter=H, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/hlyj/CN/10.12102/j.issn.1009-6493.2026.17.016, detailUrlEn=https://castjournals.cast.org.cn/joweb/hlyj/EN/10.12102/j.issn.1009-6493.2026.17.016, pdfUrlCn=https://castjournals.cast.org.cn/joweb/hlyj/CN/PDF/10.12102/j.issn.1009-6493.2026.17.016, pdfUrlEn=https://castjournals.cast.org.cn/joweb/hlyj/EN/PDF/10.12102/j.issn.1009-6493.2026.17.016, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1788969600000, fullTextJson=null, articleText=null, reference=null)
收藏切换
基于医院针刺伤报告的主题建模方法学研究：LDA与BERTopic用于风险识别与管理决策支持的比较
收藏切换
曹廷芬 , 晏超 *, * , 钟涛 , 蒋婷
护理研究 | 科研论著 2026,40(17): 3068-3073
收起
收藏切换
护理研究 |科研论著 2026 , 40 (17) : 3068 -3073
基于医院针刺伤报告的主题建模方法学研究：LDA与BERTopic用于风险识别与管理决策支持的比较
全屏
[Author(id=1322924542242464714, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322924542330545100, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542242464714, language=EN, stringName=Tingfen CAO, firstName=Tingfen, middleName=null, lastName=CAO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924542435402701, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542242464714, language=CN, stringName=曹廷芬, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1贵州航天医院（护理部），贵州 563000, bio={"content":"

曹廷芬，副主任护师，本科

"}, bioImg=null, bioContent=

曹廷芬，副主任护师，本科

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])]), Author(id=1322924542682866639, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=muyanchao0520@163.com, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1322924542779335633, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542682866639, language=EN, stringName=Chao YAN, firstName=Chao, middleName=null, lastName=YAN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924543077131218, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924542682866639, language=CN, stringName=晏超, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, *, address=1贵州航天医院（护理部），贵州 563000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])]), Author(id=1322924543156822996, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322924544851321816, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924543156822996, language=EN, stringName=Tao ZHONG, firstName=Tao, middleName=null, lastName=ZHONG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924544931013593, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924543156822996, language=CN, stringName=钟涛, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1贵州航天医院（护理部），贵州 563000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])]), Author(id=1322924545014899675, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322924545102980061, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924545014899675, language=EN, stringName=Ting JIANG, firstName=Ting, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1Guizhou Aerospace Hospital, Guizhou 563000 China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322924545170088926, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, authorId=1322924545014899675, language=CN, stringName=蒋婷, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=1贵州航天医院（护理部），贵州 563000, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322924542091469765, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, xref=1, ext=[AuthorCompanyExt(id=1322924542099858374, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Guizhou Aerospace Hospital, Guizhou 563000 China), AuthorCompanyExt(id=1322924542108246983, tenantId=1146029695717560320, journalId=1220038315760373764, articleId=1322924541126779839, companyId=1322924542091469765, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1贵州航天医院（护理部），贵州 563000)])])]
曹廷芬, 晏超*, * , 钟涛, 蒋婷
作者信息
  • 1贵州航天医院（护理部），贵州 563000
通讯作者:
晏超，E-mail：
YAN Chao, E-mail:
作者简介:

曹廷芬，副主任护师，本科

Methodological study on topic modeling based on hospital needlestick injury reports: comparison of LDA and BERTopic in risk identification and management decision support
Tingfen CAO, Chao YAN , Tao ZHONG, Ting JIANG
Affiliations
  • 1Guizhou Aerospace Hospital, Guizhou 563000 China
出版时间: 2026-09-10 doi: 10.12102/j.issn.1009-6493.2026.17.016
文章导航
收藏切换
摘要
收起
目的

以医院针刺伤报告为应用场景，构建“主题质量-可解释性-行动转化-技术特性”的综合评估框架，比较潜在狄利克雷分配（LDA）与基于预训练语言模型的深度学习主题框架（BERTopic）两种主题建模方法在风险识别与管理决策支持中的差异，并输出主题驱动的干预建议。

方法

收集某三级甲等医院2018年1月—2025年9月医院感染管理信息系统中针刺伤报告160份，完成中文分词、去停用词与同义归并等预处理后，分别训练LDA与BERTopic模型。采用主题一致性、专家盲评及训练时间、参数敏感性、自动化程度等技术指标进行对比，并将主题结果映射至分层干预框架形成可操作对策。

结果

两种模型均识别出血标本采集、针头处理过程、特殊科室复杂操作等针刺伤的关键风险环节。LDA生成5个宏观主题，适用于快速概览与常规监测；BERTopic生成4个语义聚合主题及1个噪声类，能够更清晰地刻画实习护士采血链风险、针头后处理系统失效、高感染风险病例暴露等可直接转化为管理行动的场景。BERTopic与LDA主题一致性得分分别为0.587、0.452分，专家总体评分分别为4.555、3.944分，训练时间分别为182、45 s。

结论

BERTopic在语义深度与行动转化方面更具优势，适合根因分析与专项改进；LDA更适合日常快速筛查与趋势监测。建议医院采用“日常LDA监测+阶段性BERTopic深挖”的分层应用策略，将主题结果转化为面向人群、流程与器具的结构化干预与评估指标，形成风险识别-干预-复盘的闭环管理。

针刺伤  /  主题建模  /  狄利克雷分配（LDA）  /  基于预训练语言模型的深度学习主题框架（BERTopic）  /  医疗安全  /  主题报告  /  方法学
Objective

Taking the hospital needlestick injury report as the application scenario,to construct a comprehensive assessment framework of "theme quality-interpretability-action transformation-technical characteristics" to compare the potential Dirichlet scores,the differences between the two topic modeling methods,LDA and the Deep Learning Topic Framework Based on Pre-trained Language Models(BERTopic),in risk identification and management decision support,and input the theme-driven intervention suggestions.

Methods

A total of 160 needlestick injury reports were collected from a tertiary hospital between January 2018 and September 2025.After preprocessing including Chinese word segmentation,stop-word removal,and synonym merging,LDA and BERTopic models were trained separately.Model comparison was conducted using topic coherence,expert blind evaluation,and technical indicators such as training time,parameter sensitivity,and degree of automation.The resulting topics were then mapped onto a tiered intervention framework to formulate actionable countermeasures.

Results

Both models identified key risk scenarios including blood specimen collection, needle withdrawal/needle handling,and complex procedures in special departments.LDA generated five macro-level topics,suitable for rapid overview and routine monitoring.BERTopic generated four semantically coherent topics and one noise cluster, which more clearly delineated actionable scenarios such as blood collection chain risks among intern nurses,systemic failure in post-needle processing,and exposure in high-infection-risk cases.The topic coherence scores for BERTopic and LDA were 0.587 and 0.452, respectively.The overall expert rating scores were 4.555 and 3.944,respectively.Training times were 182 seconds and 45 seconds,respectively.

Conclusion

BERTopic demonstrates superior advantages in semantic depth and action translatability, making it suitable for root cause analysis and targeted improvement initiatives.LDA is better suited for rapid daily screening and trend monitoring.This study recommends a tiered application strategy of "routine LDA monitoring plus periodic BERTopic deep dives" in hospitals.The topic results can be transformed into structured interventions and evaluation indicators targeting populations,workflows,and devices,thereby forming a closed-loop management cycle of risk identification-intervention-review.

needlestick injury  /  topic modeling  /  latent Dirichlet allocation(LDA)  /  BERTopic  /  medical safety  /  incident reports  /  methodology
曹廷芬, 晏超, 钟涛, 蒋婷. 基于医院针刺伤报告的主题建模方法学研究：LDA与BERTopic用于风险识别与管理决策支持的比较. 护理研究, 2026 , 40 (17) : 3068 -3073 . DOI: 10.12102/j.issn.1009-6493.2026.17.016
Tingfen CAO, Chao YAN, Tao ZHONG, Ting JIANG. Methodological study on topic modeling based on hospital needlestick injury reports: comparison of LDA and BERTopic in risk identification and management decision support[J]. Chinese Nursing Research, 2026 , 40 (17) : 3068 -3073 . DOI: 10.12102/j.issn.1009-6493.2026.17.016
正文
收起
针刺伤是医疗领域最严峻的职业健康威胁之一[1]，其报告制度是医院安全管理的核心[2]。这些持续产生的自由文本报告构成了一个大型的“真实世界”数据，蕴含着风险发生的关键规律[3]。然而，传统人工分析方法效率低下，难以从海量文本中系统性地提炼潜在风险模式，制约了管理效能的提升[4]。主题模型为破解此难题提供了关键技术路径[5]。其中，潜在狄利克雷分布模型（latent Dirichlet allocation,LDA）作为经典算法，虽在医疗文本分析中有所应用，但其固有的“词袋”假设使其难以捕捉深层语义，导致主题连贯性与可解释性常受局限[6]。基于预训练语言模型的深度学习主题框架（BERT-based topic model，BERTopic）融合了预训练语言模型的强大语义理解能力，在通用领域展现出卓越性能，能生成语义更清晰的主题[7]。然而，在针刺伤报告这一特定领域，现有研究多为单一模型的初步应用，严重缺乏对LDA与BERTopic的系统性实证对比，尤其缺乏从主题质量、实用性、技术特性与适用场景等多维度的综合评估，无法为研究者和实践者提供清晰的模型选择依据。因此，本研究以针刺伤报告为真实世界应用场景，研究重点在于建立可复制的模型选择与落地框架：在同一数据集上系统比较LDA与BERTopic两种模型生成的主题语义质量、可解释性、行动转化能力、支持管理决策的实用价值、各自的技术特性与最佳适用场景，据此提出面向医院管理的实施策略，以期为医疗安全文本分析的模型选择提供实证参考，为推动医院安全管理向精准化、智能化发展提供参考。
1　研究方法
收起
1.1　数据来源
本研究采用回顾性研究设计，数据来源于某大型三级甲等医院2018年1月—2025年9月医院感染管理信息系统中针刺伤报告记录。纳入标准：完整的针刺伤报告记录；包含关键字段信息（操作环节、器具类型、人员类别等）。排除标准：信息严重缺失的记录（如操作环节、器具类型同时缺失）；非针刺伤类职业暴露事件。最终纳入分析的合格记录为160份。
1.2　数据预处理
为构建适用于主题模型分析的语料库，对原始文本数据进行以下标准化预处理。1）文本构建：将非结构化的关键字段（如具体操作或环节、所涉器具、人员类别等）进行拼接，融合成一条完整的文本描述，以充分保留语义信息。2）数据清洗：去除无意义的标点符号、特殊字符及数字。3）中文分词：采用Jieba分词工具库[8]进行分词，并导入自定义的医疗专业词典（如头皮钢针、静脉采血针等）以提高分词准确性。4）去停用词：使用停用词表，并辅以自定义的医疗场景停用词（如病人、进行、发生等），过滤噪声词汇[9]。5）词性过滤与词干归一化：保留名词、动词等实词，并对同义词进行归并（如采血与采集血标本），形成规范化语料。经过预处理后，最终生成用于模型训练的文档集合D={d₁,d₂,…,dₙ}。
1.3　模型简介与配置
1.3.1　LDA及其配置
LDA是一种典型的基于“词袋”假设的生成式概率主题模型[10]。它假设文档是多个主题的混合，而每个主题则是词语上的1个概率分布。LDA通过贝叶斯推断，反向求解出给定文档集合后，最可能的主题构成和词语分布。本研究基于Gensim库实现LDA模型。关键参数设置：1）主题数（K），通过计算不同K值下的主题一致性得分，并结合人工解读，确定最优主题数；2）迭代次数，设置为2 000次，以确保模型充分收敛；3）其他参数，先验参数α和η分别采用对称先验，由模型自动学习。
1.3.2　BERTopic及其配置
BERTopic是一种基于预训练语言模型的现代主题建模技术[11]。其核心分为3步：首先利用句子嵌入模型将文档转换为语义向量，然后使用聚类算法对向量进行聚类以形成主题，最后通过基于词频-逆文档频率的加权算法提取每个聚类的主题表征词。BERTopic突破了“词袋”假设，利用深度语义理解进行文档表示，能更好地捕捉上下文信息。本研究基于BERTopic库实现。关键参数设置：1）嵌入模型，选用paraphrase-multilingual-MiniLM-L12-v2多语言模型，以优化中文文本的语义向量化效果；2）降维算法，使用UMAP，参数n_neighbors=15,n_components=5,min_dist=0.0；3）聚类算法，使用HDBSCAN，参数min_cluster_size=5，以自动识别噪声点并确定主题数量；4）主题表征，使用c-TF-IDF算法从每个聚类中提取关键词。
1.4　模型评估体系
为全面比较两种模型的性能，本研究构建了一个包含内在质量、外在实用性及技术特性的综合评估体系。
1.4.1　主题质量评估
1）内在一致性：采用主题一致性得分，衡量一个主题内高分词之间的语义相似度，分数越高表明主题内部一致性越好。2）外在可解释性：邀请6名医院感染管理领域专家采用盲法对两个模型生成的主题进行独立评估。评估维度包括风险防控指导价值、关键词与主题的关联度、主题临床意义明确性。采用Likert 5级评分。
1.4.2　实用性评估
1）风险洞察深度：评估模型是否能挖掘出超越常规认知的、细粒度的风险模式（如特定器具在特定操作环节的组合风险）。2）管理行动支持度：分析模型生成的主题能否直接转化为具体、可执行的风险防控建议或培训要点。
1.4.3　技术特性评估
1）参数敏感性：对比LDA在确定主题数K与BERTopic在确定聚类参数上的复杂度和敏感性。2）自动化程度：评估从原始数据到最终主题生成流程中，所需人工干预的步骤多寡。3）计算效率：在相同硬件环境下，记录并对比两个模型的训练时间。
1.5　统计学分析
采用描述性统计分析展示数据特征，使用相关性分析探讨主题分布与临床特征的关系。所有分析使用Python 3.9完成，主要依赖库包括pandas、numpy、gensim、bertopic等。
2　结果
收起
2.1　数据基本特征
对160份有效针刺伤报告的分析显示，在人员构成上，实习护士是主要发生人群，占61.25%；其次为本院执业护士（30.63%）。在操作环节方面，拔针（16.88%）与采集血标本（15.00%）是风险发生较高的2个环节，然后是分离针头（10.00%）和针头放入锐器盒（9.37%）。在污染源类型上，血液暴露占据绝对主导地位，占88.75%。这些特征共同揭示了“实习护士”和“基础操作环节”是风险防控的关键焦点。见表1。
2.2　LDA主题发现
LDA生成了5个主题，呈现出相对均衡的风险分布。其中，“基础护理操作风险”权重（28.1%）最高，其关键词（如拔针、头皮钢针、实习护士）反映了高频常规操作中的风险。“血标本采集风险”（25.6%）和“针头处理过程风险”（22.5%）则凸显了特定操作流程中的安全隐患，关键词中均频繁出现“血液”与特定器具。整体而言，LDA从宏观上勾勒出了风险在不同操作类型和环节中的分布轮廓。具体见表2。
2.3　BERTopic主题发现
BERTopic自动识别出4个语义清晰的主题和1个噪声主题（包含30个文档，占18.75%）。其主题展现出更强的语义聚合性与场景深度，例如，“实习护士血标本采集链风险”主题将人员、操作、器具和后果（乙型肝炎）紧密关联，形成了一个完整的风险认知单元；“针头安全处理系统失效”主题则聚焦于回套针帽、分离针头等后处理环节，揭示了系统性问题。此外，模型还识别出“高感染风险病例暴露”这一具有重要临床意义的特定主题。具体见表3。
2.4　模型性能对比
客观指标显示，BERTopic的主题一致性得分（0.587分）高于LDA（0.452分），表明其生成的主题内部语义更一致。同时，BERTopic具备内置的噪声识别能力，而LDA不具备此功能，具体见表4。专家评估结果显示BERTopic与LDA相比，总体评分较高，具体见表5。在技术特性上，LDA的训练效率更高（45 s），远超BERTopic（182 s）；然而，BERTopic在参数敏感性上表现更稳定，且整个流程的自动化程度与结果的可解释性均优于LDA。具体见表6。
2.5　主题到行动：基于层级控制的风险对策映射
为保障“结果能转化为有效对策”，本研究将BERTopic识别的典型风险场景按职业安全层级控制（工程控制—管理控制—培训/行为—防护）思路进行映射，形成可直接落地的干预清单与监测指标，具体见表7。
3　讨论
收起
3.1　数据特征凸显风险防控的薄弱环节
从数据基本特征来看，实习护士群体是构成针刺伤的主要人群，占61.25%，且风险集中于拔针（16.88%）、血标本采集（15.00%）等基础环节，这一现象不能简单归因于个人经验不足，更是反映出管理者在培训、操作、防护、管理等环节上的薄弱性。人员方面，实习护士培训体系存在“重理论、轻实操、缺反馈”的现象，尤其在安全操作与应急处理方面的训练不够；器具使用方面，“非安全型”针具使用率仍然较高，说明医疗机构在安全装置的配备、培训与依从性监督均存在较大的提升空间。流程方面，操作流程（如拔针、针头分离）未充分融入防刺伤设计，关键环节缺乏标准化流程。因此，针刺伤的风险防控不仅仅是加强培训，更应该转向结构化、可落地的系统性措施改进。
3.2　LDA与BERTopic技术特性差异为临床干预提供了不同切入点
LDA基于词频统计原理[12]，将风险特征相对均衡地分布在5个主题中，呈现出较为宏观的风险图谱，有助于管理者快速了解风险分布，适合定期报告与趋势监测。如“基础护理操作风险”主题提示要定期评估常规操作中的安全隐患，应将其纳入质量监测指标体系。BERTopic凭借其深度语义理解能力，呈现出更加聚焦的风险分布模式[13]，凸显高风险链条与系统失效点，如BERTopic生成“实习护士血标本采集链风险”主题，成功地将人员类别、操作环节、器具类型和潜在后果等多个维度有机融合，形成了完整的风险认知链条，指出了要向培训—操作—防护全链条的系统性改进；“针头安全处理系统失效”主题则提示要重点优化针具处理流程，推广使用带安全装置的针具。
3.3　BERTopic的量化优势与临床实用性
从模型性能的对比结果来看，BERTopic在主题一致性、噪声识别等方面具有优势[14]，更适合根因分析（RCA）与质量改进项目，其输出的语义明确主题可直接转化为临床干预措施：针对实习护士血标本采集链风险，可通过模拟训练，强化采血全流程的安全操作与应急处理干预；针对高感染风险病例暴露，可通过建立高危病例操作预警机制干预，强化防护装备的使用与暴露后处置流程；针对噪声点中的信息不完整报告，可优化不良事件报告单设计，强制关键字段填写，提升数据质量。
3.4　LDA与BERTopic在技术特性上形成互补
在技术特性方面，两种模型呈现出明显的互补特征。LDA模型以其45 s的训练速度展现出在快速响应方面的优势，适合早期风险识别；而BERTopic的深度分析能力则适合定期做专项评估，用于制定与优化干预策略。两者结合可形成快速筛查-深度分析-精准干预-效果评估的闭环管理流程，提升安全管理的动态性与科学性。
4　建议
收起
基于以上发现，建议医疗机构从数据、分析、行动、评估4个环节构建针刺伤防控闭环。1）数据质量：优化不良事件/职业暴露报告单，设置强制字段与结构化选项，定期开展报告质量抽检与反馈；2）分层分析：日常周/月度采用LDA快速监测与预警，季度/年度采用BERTopic开展根因分析与场景化复盘；3）主题驱动干预：围绕实习护士采血链、针头后处理、特殊科室锐器交接与高感染风险暴露等主题，按层级控制优先推进工程控制（安全型器具、就近锐器盒、无接触传递）与关键流程再设计，并配套带教准入、模拟训练与稽核；4）指标评估：将安全型针具使用率、回套针帽违规率、暴露后处置时效、报告完整率等纳入质控指标，持续复盘与迭代。
5　局限性
收起
本研究存在一定的局限性：单中心的研究设计可能影响研究结果的普适性，回顾性研究特征也难以完全控制数据记录质量的差异性；未来的研究方向应通过多中心合作验证模型的通用性，以及探索结合临床预后指标的模型优化路径，进一步提升主题模型在医疗质量管理中的实用价值。
6　小结
收起
本研究通过实证比较表明，BERTopic在主题质量、语义深度与临床实用性方面优于LDA，更适合针刺伤风险的系统分析与精准干预；LDA则适合快速筛查与监测。建议医疗机构采用“深浅结合、模型互补”的应用策略，制定具体、可执行的临床干预措施，从而实现从风险识别到风险防控的闭环管理，推动医疗质量安全管理向系统化、智能化方向发展。

文献
收起
参考文献 引证文献
排序方式：
1
Mccabe F J,Dunne N,Farrington S K,et al.Needlestick injury incidence and reporting in Irish surgical trainees[J].The Surgeon,2025,23(3):133-137.
2
Zhang L H,Li Q,Guan L,et al.Prevalence and influence factors of occupational exposure to blood and body fluids in registered Chinese nurses:a national cross-sectional study[J].BMC Nursing,2022,21(1):298.
3
Homs-Romero E,Romero-Collado A,Verdú J,et al.Validity of chronic venous disease diagnoses and epidemiology using validated electronic health records from primary care:a real-world data analysis[J].Journal of Nursing Scholarship,2021,53(3):296-305.
4
Aeffner F,Wilson K,Martin N T,et al.The gold standard paradox in digital image analysis:manual versus automated scoring as ground truth[J].Archives of Pathology & Laboratory Medicine,2017,141(9):1267-1275.
5
Liu F,Panagiotakos D.Real-world data:a brief review of the methods,applications,challenges and opportunities[J].BMC Medical Research Methodology,2022,22(1):287.
6
Veugen T,Dunning V,Marcus M,et al.Secure latent dirichlet allocation[J].Frontiers in Digital Health,2025,7:1610228.
7
Tejani A S.To BERT or not to BERT:advancing non-invasive prediction of tumor biomarkers using transformer-based natural language processing(NLP)[J].European Radiology,2023,33(11):8014-8016.
8
Yang B B,Lu W,Xuan Y,et al.The influences of social support expressed from doctors and disclosed from peers on patient decision-making:an analysis from the online health community[J].Scientific Reports,2025,15:2703.
9
Lee J,Yoon W,Kim S,et al.BioBERT:a pre-trained biomedical language representation model for biomedical text mining[J].Bioinformatics,2020,36(4):1234-1240.
10
Wilcox K T,Jacobucci R,Zhang Z Y,et al.Supervised latent Dirichlet allocation with covariates:a Bayesian structural and measurement model of text and covariates[J].Psychological Methods,2023,28(5):1178-1206.
11
Miao B Y,Sushil M,Xu A,et al.Characterisation of digital therapeutic clinical trials:a systematic review with natural language processing[J].The Lancet Digital Health,2024,6(3):e222-e229.
12
Hagg L J,Merkouris S S,O'Dea G A,et al.Examining analytic practices in latent dirichlet allocation within psychological science:scoping review[J].Journal of Medical Internet Research,2022,24(11):e33166.
13
Ke Y H,Yang R,Liu N.Comparing open-access database and traditional intensive care studies using machine learning:bibliometric analysis study[J].Journal of Medical Internet Research,2024,26:e48330.
14
Ma L,Chen R,Ge W G,et al.AI-powered topic modeling:comparing LDA and BERTopic in analyzing opioid-related cardiovascular risks in women[J].Experimental Biology and Medicine,2025,250:10389.
2026年第40卷第17期
33
15
引用本文
BibTeX
文章信息
doi: 10.12102/j.issn.1009-6493.2026.17.016
  • 接收时间：2025-12-23
  • 首发时间：2026-10-30
  • 出版时间：2026-09-10
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期：2025-12-23
  • 修回日期：2026-07-14
基金
作者信息
    1贵州航天医院（护理部），贵州 563000

通讯作者:

晏超，E-mail：
YAN Chao, E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/hlyj/CN/10.12102/j.issn.1009-6493.2026.17.016
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数
科
Family		 属数
Number of
genus		 种数
Number of
species		 占总种数比例
Percentage of
total species (%)		 属
Genus		 种数
Number of
species		 占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
相关链接
论文
最新文章
热读文章
热点专题
内参
蓝皮书
产业发展报告
传播力报告
期刊分级目录
资讯
学术新闻
行业新闻
学术会议
行业会议
关于
关于我们
联系我们
北京市海淀区学院南路86号
100081
010-62662699（运营及技术支持）
qkjq@cast.org.cn

中国科学技术协会版权所有 京ICP 备10216604号-4 海淀分局备案 1101084647
科技导报社主办 中国科协信息中心技术支持