数字工具与人工智能支持抗菌药物管理的系统综述
最常用工具为临床决策支持系统(58%),多数研究显示治疗适宜性改善与抗生素消耗下降,但对临床与微生物学指标的影响异质且证据质量低。
要点
- 临床决策支持系统是医院抗菌药物管理项目中最常用的数字工具。
- 数字工具多可改善处方质量指标并降低抗菌药物消耗。
- 对临床和微生物学结局的影响不一致且研究质量偏低。
核心结论
- 数字工具应用于抗菌药物管理项目可改善处方质量指标并减少抗菌药物消耗,但对临床与微生物学结局的影响仍有限。
- 需标准化指标与高质量方法学研究以确认其对管理项目结局的真实影响。
研究设计与样本
- 系统综述,遵循PRISMA方法学,检索PubMed、Scopus与Cochrane Library(2014-25年)及灰色文献。
- 由两名研究者独立完成筛选、评价与数据提取,第三名研究者解决分歧,并按研究类型评估偏倚风险与方法学质量。
- 方案已在PROSPERO注册,ID为CRD42024601221。
主要发现
- 最常用的工具为临床决策支持系统(58%),其次为电子健康记录警报、机器学习模型、远程医疗与移动应用。
- 多数研究显示治疗适宜性改善与抗生素消耗减少。
- 对临床和微生物学指标的影响呈现异质性且质量较低。
临床怎么用
- 医院抗菌药物管理项目可优先考虑部署临床决策支持系统等数字工具以改善处方质量并降低抗菌药物消耗。
- 应用数字工具时不应仅依赖其对临床与微生物学结局的改善预期,需结合本地指标持续评估。
局限性
- 纳入研究对临床和微生物学指标的影响异质性大且证据质量低。
- 原文未报告各工具效应的具体数值与合并结果,难以量化比较。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
INTRODUCTION: Antimicrobial resistance is recognized by the WHO as one of the leading global public health challenges. Antimicrobial Stewardship Programmes (ASPs) aim to promote the rational and efficient use of antimicrobials, although their implementation faces various limitations. Digital tools and artificial intelligence (AI) emerge as strategies to improve the efficiency and outcomes of these programmes. OBJECTIVE: To analyse the role of digital tools and AI in hospital-based ASP interventions and their impact on clinical, microbiological and antimicrobial use indicators. MATERIAL AND METHODS: A systematic review was conducted following PRISMA methodology. A bibliographic search was performed in PubMed, Scopus and Cochrane Library (2014-25), as well as grey literature. Articles on hospitalized patients requiring antimicrobial treatment were included according to PICOS criteria. Risk of bias and methodological quality were assessed according to the type of study. Two independent researchers conducted the selection, evaluation and data extraction. A third researcher resolved discrepancies. RESULTS: Thirty-one studies met the inclusion criteria. The most commonly used tools were Clinical Decision Support Systems (58%), followed by alerts in electronic health records, machine learning models, telemedicine and mobile applications. Most studies showed improvements in therapeutic appropriateness and reduction in antibiotic consumption. However, the impact on clinical and microbiological indicators was heterogeneous and of low quality. CONCLUSION: Digital tools applied in ASPs improve prescription quality indicators and reduce antimicrobial consumption. Nevertheless, their impact on clinical and microbiological outcomes remains limited. Standardized indicators and robust methodological studies are needed to confirm their real impact on ASP outcomes. PROTOCOL REGISTRATION: This protocol has been registered in PROSPERO with the ID: CRD42024601221.
引用
- Cortés Sánchez CJ, Salazar González F, Gómez Portolés JM, et al. The role of digital tools and artificial intelligence in supporting antimicrobial stewardship: a systematic review. J Antimicrob Chemother. 2026. PMID: 42313419. doi:10.1093/jac/dkag213
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。