质量改进策略在预防神经外科重症监护病房患者医疗保健相关感染中的有效性：一项系统评价。
完整摘要 · 中文译文
背景：医疗保健相关感染（HAIs）仍然是全球重症监护病房（ICUs）中一个主要且在很大程度上可预防的致病率、额外死亡率、住院时间延长以及医疗费用不断上升的原因。神经外科患者尤其易感，因为他们频繁接触有创设备，如脑室外引流管（EVDs）、中心静脉导管、导尿管和机械通气，同时伴有神经功能受损状态和长期ICU住院。尽管质量改进（QI）策略已在重症监护环境中广泛实施，但针对神经外科和神经重症监护人群的具体证据仍然零散。 目标：系统评估QI干预措施在降低神经外科和神经科ICU患者HAI发生率方面的有效性，并评估其对次要结局的影响，包括ICU住院时间、设备使用率和集束化干预依从性。 方法：遵循PRISMA 2020指南，于2010年1月至2026年5月间在PubMed/MEDLINE、Scopus、Embase、Cochrane图书馆和Web of Science中进行了全面检索。纳入主要证据合成的合格研究包括随机对照试验、准实验研究和观察性队列研究，这些研究评估了针对神经外科或神经科ICU人群，或可提取神经重症监护数据的混合ICU环境中，结构化QI干预措施在预防HAI方面的效果。筛选过程中发现的系统评价被保留用于背景性解读和参考。 筛选，但因存在重叠的一手数据重复，被排除在主要证据综合之外。偏倚风险根据研究设计分别采用ROBINS-I、Cochrane RoB 2和纽卡斯尔-渥太华量表进行评估。 结果：来自14个国家的17项一手研究符合纳入标准，并被纳入主要叙述性综合。3项系统评价被保留用于背景解释。结构化的质量改进（QI）干预措施在各研究中持续降低了设备相关感染率。最有力的证据来自脑室外引流（EVD）专项护理集束化方案，实施集束化方案后，EVD相关感染率从每1,000导管日14.35例下降至3.21例，其中一项研究报告在跨专业集束化方案采纳后连续20个月无感染。设备相关感染预防集束化方案也显示出具有临床意义的降低效果，包括在结构化教育干预后，呼吸机相关性肺炎（VAP）发生率从每1,000呼吸机日68.2例降至47.4例，导尿管相关尿路感染（CAUTI）发生率从每1,000导管日24.9例降至13.4例。在标准化中心静脉导管集束化方案实施后，导管相关血流感染（CLABSI）发生率显著下降，其中一项研究在持续监测下达到零感染。纳入的一手研究中偏倚风险主要为低至中等。 结论：结构化的质量改进（QI）干预措施，特别是EVD专项护理集束化方案、设备相关感染预防方案以及带依从性监测的员工教育项目，持续降低了神经外科和神经科重症监护病房中医疗保健相关感染（HAI）的发生率。 重症监护病房（ICU）环境。尽管证据基础主要为非随机研究，但研究结果在不同医疗体系和机构背景下方向一致。在神经重症监护中普遍采用基于证据的感染预防框架得到强烈支持。
完整摘要 · 英文原文
BACKGROUND: Healthcare-associated infections (HAIs) remain a major and largely preventable cause of morbidity, excess mortality, prolonged hospitalization, and escalating healthcare costs in intensive care units (ICUs) worldwide. Neurosurgical patients are particularly vulnerable because of frequent exposure to invasive devices such as external ventricular drains (EVDs), central venous catheters, urinary catheters, and mechanical ventilation, combined with impaired neurological status and prolonged ICU stays. Although quality improvement (QI) strategies have been widely implemented across critical care settings, evidence specific to neurosurgical and neurocritical care populations remains fragmented. OBJECTIVES: To systematically evaluate the effectiveness of QI interventions in reducing HAI incidence among neurosurgical and neurological ICU patients and to assess their impact on secondary outcomes including ICU length of stay, device utilization, and bundle compliance. METHODS: A comprehensive search was conducted across PubMed/MEDLINE, Scopus, Embase, the Cochrane Library, and Web of Science from January 2010 to May 2026, following PRISMA 2020 guidelines. Eligible studies for the primary evidence synthesis included randomized controlled trials, quasi-experimental studies, and observational cohort studies evaluating structured QI interventions for HAI prevention in neurosurgical or neurological ICU populations, or mixed ICU settings with extractable neurocritical care data. Systematic reviews identified during screening were retained for contextual interpretation and reference screening but were excluded from the primary evidence synthesis to avoid duplication of overlapping primary data. Risk of bias was assessed using ROBINS-I, Cochrane RoB 2, and the Newcastle-Ottawa Scale according to study design. RESULTS: Seventeen primary studies from 14 countries met eligibility criteria and were included in the primary narrative synthesis. Three systematic reviews were retained for contextual interpretation. Structured QI interventions consistently reduced device-associated infection rates across studies. The strongest evidence was observed for EVD-specific care bundles, with EVD-related infection rates decreasing from 14.35 to 3.21 per 1,000 catheter-days following bundle implementation, and one study reporting 20 consecutive infection-free months after interprofessional bundle adoption. Device-associated infection prevention bundles also demonstrated clinically meaningful reductions, including VAP rates decreasing from 68.2 to 47.4 per 1,000 ventilator-days and CAUTI rates from 24.9 to 13.4 per 1,000 catheter-days following structured educational interventions. CLABSI rates declined substantially following standardized central venous catheter bundle implementation, reaching zero in one study with sustained monitoring. Risk of bias was predominantly low to moderate across included primary studies. CONCLUSIONS: Structured QI interventions, particularly EVD-specific care bundles, device-associated infection prevention protocols, and staff education programs with compliance monitoring, consistently reduce HAI incidence in neurosurgical and neurological ICU settings. Although the evidence base is predominantly non-randomized, findings are directionally consistent across diverse healthcare systems and institutional contexts. Universal adoption of evidence-based infection prevention frameworks in neurocritical care is strongly supported.