FMEA风险评估法在ICU多重耐药菌医院感染防控中的应用

目的 采用失效模式与效应分析法（FMEA）筛选出重症监护病房（ICU）多重耐药菌高风险因素，为防控提供参考依据。方法 FMEA风险评估法筛选出ICU多重耐药菌医院感染的高风险因素，提出改进措施并落实，比较干预前后的效果。结果 筛选出的高风险事件为患者过多需加床，清洁工隔离意识薄弱，实习生和轮科生隔离意识薄弱。干预前多重耐药菌医院感染发生率为4.78%（89/1 861），干预后为2.08%（34/1 636），干预前后比较，差异有统计学（χ²=18.76，P<0.05）。干预后耐碳青霉烯类鲍曼不动杆菌医院感染发生率为0.31%，低于干预前的1.40%，差异有统计学意义（P<0.05）；耐甲氧西林金黄色葡萄球菌、产超广谱β-内酰胺酶（ESBLs）大肠埃希菌、产ESBLs肺炎克雷伯菌、耐碳青霉烯类肠杆菌的医院感染发生率干预前后比较，差异均无统计学意义（均P>0.05）。结论 FMEA可指导医院感染管理科和ICU发现多重耐药菌感染防控中的薄弱环节，有的放矢地进行预防和控制。

Application of FMEA risk assessment in prevention and control of multidrug-resistant organism healthcare-associated infection in intensive care unit

Objective To screen out high risk factors for multidrug-resistant organisms(MDROs) in intensive care unit (ICU) through failure mode and effect analysis (FMEA), and provide reference for prevention and control. Methods FMEA risk assessment method was used to screen out high risk factors for MDRO healthcare-associated infection(HAI) in ICU, improvement measures were put forward and implemented, effect before and after intervention was compared. Results The screened high-risk events were extra bed for excessive patients, weak consciousness of isolation among cleaners, interns and rotary students. Incidence of MDRO HAI was 4.78% (89/1 861) before intervention and 2.08% (34/1 636) after intervention, difference was statistically significant before and after intervention (χ²=18.76,P<0.05). Incidence of carbapenem-resistant Acinetobacter baumannii (CRAB) HAI was 0.31% after intervention, which was lower than 1.40% before intervention (P<0.05); there was no significant difference in HAI rates of methicillin-resistant Staphylococcus aureus, extended-spectrum β-lactamases (ESBLs)-producing Escherichia coli, ESBLs-producing Klebsiella pneumoniae and carbapenem-resistant Enterobacteriaceae before and after intervention (all P>0.05). Conclusion FFMEA can guide department of HAI management and ICU to find out the weak links in the prevention and control of MDRO infection, and carry out targeted prevention and control measures.