国内医院外科系统医院感染现状的循证分析

Evidence-Based Analysis of Nosocomial Infection in Surgery Department of National Hospital

目的全面收集近年国内医院外科系统医院感染文献进行数据提取与分析,以了解国内综合医院外科系统医院感染的一般性特征,为进一步建立外科系统医院感染预警数据模型提供参考。方法以医院感染、交叉感染、院内感染、预防、控制等为检索词,计算机检索CBM、CNKI、VIP、WanFang Data等数据库在2001～2009年期间发表的相关文献,主要纳入二、三级综合医院报告数据,并进行分类和Meta分析。结果共纳入64篇文献,1 990 929例住院病例。统计结果显示,该64所医院的平均医院感染率为4.46%,平均例次感染率为4.73%;内科总感染率为23.28%,外科总感染率为17.33%,但内、外科医院感染发生例次率差异无统计学意义;革兰阴性菌感染占47.71%,革兰阳性菌感染占21.31%;平均抗生素使用比例为60.59%;医院感染平均漏报率为12.42%;医院感染高发于2～5月份。结论我国现有医院外科系统医院感染现状研究大多为证据强度不高的回顾性调查和横断面调查,且报道数据分类多不全面和不一致;内、外科间的医院感染率差异无统计学意义;漏报率较高导致原始调查数据失真将影响未来预警数学模型的建立;医院感染率存在周期性变化的特点。

Objective To investigate on the epidemiologic characteristics of nosocomial infection in surgery departments of general hospitals by analyzing the data collected from documents which were published in recent years, so as to provide references for the construction of precautionary system model. Methods Applying comprehensive search strategies, we searched various electronic databases as CBM （1978 to 2008）, CNKI （1912 to 2008）, VIP （2001 to 2009） and WanFang Data （2001 to 2009）. MeSH terms and/or text words induded： nosocomial infections, cross infection, hospital infection, prevent and control. Data from top and second grade hospital were included in this analysis. Results Sixty four articles and a total of 1 990 929 inpatients were included. Results showed： average nosocomial infection rate was 4.46%; the total rates of medicine department and surgery department were 23.28% and 17.33% respectively and no significant difference was found between the two departments; the infection rates of G^- and G^＋ germ were 47.71% and 21.31% respectively; the rates of average antibiotics use was 60.59% and the rate of missing report was 12.42%. Noscomial infection was related to season change and the wave peak was from February to May. Conclusion Most of the included studies were retrospective studies and cross-sectional studies. The type of data was inconsistency and incomplete, causing weak strength of evidence. High missing rate of reports makes the precautionary model hard to build in future.