某大型综合医院感染现患率分析与预测研究

目的 了解2010-2011年医院感染日现患率及其变化趋势,对2012年医院感染日现患率进行预测.方法 利用医院感染实时监控系统(RT-NISS)收集某院2010年1月1日-2011年12月31日的日现患率数据,分析其变化趋势；基于两年的数据,利用加权平均法和BP网络模型对2012年1月1日-4月30日的日现患率进行预测,并将实际值与预测值对比分析预测效果.结果 某院2010、2011年平均现患率分别为5.26％、5.53％；两年日现患率变化趋势基本一致,2月13-19日过节期间现患率最高,两年均值分别为l0.19％、13.53％,其次为10月1-7日过节期间两年均值分别为7.51％、6.98％；加权平均法预测值与实际值差异不显著,BP网络模型预测结果与实际值差异有统计学意义.结论 长假期间的医院感染现患率高于平时,加权平均法和BP网络模型均能对医院感染现患率的变化趋势进行预测,但基于两年的数据加权平均法优于BP网络模型.

Prevalence of nosocomial infections in a general hospital:a predictive study

OBJECTIVE To estimate the daily prevalence rate of nosocomial infections(NI) and variation trends and to predict the prevalence of NI in 2012.METHODS The real time nosocomial infection surveillance system(RT-NISS) was employed to collect the prevalence data from Jan 1,2010 to Dec 30,2011,the variation trend was analyzed,and weighted average method and BP network model were adopted to predict the prevalence from Jan 1,2012 to Apr 30,2012;the actual value and predicted value were compared to analyze the prediction effect.RESULTS The average prevalence rate was 5.26% in 2010,and 5.53% in 2011;the variation trend of the daily prevalence rate during the two years was basically consistent,the prevalence rate was the highest during the holiday(from 13th to 19th of Feb),the mean value of the two years were 10.19% and 13.53% respectively,followed by the holiday from Oct 1st to Oct 7TH(7.51% vs 6.98%) respectively;there was not significant difference between the actual value and the predicated value acquired by weighted average method,and the difference between the predicated value and the actual value acquired by the BP network model was statistically significant.CONCLUSION The prevalence rate of nosocomial infections is higher during the holidays than the usual time,both the weighted average method and BP network model can be used to predict the variation of prevalence of nosocomial infections,as compared the data of the two years,the weighted average method is superior to the BP network model.