【摘要】： 背景和目的人禽流感(Human avian influenza,HAI)是人接觸禽流感病毒污染的排泄物或者分泌物而感染,并出現(xiàn)以呼吸道感染、粘膜充血等癥狀為主要表現(xiàn)的人畜共患疾病,死亡率較高。部分高致病性禽流感毒株(如H5N1)的患者可出現(xiàn)呼吸衰竭和多器官損害。2003年12月～2009年2月,人禽流感疫情相繼波及全球15個(gè)國(guó)家,涵蓋亞洲、歐洲、非洲、美洲四大洲,共計(jì)408例人禽流感病例,其中254人死亡,病死率高達(dá)62.3%。因此,有必要通過分析全球歷年來實(shí)驗(yàn)室確診人甲型流感(H5N1)病例的流行病學(xué)資料以持續(xù)監(jiān)測(cè)其發(fā)病模式是否發(fā)生改變,探索短期病例數(shù)的預(yù)測(cè)方法和影響病例死亡的可能因素;也為建立我國(guó)的監(jiān)測(cè)及預(yù)警方法打下基礎(chǔ)。 方法收集2003年12月至2009年2月全球15個(gè)國(guó)家向WHO報(bào)告的實(shí)驗(yàn)室確診人甲型流感(H5N1)病例的資料建立數(shù)據(jù)庫(kù);通過對(duì)全球人禽流感病例的逐年的年齡和性別分布、發(fā)病至住院的中位時(shí)間及發(fā)病至死亡的中位時(shí)間、家庭聚集性變化、季節(jié)趨勢(shì)和病死率等指標(biāo)進(jìn)行分析,判斷人禽流感發(fā)病模式有無變化;用灰色動(dòng)態(tài)模型和時(shí)間序列ARIMA模型對(duì)人禽流感病例月發(fā)病例數(shù)和月累計(jì)發(fā)病例數(shù)進(jìn)行預(yù)測(cè);運(yùn)用多水平logstic模型和單水平logstic回歸篩選可能影響病例死亡的因素。 結(jié)果 1.人禽流感發(fā)病模式監(jiān)測(cè) 各年齡組男女均可發(fā)病,男女性別比為0.9:1,各年份間和各年齡段間性別比差異無統(tǒng)計(jì)學(xué)意義;發(fā)病到住院的中位時(shí)間為4天(范圍0～20天),發(fā)病到死亡的中位時(shí)間為9天(2～31天),各年份間差異無統(tǒng)計(jì)學(xué)意義;發(fā)病高峰期為11月20日～3月6日,不同年份和不同國(guó)家間差異無統(tǒng)計(jì)學(xué)意義;總病死率為62.3%,且不隨時(shí)間而變化;歷年家庭聚集性也無增加趨勢(shì)。 2.人禽流感病例數(shù)預(yù)測(cè)效果 對(duì)月累計(jì)發(fā)病例數(shù)進(jìn)行預(yù)測(cè),時(shí)間序列的中位預(yù)測(cè)誤差為0.94%,灰色模型的中位預(yù)測(cè)誤差為1.5%;對(duì)月發(fā)病例數(shù)進(jìn)行預(yù)測(cè),時(shí)間序列的中位預(yù)測(cè)誤差為4.46%,灰色模型的中位預(yù)測(cè)誤差為23.36%。 3.影響人禽流感病例存活的相關(guān)因素 病例年齡每減少1歲,存活的可能性將增加1.03倍;病例發(fā)病到住院時(shí)間間隔每減少1天,病例存活可能性將增加1.20倍;有家庭聚集性的人禽流感病例,其存活的概率是無家庭聚集性病例的2.71倍。 結(jié)論 1.人禽流感發(fā)病模式監(jiān)測(cè) 全球人禽流感H5N1發(fā)病模式過去6年沒有重大改變:各年齡段男女均可發(fā)病;冬春季節(jié)高發(fā)的特征比較穩(wěn)定;歷年家庭聚集性無增加趨勢(shì)。 2.人禽流感病例數(shù)的預(yù)測(cè)效果 時(shí)間序列ARIMA模型是對(duì)人禽流感H5N1月發(fā)病數(shù)進(jìn)行預(yù)測(cè)的比較理想方法。 3.影響人禽流感病例存活的相關(guān)因素 年齡、發(fā)病到住院天數(shù)和有無家庭聚集性可能對(duì)病例的存活有影響。
[Abstract]:Background and objective Human avian influenza (Human avian influenza,HAI) is infected by human contact with excreta or secretion contaminated by avian influenza virus, and the main symptoms such as respiratory tract infection and mucous congestion are zoonotic diseases, and the mortality rate is high. Respiratory failure and multiple organ damage can occur in some patients with highly pathogenic avian influenza strains (such as H5N1). From December 2003 to February 2009, the human avian influenza epidemic spread to 15 countries around the world, covering four continents: Asia, Europe, Africa and the United States. A total of 408 cases of avian influenza were reported, of which 254 died, with a fatality rate of 62.3%. It is therefore necessary to continuously monitor changes in the pattern of influenza A through the analysis of epidemiological data from laboratories around the world to confirm cases of human influenza A (H5N1) over the years, To explore the prediction method of short-term case number and the possible factors affecting case death. It also lays a foundation for the establishment of monitoring and early warning methods in China. Methods from December 2003 to February 2009, the data of laboratory confirmed human influenza A (H5N1) cases reported to WHO in 15 countries around the world were collected to establish a database. Based on the analysis of the annual age and sex distribution of human avian influenza cases, the median time from onset to hospitalization, the median time from onset to death, the change of family aggregation, seasonal trend and mortality, To determine whether the incidence pattern of human avian influenza has changed or not. Grey dynamic model and time series ARIMA model were used to predict the number of monthly and cumulative cases of human avian influenza, and multi-level logstic model and single-level logstic regression were used to screen the factors that might affect the death of human avian influenza. Result 1. The incidence pattern of human avian influenza was monitored between men and women in all age groups, and the sex ratio of male to female was 0.9 鈮，

本文編號(hào)：2490991