中国地方病学杂志
中國地方病學雜誌
중국지방병학잡지
CHINESE JOURNAL OF ENDEMIOLOGY
2012年
1期
84-87
,共4页
布鲁杆菌病%自回归移动平均模型%预测
佈魯桿菌病%自迴歸移動平均模型%預測
포로간균병%자회귀이동평균모형%예측
Brucellosis%Autoregressive integrated moving average model%Forecasting
目的 对比分析近年来山西省与全国布鲁杆菌病(简称布病)疫情流行情况,预测布病疫情发展趋势,为布病疫情监测与防控提供参考.方法 收集山西省2006年1月至2010年12月布病疫情月报告资料和中国疾病预防控制中心2005年1月至2010年12月发布的全国布病疫情报告资料,根据布病的逐年增长量、发展速度、增长速度等指标,比较山西省和全国近年来布病疫情流行现状.利用布病疫情月报告发病人数,拟合时间序列自回归移动平均模型(ARIMA),通过对拟合效果的验证,预测山西省和全国未来2年内的布病发病状况.结果 山西省布病疫情2008年达到高峰,新发病例5397例,较2007年增加了900例,从2008年后发病人数下降,2010年下降了17.67％(906/5128)；全国布病发病人数从2006 - 2009年一直呈上升趋势,在2007-2008年迅速增长,增长速度高达39.16％(8442/21 560),虽然2010年布病的发病人数比2009年下降了2041例,但下降速度仅为5.14％(2041/37 734)；无论山西省还是全国,5-7月均为布病发病的高峰期.拟合时间序列ARIMA模型,成功地获得了山西省和全国的ARIMA模型[(1,0,1)(1,1,0)12、(1,0,1)(0,1,1)12].验证模型的拟合效果显示,两个ARIMA模型对布病发病人数预测值的动态趋势与实际情况基本一致,全部落到了预测值的95％可信限内,模型拟合效果良好.模型预测结果显示,在未来2年内,山西省布病的发病人数总体趋势基本平稳,没有出现大的变动；全国布病发病人数在2011年和2012年预计出现小幅度的回升,每年的5-7月份是布病发病的高峰期.结论 山西省布病控制措施得力,布病发病人数呈下降趋势；ARIMA模型预测布病疫情效果好,可以为下一步疫情预测与评价提供有价值的参考.
目的 對比分析近年來山西省與全國佈魯桿菌病(簡稱佈病)疫情流行情況,預測佈病疫情髮展趨勢,為佈病疫情鑑測與防控提供參攷.方法 收集山西省2006年1月至2010年12月佈病疫情月報告資料和中國疾病預防控製中心2005年1月至2010年12月髮佈的全國佈病疫情報告資料,根據佈病的逐年增長量、髮展速度、增長速度等指標,比較山西省和全國近年來佈病疫情流行現狀.利用佈病疫情月報告髮病人數,擬閤時間序列自迴歸移動平均模型(ARIMA),通過對擬閤效果的驗證,預測山西省和全國未來2年內的佈病髮病狀況.結果 山西省佈病疫情2008年達到高峰,新髮病例5397例,較2007年增加瞭900例,從2008年後髮病人數下降,2010年下降瞭17.67％(906/5128)；全國佈病髮病人數從2006 - 2009年一直呈上升趨勢,在2007-2008年迅速增長,增長速度高達39.16％(8442/21 560),雖然2010年佈病的髮病人數比2009年下降瞭2041例,但下降速度僅為5.14％(2041/37 734)；無論山西省還是全國,5-7月均為佈病髮病的高峰期.擬閤時間序列ARIMA模型,成功地穫得瞭山西省和全國的ARIMA模型[(1,0,1)(1,1,0)12、(1,0,1)(0,1,1)12].驗證模型的擬閤效果顯示,兩箇ARIMA模型對佈病髮病人數預測值的動態趨勢與實際情況基本一緻,全部落到瞭預測值的95％可信限內,模型擬閤效果良好.模型預測結果顯示,在未來2年內,山西省佈病的髮病人數總體趨勢基本平穩,沒有齣現大的變動；全國佈病髮病人數在2011年和2012年預計齣現小幅度的迴升,每年的5-7月份是佈病髮病的高峰期.結論 山西省佈病控製措施得力,佈病髮病人數呈下降趨勢；ARIMA模型預測佈病疫情效果好,可以為下一步疫情預測與評價提供有價值的參攷.
목적 대비분석근년래산서성여전국포로간균병(간칭포병)역정류행정황,예측포병역정발전추세,위포병역정감측여방공제공삼고.방법 수집산서성2006년1월지2010년12월포병역정월보고자료화중국질병예방공제중심2005년1월지2010년12월발포적전국포병역정보고자료,근거포병적축년증장량、발전속도、증장속도등지표,비교산서성화전국근년래포병역정류행현상.이용포병역정월보고발병인수,의합시간서렬자회귀이동평균모형(ARIMA),통과대의합효과적험증,예측산서성화전국미래2년내적포병발병상황.결과 산서성포병역정2008년체도고봉,신발병례5397례,교2007년증가료900례,종2008년후발병인수하강,2010년하강료17.67％(906/5128)；전국포병발병인수종2006 - 2009년일직정상승추세,재2007-2008년신속증장,증장속도고체39.16％(8442/21 560),수연2010년포병적발병인수비2009년하강료2041례,단하강속도부위5.14％(2041/37 734)；무론산서성환시전국,5-7월균위포병발병적고봉기.의합시간서렬ARIMA모형,성공지획득료산서성화전국적ARIMA모형[(1,0,1)(1,1,0)12、(1,0,1)(0,1,1)12].험증모형적의합효과현시,량개ARIMA모형대포병발병인수예측치적동태추세여실제정황기본일치,전부락도료예측치적95％가신한내,모형의합효과량호.모형예측결과현시,재미래2년내,산서성포병적발병인수총체추세기본평은,몰유출현대적변동；전국포병발병인수재2011년화2012년예계출현소폭도적회승,매년적5-7월빈시포병발병적고봉기.결론 산서성포병공제조시득력,포병발병인수정하강추세；ARIMA모형예측포병역정효과호,가이위하일보역정예측여평개제공유개치적삼고.
Objective To analyze the Brucellosis incidence and to predict the trends of the disease in Shanxi province and the national in recent years,which could provide the reference for surveillance,prevention and control of the disease.Methods Brucellosis data which was reported monthly during January 2006 and December 2010 in Shanxi province and the data released by Chinese Center for Disease Control and Prevention during January 2005 and December 2010 were collected.Several indexes,such as the annual increasing number,the development rate,growth rate and other indicators were applied to compare Shanxi province with the national Brucellosis epidemic in recent years.What's more,the seasonal autoregressive integrated moving average model (ARIMA) was fitted respectively with the data of Brucellosis incident number reported monthly,so as to predict the prevalence status in the coming two years by verifying the fitting effect.Results Brucellosis prevalence of Shanxi province reached the peak in 2008,and the incidence number was 5397,which was 900 more than 2007.From the onset of decline after 2008,the prevalence decreased by 17.67％ (906/5128) in 2010.However,national incidence of Brucellosis kept increasing before 2009 and the prevalence increased rapidly from 2007 to 2008,and the growth rate reached 39.16％ (8442/21 560).Although the number of Brucellosis fell by 2041 cases in 2010 than in 2009,the rate of decline was only 5.14％(2041/37 734).The fastigium of Brucellosis was from May to July yearly whether Shanxi province or the country.The ARIMA models of Shanxi province and the nation were ARIMA [(1,0,1)(1,1,0)12] and ARIMA[(1,0,1)(0,1,1)12],respectively,according to the incidence numbers reported monthly.The fitting effect of models showed that the predicted values of the two models were both consistent with the actual situation and all predicted values fell within the 95％ confidence limits,which depicted that they both fitted well.The predicted values depict that the incidence of Brucellosis overall trend was basically stable in Shanxi province,while the numbers in the nation would increase in a small extent in 2011 and 2012.The fastigium of Brucellosis was still from May to July yearly.Conclusions Brucellosis control measures are effective in Shanxi province,incidence of Brucellosis declining.The ARIMA model could predict the number of Brucellosis well,which can provide a valuable reference for the predication and evaluation of Brucellosis epidemic in the future.
