PDF(965 KB)
上饶市2005—2022年肾综合征出血热流行特征及发病预测
徐鹏, 赵玉芳, 刘瑶, 刘娟, 刘敏仔
安徽预防医学杂志 ›› 2025, Vol. 31 ›› Issue (2) : 137-140.
PDF(965 KB)
PDF(965 KB)
上饶市2005—2022年肾综合征出血热流行特征及发病预测
Epidemiological characteristics and incidence prediction of hemorrhagic fever with renal syndrome in Shangrao City from 2005 to 2022
目的 了解江西省上饶市肾综合征出血热(HFRS)的流行特征,建立BP神经网络模型预测HFRS的发病趋势,为其科学防控提供依据。方法 收集2005—2022年上饶市HFRS病例信息,描述HFRS的时间分布、人群分布和地区分布,计算发病率,率的比较用χ2检验。采用MATLAB 2018a 构建BP神经网络模型,建立HFRS月发病率预测模型。结果 2005—2022年上饶市HFRS年平均发病率为1.95/10万,病死率为0.22%,不同年度发病率差异有统计学意义(χ2=221.782,P<0.001)。HFRS的发病高峰为4—6月和11月至次年1月,两个时间段累计发病数分别占总数的30.18%(698/2 313)和38.74%(896/2 313)。男性、女性发病率分别为2.63/10万、1.22/10万,男性发病率高于女性,差异有统计学意义(χ2=299.270,P<0.001)。在所有病例中,年龄分布以50~69岁为主,占42.63%;职业分布以农民为主,占68.61%。发病率前5位的县(区)分别横峰县(5.41/10万)、铅山县(5.07/10万)、玉山县(3.98/10万)、广信区(3.46/10万)、弋阳县(2.68/10万)。前12个月发病率预测下一个月的发病率建立的BP神经网络模型的平均绝对百分比误差(MAPE)为25.76%。结论 上饶市HFRS发病具有季节性流行、50岁以上人群高发及部分地区聚集发病的特征。建立的BP神经网络模型可以预测上饶市HFRS的发病情况。
Objective To understand the epidemiological characteristics of hemorrhagic fever with renal syndrome (HFRS) in Shangrao City,Jiangxi Province,and to establish a BP neural network model to predict the incidence trend of HFRS,providing a basis for scientific prevention and control. Methods The clinical data of HFRS patients in Shangrao City from 2005 to 2022 were collected,the time distribution,population distribution and regional distribution of HFRS were described,and the incidence rate was calculated.Chi-square test was used to compare the rates.MATLAB 2018a was used to construct the BP neural network model and establish the prediction model for HFRS monthly incidence. Results The average annual incidence rate of HFRS in Shangrao City from 2005 to 2022 was 1.95 per 100 000,with fatality rate of 0.22%.There was statistically significant difference in incidence rate among different years (χ2=221.782,P<0.001).The peak periods for HFRS occurrence were from April to June and from November to January of the following year,with the cumulative reported cases accounting for 30.18% (698/2 313) and 38.74% (896/2 313) of the total reported cases,respectively.The incidence rate of males was 2.63 per 100 000,which was higher than 1.22 per 100 000 of females,the difference was statistically significant (χ2=299.270,P<0.001).Among all reported cases,the age distribution was mainly between 50 and 69 years old,accounting for 42.63%,and the occupational distribution was primarily farmers,accounting for 68.61%.The top 5 counties (districts) with the highest incidence rates were Hengfeng County (5.41 per 100 000),Qianshan County (5.07 per 100 000),Yushan County (3.98 per 100 000),Guangxin District (3.46 per 100 000),and Yiyang County (2.68 per 100 000).The BP neural network model was established to predict the incidence rate for the next month based on the incidence rates of the previous 12 months,with a mean absolute percentage error (MAPE) of 25.76%. Conclusion The occurrence of HFRS in Shangrao City is characterized by seasonal epidemics,high incidence in people over 50 years old,and clustering in certain areas.The established BP neural network model can predict the incidence of HFRS in Shangrao City.
Hemorrhagic fever with renal syndrome / Epidemiological characteristics / Back propagation neural network model (BP)
| [1] |
车雷, 王晓丽, 张国斌, 等. 2011—2020年沈阳市肾综合征出血热宿主动物监测分析[J]. 中华卫生杀虫药械, 2022, 28(2):169-173.
|
| [2] |
李青, 史永林, 刘红, 等. 鼠带毒指数预测肾综合征出血热发病率的探讨[J]. 安徽预防医学杂志, 2015, 21(1):8-9.
|
| [3] |
赵宛莹, 郭雪艳, 王鸿宇, 等. 山东省潍坊市2013—2020年肾综合征出血热流行特点及影响因子研究[J]. 现代预防医学, 2023, 50(14):2651-2655+2682.
|
| [4] |
王姿欢, 孙玉兰, 陈艳伟, 等. 2015—2020年北京市肾综合征出血热流行病学特征分析[J]. 中国人兽共患病学报, 2022, 38(3):256-259.
目的 描述现住址为北京市的肾综合征出血热(HFRS)病例的流行病学特征。方法 使用WPS表格录入流行病学调查数据,进行病例个案信息数据整理后,使用SPSS 20.0软件进行统计学描述。结果 2015年至2020年北京市共报告HFRS病例46例。2015年至2020年HFRS病例数(年平均发病率)分别为11例(0.06/10万人)、9例(0.04/10万人)、7例(0.03/10万人)、8例(0.04/10万人)、9例(0.04/10万人)和2例(0.01/10万人)。男女性别比为4.7∶1。年龄介于23~69岁,中位数为42岁,职业人数最多的为干部职员(11人)。发病-就诊、就诊-住院、就诊-诊断时间间隔中位数分别为为3 d、1 d和3 d。22例(47.8%)在发病前2个月内有外出(或旅游)史。结论 2015年至2020年北京市HFRS散发,男性青壮年为主,全年均有发病,无明显地域聚集性。病例存在就诊及诊断不及时现象,曾有外出或旅游史者占比较高。
|
| [5] |
宋宁娟, 谢昀, 丁晟, 等. 江西省2013—2018年肾综合征出血热空间自相关分析[J]. 中华疾病控制杂志, 2020, 24(8):876-880.
|
| [6] |
郝启迪, 米莎莎, 李凡卡. 利用ARIMA和BP神经网络模型分析新疆生产建设兵团乙肝流行特征[J]. 现代预防医学, 2021, 48(18):3265-3269+3284.
|
| [7] |
李剑, 刘莉红, 何志勇, 等. BP神经网络和SARIMA模型在手足口病发病率预测中的比较[J]. 安徽预防医学杂志, 2023, 29(6):454-458+496.
|
| [8] |
刘天, 姚梦雷, 侯清波, 等. SARIMA、GAM和LSTM在肾综合征出血热预测中的应用效果比较. 中华地方病学杂志, 2022, 41(9):709-714.
|
| [9] |
张天琛, 章承锋, 夏光辉, 等. 2005—2020年江西省肾综合征出血热流行特征分析[J]. 现代预防医学, 2022, 49(17):3073-3078.
|
| [10] |
宋宁娟, 谢昀, 袁辉, 等. 江西省2012—2017年肾综合征出血热疫情分析[J]. 中华疾病控制杂志, 2019, 23(6):685-689.
|
| [11] |
殷强玲. 中国流行性出血热重点流行区汉坦病毒遗传进化特征分析[D]. 北京: 中国疾病预防控制中心, 2019.
|
| [12] |
王敏, 杨辉, 余樟有, 等. 浙江省衢州市2006—2020年肾综合征出血热人群流行特征及宿主动物监测研究[J]. 中国媒介生物学及控制杂志, 2022, 33(4):480-484+509.
目的 了解浙江省衢州市肾综合征出血热（HFRS）流行特征和宿主动物构成及汉坦病毒携带情况，为制定科学防控措施提供参考依据。方法 对衢州市2006－2020年HFRS发病资料进行分析，采用夹夜法调查小兽捕获率，采集小兽肺和血进行汉坦病毒抗原和抗体检测，并分析小兽种类构成及其带病毒情况。率的比较采用χ<sup>2</sup>检验。结果 2006－2020年衢州市共报告HFRS病例720例，年平均发病率为2.13/10万，其中死亡病例2例，病死率为0.28%。发病人群以30～69岁为主，占92.08%，农民占80.14%，男女性别比为2.35∶1。发病高峰集中在每年的5－7月、10月至次年1月，以开化县年平均发病率最高（8.81/10万）。监测点室内、外小兽捕获率分别为5.57%和6.14%，二者差异有统计学意义（χ<sup>2</sup>=7.374，P=0.007），室内以褐家鼠为优势种，占41.18%，室外以黑线姬鼠为优势种，占62.97%，室内外小兽种类构成差异有统计学意义（χ<sup>2</sup>=1 343.773，Pχ<sup>2</sup>=17.260，P=0.004）；小兽血清抗体阳性率为6.71%，不同种类小兽抗体检测阳性率差异有统计学意义（χ<sup>2</sup>=32.923，P结论 衢州市小兽密度及其带病毒率较高，HFRS病例以中老年人、男性、农民为主，发病高峰呈夏季及冬季双峰型，衢州市仍需加强HFRS监测、防鼠灭鼠、健康教育及疫苗接种相结合的综合防治措施。
|
| [13] |
赵棋锋, 刘明奇, 马珊珊, 等. 2006—2022年绍兴市肾综合征出血热流行特征分析[J]. 预防医学, 2023, 35(6):514-516+521.
|
| [14] |
徐雪华, 李金萍, 邓文海, 等. 山东省莱州市1984—2022年肾综合征出血热流行病学特征分析[J]. 中国农村卫生, 2023, 15(8):51-56.
|
| [15] |
郑艳泽, 赵兴, 李海朋, 等. 2011—2021年连云港市肾综合征出血热流行病学特征分析[J]. 应用预防医学, 2023, 29(1):7-10.
|
| [16] |
孙倩莱, 张苏蕾, 曾舸, 等. 2009—2019年湖南省肾综合征出血热病例及宿主动物监测分析[J]. 中国人兽共患病学报, 2021, 37(4):351-355.
目的 掌握湖南省肾综合征出血热病例的流行特征和临床特点,为防治工作提供参考依据。方法 采用描述性流行病学方法分析2009-2019年湖南省肾综合征出血热病例临床和流行病学特征以及宿主动物监测情况。结果 2009-2019年湖南省共报告肾综合征出血热病例7 001例,年均报告发病率为0.95/10万。患者以男性青壮年农民为主,各市州病例发病日期距离就诊日期中位数集中在3~5 d,就诊日期距离诊断日期中位数集中在0~2 d。病例症状以轻型、中型为主(82.45%)。临床症状和体征以发热(93.49%)、乏力(91.57%)、起病急(85.39%)等多见。相关暴露因素中,主要有鼠或鼠排泄物接触史。2009-2019年各监测点平均鼠密度为3.22%,室内鼠密度3.47% 高于野外(1.77%,χ<sup>2</sup>=648.794,P=0.000)。鼠肺标本HFRS总抗原阳性率1.75%,4个监测点HFRS的抗原阳性率差异有统计学意义(χ<sup>2</sup>=29.445,P=0.000),双峰县鼠病毒携带率最高,邵东县鼠病毒携带率最低。结论 近年来湖南省出血热疫情出现上升趋势,汉城型和汉滩型混合分布,患者临床表现不典型,就诊不够及时,需要加强健康教育工作,提高民众自我保护和及时就诊意识。
|
| [17] |
杨文姣, 肖俊玲, 丁国武. ARIMA模型和BP神经网络模型在甘肃省结核病发病率预测中的应用[J]. 中华疾病控制杂志, 2019, 23(6):728-732.
|
| [18] |
祁娟, 陈守义, 陈海燕, 等. 基于三种时间序列模型对广州市肾综合征出血热发病率的预测比较研究[J]. 医学动物防制, 2021, 37(5):485-489.
|
| [19] |
Hemorrhagic fever with renal syndrome (HFRS), caused by hantavirus, is a serious public health problem in China. Despite intensive countermeasures including Patriotic Health Campaign, rodent control and vaccination in affected areas, HFRS is still a potential public health threat in China, with more than 10,000 new cases per year. Previous epidemiological evidence suggested that meteorological factors could influence HFRS incidence, but the studies were mainly limited to a specific city or region in China. This study aims to evaluate the association between monthly HFRS cases and meteorological change at the country level using a multivariate distributed lag nonlinear model (DLNM) from 2004 to 2018. The results from both univariate and multivariate models showed a non-linear cumulative relative risk relationship between meteorological factors (with a lag of 0–6 months) such as mean temperature (Tmean), precipitation, relative humidity (RH), sunshine hour (SH), wind speed (WS) and HFRS incidence. The risk for HFRS cases increased steeply as the Tmean between − 23 and 14.79 °C, SH between 179.4 and 278.4 h and RH remaining above 69% with 50–95 mm precipitation and 1.70–2.00 m/s WS. In conclusion, meteorological factors such as Tmean and RH showed delayed-effects on the increased risk of HFRS in the study and the lag varies across climate factors. Temperature with a lag of 6 months (RR = 3.05) and precipitation with a lag of 0 months (RR = 2.08) had the greatest impact on the incidence of HFRS.
|
| [20] |
Hemorrhagic fever with renal syndrome (HFRS) is a rodent-borne infectious disease caused by hantaviruses. About 90% of global cases were reported in China. We collected monthly data on counts of HFRS cases, climatic factors (mean temperature, rainfall, and relative humidity), and vegetation (normalized difference vegetation index (NDVI)) in 109 Chinese counties from January 2002 to December 2013. First, we used a quasi-Poisson regression with a distributed lag non-linear model to assess the impacts of these four factors on HFRS in 109 counties, separately. Then we conducted a multivariate meta-analysis to pool the results at the national level. The results of our study showed that there were non-linear associations between the four factors and HFRS. Specifically, the highest risks of HFRS occurred at the 45th, 30th, 20th, and 80th percentiles (with mean and standard deviations of 10.58 ± 4.52 °C, 18.81 ± 17.82 mm, 58.61 ± 6.33%, 198.20 ± 22.23 at the 109 counties, respectively) of mean temperature, rainfall, relative humidity, and NDVI, respectively. HFRS case estimates were most sensitive to mean temperature amongst the four factors, and the lag patterns of the impacts of these factors on HFRS were heterogeneous. Our findings provide rigorous scientific support to current HFRS monitoring and the development of early warning systems.
|
/
| 〈 |
|
〉 |
