Study On Indicators And Modelling Of Risk Assessment On Malaria Re-establishment

Objectives:To develop a risk assessment indicator system for malaria re-establishment based on a framework which combines receptivity and vulnerability to malaria,to build and verify a risk assessment model for evaluating the risk of malaria re-establishment,and to classify the risk range of the re-establishment in China.This study will provide a scientific reference for surveillance and prevention of reintroduction of the parasite.Methods:The study included three parts.In the first part,the risk assessment indicator system for malaria re-establishment were developed by combining traditional risk assessment methods and an infectivity-receptivity-vulnerability method,which was recommended by the World Health Organization.On the basis of a review of the literature and a research group discussion,preliminary candidate indicators were screened.Using the Delphi method,a two-round expert consultation was conducted to evaluate the indicator system and to confirm the weight of each indicator.Thus an indicator system was built based on a basic framework including receptivity and vulnerability.The second part of the study was to build the risk re-establishment assessment model.Seventeen villages in eight towns,in the China-Myanmar border regions in the Yunnan Province,were selected to investigate the anopheline community structure.Five villages in four towns were selected to understand receptivity indicators including the species,density,human biting rate,human blood index,and parous rate of malarial vectors.In addition,the vulnerability indicators were measured,including number of imported cases,imported plasmodium spp.,and proportion of mobile population in the Yingjiang County in the Yunnan Province.A small-scale malaria re-establishment risk assessment model(SMRRAM)was built to calculate the values and degrees of receptivity and vulnerability in each village by adopting the indicator system and the weight of each indicator that were created in the first part of the study.A multiplicative-model-based malaria re-establishment index(MRI)was thus calculated to assess the risk of malaria re-establishment in the selected villages.The third part of the study was the verification and application of the risk assessment model.Based on the reported data of imported malaria cases in 2016 in China,a two-step mathematical model was developed to select key indicators,and then a large-scale malaria re-establishment risk assessment model(LMRRAM)was built to calculate the risk degree of re-establishment in each county and to classify the risk range of malaria re-establishment in China accordingly,and consequently to cover the shortage of the current risk assessment method that was only based on the species of the vectors.Results:The results of the Delphi method showed that expert positive coefficient was 100.0%.The Kendall coordination coefficient of experts' scores was 0.488 on the importance indicators(?2 = 132.698,P<0.001).A three-level indicator system for assessing malaria re-establishment was confirmed including two categories(receptivity and vulnerability),four factors(vector factor,meteorological factor,imported case factor,and mobile population factor),and sixteen indicators(species of vectors,density of vectors,human biting rate,human blood index,parous rate,monthly mean temperature,monthly precipitation,duration from entry date to illness onset date of imported cases,duration from illness onset date to date of treatment,number of imported cases,species of the parasite,categories of foci of imported cases,proportion of mobile population,the epidemic status of the areas for mobile population,the duration of living overseas,and mosquito protection measures adopted or not).The Cronbach a coefficient of the indicator system was 0.883,and the standardized coefficient was 0.904,which meant that the reliability of the system was satisfied.The result of factor analysis revealed that communalities between any two variables were 1,and the cumulative of extraction sums of squared loadings was 80.0%,which revealed that the factors could explain most the information of the variables and the efficacy of the analysis was acceptable.The weights of the sixteen indicators were 0.076,0.069,0.066,0.057,0.051,0.060,0.057,0.059,0.067,0.058,0.070,0.068,0.049,0.069,0.059,0.066,respectively.The median weights of vector indicators,imported case indicators,weather indicators,and mobile population indicators were 0.066,0.067,0.059,and 0.063,respectively.The indicator with highest weight(0.076)was species of malarial vector among vector indicators and the indicator with highest weight(0.070)was species of plasmodium spp.among imported case indicators.Over the study period,10053 Anopheles mosquitoes were collected from the eight towns,and 15 Anopheles species were identified,the most common of which were Anopheles sinensis(75.4%),An.kunmingensis(15.6%),and An.minimus(3.5%).The vectors were An.minimus,An.sinensis,or these two species mixed in the five villages.Differences of the density of the vectors were observed among villages.The density of An.minimus ranged from 0.66/light-trap/night to 4.49/light-trap/night,and the density of An.sinensis ranged from 0.60/light-trap/night to 103.54/light-trap/night.The human-biting rates of An.minimus and An.sinensis were 3.3/bait/night and 3.6/bait/night,the parous rates of them were 90.5%and 93.3%,and the human blood indices of them were 0.065 and 0.184,respectively.The mean monthly temperature ranged from 14? to 22? and monthly precipitation ranged from 183mm to 217mm in the villages.Thirty-two imported cases were identified in the 5 villages,with a 4-year average of 1 case/year(range:0-5 cases/year).The mean duration from entry date to illness onset date of imported cases ranged from 5.6 days to 12.0 days in the villages,the mean duration from illness onset date to date of treatment of the cases ranged from 1.8 days to 3.6 days,and the densities of importation ranged from 0.000 to 0.033,respectively.93.8%of the imported cases were infected with Plasmodium vivax,others were P.falciparum.The categories of foci of imported cases were foci which transmission had occurred and foci which had the conditions for the transmission.The proportion of mobile population ranged from 28.4%to 72.0%and most of them had travel history to Myanmar.The duration of living overseas ranged from 1.6 months to 11 months,and 98.0%of mobile population adopted mosquito protection measures.According to the SMRRAM model,Ka Ya He had the highest receptivity with a score of 1.641 and degree of 5,and followed by Jing Po Zhai,Hu Que Ba,and Xin Cun.Zhuan Po Zhai had the lowest receptivity with a score of 1.142 and degree of 1.Xin Cun had the highest vulnerability with the score of 1.757 and the degree of 5,and followed by Jing Po Zhai,Ka Ya He and Hu Que Ba.Zhuan Po Zhai had the lowest vulnerability with a score of 0.498 and degree of 1.Ka Ya He had the highest MRI with a score of 15,followed by Jing Po Zhai,Xin Cun,and Hu Que Ba,where MRI scores were 12,5,and 4,respectively.Zhuan Po Zhai had the lowest MRI with a score of 1.The malaria re-establishment risk was high in Jing Po Zhai and Ka Ya He,and low in Xin Cun,Zhuan Po Zhai,and Hu Que Ba.After adopting a two-step mathematical model to screen the indicators of SMRRAM model built in part two of this study,a LMRRAM model was developed including species of vector and species of the parasite to conduct the verification and application of the model on the county level in China.The results of the modelling revealed that LMRRAM was also a three-level indicator system which included two categories(receptivity and vulnerability),two factors(vector factor and imported case factor),and two indicators(species of vectors and species of the parasite).The framework,simulation method and capability of the output of LMRRAM model was the same to SMRRAM model.LMRRAM model could be successfully ran stable in the 1490 selected counties,and could be used to assess the malaria re-establishment risk in China.Among the 1490 counties,1358 had low risk level(a proportion of 91.1%),80 had medium risk level(a proportion of 5.4%),35 had high risk level(a proportion of 2.3%),13 had moderately high risk level(a proportion of 0.9%),and 4 had very high risk level(a proportion of 0.3%).There were eight provinces with high re-establishment risk counties,including Yunnan Province(14 counties),Guangxi Zhuang Autonomous Region(7 counties),Guangdong Province(5 counties),Hunan Province(4 counties),Fujian Province(2 counties),Hainan Province(1 county),Chongqing City(1 county),and Guizhou Province(1 county).Three provinces had counties with high re-establishment risk counties,including Yunnan Province(11 counties),Hainan Province(1 county),and Guangxi Zhuang Autonomous Region(1 county).Only Yunnan Province had very high re-establishment risk counties(4 counties).Conclusions:Based on the framework of receptivity and vulnerability,the indicator system for assessing malaria re-establishment is a three-level system.Vector factor and imported case factor are two key factors,and species of vectors and species of the parasite are two key indicators for the assessment of the re-establishment.LMRRAM and SMRRAM model have the same framework,simulation method and capability of the output,and suit for assessing the risk of the re-establishment in the small-scale(village)and large-scale(county)levels,respectively.The southern provinces of China(such as Yunnan Province,Guangxi Zhuang Autonomous Region,Hainan Province,and Guangdong Province)have high risk of malaria re-establishment.