Optimal Population-Level Infection Detection Strategies for Malaria Control and Elimination in a Spatial Model of Malaria Transmission.
Optimal Population-Level Infection Detection Strategies for Malaria Control and Elimination in a Spatial Model of Malaria Transmission.
复制标题
DOI：
10.1371/journal.pcbi.1004707
复制
发表时间：
2016-01
影响因子：
4.3
通讯作者：
Wenger EA
中科院分区：
文献类型：
--
作者：
Gerardin J;Bever CA;Hamainza B;Miller JM;Eckhoff PA;Wenger EA
使用抗疟疾药物的大规模运动可能是在当地消除疟疾的有力工具，但目前的诊断技术不够敏感，无法识别所有感染个体。同时，对未感染者的过度治疗增加了加速出现耐药性和失去社区接受度的风险。传播强度的局部异质性可能使针对指数病例的运动策略能够成功地针对亚专利感染，同时限制过度治疗。虽然有选择地瞄准传播热点已被提议作为疟疾控制的一种战略，但这种瞄准尚未在消除疟疾的背景下进行试验。利用来自赞比亚南部四个卫生设施集水区调查的家庭地点、人口统计数据和流行率数据，以及基于代理的疟疾传播和免疫获得模型，根据社区年龄相关的疟疾流行情况，对每个家庭的传播强度进行了拟合。为每个集水区的每个家庭构建了一套个体感染轨迹，考虑到异质暴露和免疫。模拟和评估了各种运动策略——大规模给药、大规模筛查和治疗、局部大规模给药、雪球反应性病例检测、集中抽样和假设的血清学诊断——在发现感染、尽量减少过度治疗、减少临床病例数和阻断传播方面的表现。在疟疾控制方面，在除最高传播条件外的所有情况下，假定的治疗都导致大量过度治疗，而没有进一步降低发病率。与非靶向方法相比，药物运动选择性靶向热点是一种无效的消除工具，因为现有现场诊断的灵敏度有限。血清学诊断可能是消除疟疾的一种有效工具，但需要更高的覆盖率才能达到与大规模分发推定治疗类似的结果。全世界有数百万人面临疟疾的危险，疟疾是一种由蚊子传播的寄生虫传染病。近年来，在减少疟疾负担方面取得了巨大进展，许多区域正在制定消除战略。消灭疟疾的一种方法是用抗疟疾药物治疗大量人群，从而耗尽人类宿主体内的寄生虫库。然而，目前的现场诊断不够灵敏，无法正确识别所有受感染的个体。假定对全体人口使用抗疟疾药物将有效清除感染，但也可能导致严重的过度治疗，并鼓励寄生虫产生耐药性。我们也许能够根据他们的家庭成员和邻居是否检测呈阳性来预测哪些检测呈阴性的人实际上是被感染的。利用疟疾免疫获取的数学模型和赞比亚南部疟疾流行的空间数据集，我们模拟了识别受感染个体的策略，并比较了每种策略耗尽感染库和避免过度治疗的能力。我们根据一个地区的疟疾流行情况，对最佳战略提出不同的建议。
Mass campaigns with antimalarial drugs are potentially a powerful tool for local elimination of malaria, yet current diagnostic technologies are insufficiently sensitive to identify all individuals who harbor infections. At the same time, overtreatment of uninfected individuals increases the risk of accelerating emergence of drug resistance and losing community acceptance. Local heterogeneity in transmission intensity may allow campaign strategies that respond to index cases to successfully target subpatent infections while simultaneously limiting overtreatment. While selective targeting of hotspots of transmission has been proposed as a strategy for malaria control, such targeting has not been tested in the context of malaria elimination. Using household locations, demographics, and prevalence data from a survey of four health facility catchment areas in southern Zambia and an agent-based model of malaria transmission and immunity acquisition, a transmission intensity was fit to each household based on neighborhood age-dependent malaria prevalence. A set of individual infection trajectories was constructed for every household in each catchment area, accounting for heterogeneous exposure and immunity. Various campaign strategies—mass drug administration, mass screen and treat, focal mass drug administration, snowball reactive case detection, pooled sampling, and a hypothetical serological diagnostic—were simulated and evaluated for performance at finding infections, minimizing overtreatment, reducing clinical case counts, and interrupting transmission. For malaria control, presumptive treatment leads to substantial overtreatment without additional morbidity reduction under all but the highest transmission conditions. Compared with untargeted approaches, selective targeting of hotspots with drug campaigns is an ineffective tool for elimination due to limited sensitivity of available field diagnostics. Serological diagnosis is potentially an effective tool for malaria elimination but requires higher coverage to achieve similar results to mass distribution of presumptive treatment. Millions of people worldwide live at risk for malaria, a parasitic infectious disease transmitted by mosquitoes. Great progress has been made in reducing malaria burden in recent years, and many regions are now devising strategies for elimination. One way to eliminate malaria is to deplete the reservoir of parasites in human hosts by treating large groups of people with antimalarial drugs. However, current field diagnostics are not sensitive enough to correctly identify all infected individuals. Presumptively administering antimalarial drugs to whole populations will effectively clear infections but can also lead to substantial overtreatment and encourage the evolution of drug resistance in parasites. We might be able to predict which individuals who test negative are actually infected based on whether their household members and neighbors are testing positive. Using a mathematical model of malaria immunity acquisition and a spatial dataset of malaria prevalence in southern Zambia, we simulate strategies of identifying infected individuals and compare each strategy’s ability to deplete the infectious reservoir and avoid overtreatment. We make different recommendations for optimal strategies depending on a region’s malaria prevalence.