Semiparametric analysis of the household transmission of cholera

 描述（由申请人提供）：霍乱已成为地球仪多个地点的地方性流行病，并可能导致毁灭性的地方流行病，如2008-2009年的津巴布韦和2010-2012年的海地。在过去的50年里，对霍乱传播的科学理解 强调来自环境的感染。最近，越来越多的证据表明，人与人之间的传播可能在地方性和流行性霍乱中发挥重要作用。从人类中采集的霍乱弧菌菌株通常与从环境中采集的菌株不匹配，从人类患者中排出的霍乱弧菌处于短暂的高感染状态，其感染剂量比最近未通过人类的低10-100倍。一个新出现的假设是，环境宿主和“缓慢”的人-环境-人传播途径维持霍乱宿主并引发流行病，但“快速”的人-人传播推动了流行病的爆炸性增长。如果这一假设属实，就有理由更加重视针对霍乱病例的家庭或其他密切接触群体的干预措施，以控制地方性和流行性霍乱。尽管它对公共卫生具有潜在的重要性，但这一假设的许多证据都是间接的。长期目标是检验这一假设，并了解其对霍乱控制和研究设计的影响。拟议研究的目标是重新分析纵向霍乱发病率数据，从家庭的指标病例，使用半参数回归模型传染病传播。这些模型允许估计传染性和易感性的风险比，而无需对感染个体的传染性的时间过程进行参数假设。与以往的家庭传播数据分析的统计方法相比，这些模型更强大，更灵活，更稳定的数值。这些方法也有可能纳入正在进行的几项霍乱研究中收集的病原体基因序列数据，并可能更精确地估计传播参数。这项研究是创新的，因为它应用了新的统计方法来测试公共卫生中的一个重要假设。这一点很重要，因为了解人与人之间传播的作用对于有效控制地方性和流行性霍乱至关重要。

 DESCRIPTION (provided by applicant): Cholera has become endemic in multiple sites around the globe and can cause devastating local epidemics, such as Zimbabwe in 2008-2009 and Haiti in 2010-2012. In the last 50 years, the scientific understanding of cholera transmission has emphasized infection from the environment. Recently, evidence has accumulated that person-to-person transmission may play an important role in endemic and epidemic cholera. Vibrio cholerae strains sampled from humans often do not match well with strains sampled from the environment, and Vibrio cholerae excreted from human patients is in a transient hyperinfective state with an infectious dose 10-100 times lower than that not recently passed through humans. An emerging hypothesis is that environmental reservoirs and "slow" human-environment-human transmission pathways maintain cholera reservoirs and spark epidemics but "fast" human-to-human transmission drives the explosive growth of epidemics. If true, this hypothesis would justify increased emphasis on interventions targeted to households or other close contact groups of cholera cases for the control of endemic and epidemic cholera. Despite its potential importance to public health, much of the evidence for this hypothesis is circumstantial. The long-term goal is to test this hypothesis and to understand its implications for cholera control and study design. The goal of the proposed research is to re-analyze longitudinal cholera incidence data from the households of index cases using semiparametric regression models for infectious disease transmission. These models allow the estimation of hazard ratios for infectiousness and susceptibility without making parametric assumptions about the time course of infectiousness in infected individuals. Compared to previous statistical methods for the analysis of household transmission data, these models are more robust, more flexible, and more numerically stable. These methods also have the potential to incorporate pathogen genetic sequence data, which is being collected in several ongoing cholera studies and may allow more precise estimation of transmission parameters. This research is innovative because it applies novel statistical methods to test an important hypothesis in public health. It is significant because understanding the role of person-to-person transmission will be crucial to the effective control of endemic and epidemic cholera.