Epidemiologic and spatial methods to improve estimates of neurological disorders from population based studies

 描述(由申请人提供)：癫痫和头痛是两种神经疾病，在低收入和中等收入国家(LMIC)对其患病率和相关因素的估计仍然很差。事实上，LMIC的神经流行病学研究容易受到关键流行病学偏差的影响，这可能会导致低估或高估假定危险因素的流行率和关联度。这些偏差产生的原因是：1)诊断试验不完善导致的疾病分类错误；2)病例的空间相关性；3)调查抽样方法。目前的研究还没有全面检查这些潜在的偏见在神经流行病学中的影响。这个拟议的项目将通过应用流行病学和空间分析方法来确定和解决对两种神经疾病的人口估计的流行病学偏差：癫痫和撒哈拉以南非洲(SSA)个人的严重慢性头痛，以弥补这一差距。寄生虫感染，特别是脑囊虫病(NCC)，以及在获得孕产妇保健机会较少的情况下发生的产科事件，已被SSA有限的研究确定为与癫痫有关的主要因素。NCC经常导致几种神经系统疾病，最常见的是癫痫、癫痫和逐渐恶化的严重慢性头痛。其他促成风险的因素可能存在，但尚未衡量，并可能对相关因素的影响程度产生偏差。拟议的研究将涉及下列目标：目标1：在估计布基纳法索癫痫和严重慢性头痛的流行率时，量化错误分类带来的偏差；目标2：检查与布基纳法索癫痫和严重慢性头痛流行率有关的个人和地理层面的因素，同时通过拟合考虑数据空间相关性的贝叶斯分层空间模型来解释错误分类错误；目标3：调查纳入调查数据的抽样机制和分层后权重是否影响乌干达各教区的癫痫患病率估计。这些方法的应用将提高对癫痫和严重慢性头痛的估计的有效性，并量化LMIC进行的神经流行病学研究中的偏差的影响。这项研究的发现将促进对癫痫和头痛的理解，并有助于在SSA进行的关于神经疾病的有限研究。

 DESCRIPTION (provided by applicant): Epilepsy and headaches are two neurological disorders for which estimates of prevalence and associated factors remain poorly described in low and middle income countries (LMIC). Indeed, neuroepidemiologic studies from LMIC are subject to key epidemiologic biases that may lead to under- or overestimates of the prevalence and of magnitude of association of putative risk factors. These biases arise from failure to account for: 1) misclassification error of the disease due to imperfect diagnostic tests; 2) the spatial correlation of cases and 3) survey sampling methods. Current research has not comprehensively examined the impact of these potential biases in neuroepidemiology. This proposed project will address this gap by applying epidemiologic and spatial analytic methods to identify and address these epidemiologic biases for population estimates on two neurological disorders: epilepsy and severe chronic headaches among individuals in sub-Saharan Africa (SSA). Parasitic infections, particularly neurocysticercosis (NCC), and obstetric events in the context of low access to maternal health care have been identified as leading factors associated with epilepsy by the limited research in SSA. NCC often leads to several neurological disorders with the most common being seizures, epilepsy and progressively worsening severe chronic headaches. Additional contributing risk factors likely exist but have not been measured and may bias the magnitude of effect of the associated factors. The proposed study will address the following aims: Aim 1: To quantify the bias introduced by misclassification error when estimating the prevalence of epilepsy and severe chronic headaches in Burkina Faso; Aim 2: To examine the individual- and geographic level factors associated with the prevalence of epilepsy and severe chronic headaches in Burkina Faso while accounting for misclassification error by fitting a Bayesian hierarchical spatial model that allows for spatial correlation of the data; Aim 3: To investigate whether incorporating the sampling mechanism and post-stratification weights from survey data affect prevalence estimates of epilepsy across parishes in Uganda. The application of these methods will improve the validity of estimates for epilepsy and severe chronic headaches and quantify the impact of the biases in neuroepidemiologic studies conducted in LMIC. The findings from this research will advance the understanding of epilepsy and headaches and contribute to the limited research on neurological conditions conducted in SSA.