Improving Age- and Cause-Specific Under-Five Mortality Rates (ACSU5MR) by Systematically Accounting Measurement Errors to Inform Child Survival Decision Making in Low Income Countries

项目摘要 据估计，2020年全球有500万儿童在5岁前死亡。为了提高儿童存活率，美国 政府和国际社会投资制定、评估和实施《行动纲领》-- 有针对性的、针对疾病的拯救生命的儿童干预措施，如疟疾疫苗或阿奇霉素 解决五岁以下儿童死亡的主要原因，包括疟疾、腹泻、肺炎和脑膜炎。例程 及时估计特定年龄和原因的五岁以下儿童死亡率(ACSU5M)对于理解 五岁以下儿童死亡原因的异质性和评估儿童生存政策和 计划的有效性。ACSU5M估计任务精度远远超过有效目标所需的精度 成年人的政策和计划，但经验性数据很少。人口学和流行病学证据 得出的结论是，儿童死亡原因在1-59个月期间并不一致。国家经验主义 低于1-59个月的特异度水平的数据通常在资源匮乏、民用有限的环境中不可用 注册系统。这些数据和估计具有相当大的科学价值，可以为这一发展提供信息 以及针对特定年龄的儿童干预措施及其扩大的影响评估。之前的研究已经 有四个主要缺点：(I)使用定制收集的数据来了解单一的年龄动态 原因；(2)仅在广泛的年龄组中估计ACSU5M；(3)忽略由经验引起的不确定性 测量ACSU5M，例如来自常规住户调查的流行率测量误差；和(4) 未能解决数据收集战略中的成本效益问题。我们利用一支拥有广泛经验的团队 在特定原因和五岁以下死亡率测量和估计方面的经验提出了一系列 旨在针对这些缺陷，通过具体评估和核算测量误差来改进 低收入国家的ACSU5M估计。我们的提案通过以下方式评估数据收集策略 验证研究、焦点小组讨论和分组随机试验，并开发最先进的 统计方法，以改进ACSU5M估计的投入和背后的方法。我们的 统计工作建立在我们正在进行的NICHD R21HD095451基础上，以开发灵活的贝叶斯模型 使用部分注册数据合并了多个不确定源。与全国范围内的伙伴关系 莫桑比克的死亡监测行动(COMSA-莫桑比克)和孟加拉卫生的MatLab 和人口监测系统(HDSS)为评估和创新数据提供基础设施 收集战略、用于方法开发的高质量数据以及用于传播的目标终端用户。如果 如果成功，拟议的研究将进一步提高对ACSU5M测量误差的理解 源自主要的数据收集策略，并显著推进ACSU5M评估，以系统化 解决数据稀缺问题，以便为决策提供信息，以改善低收入国家儿童的生存。

Project Summary An estimated 5.0 million children died before age 5 years globally in 2020. To improve child survival, the US government and international community invest in the development, evaluation and implementation of age- targeted, disease-specific life-saving childhood interventions, such as a malaria vaccine or azithromycin to address leading causes of under-five mortality including malaria, diarrhea, pneumonia and meningitis. Routine and timely estimates of age-and cause-specific under-five mortality (ACSU5M) are critical for understanding heterogeneity in causes of deaths within the under-five window and evaluating child survival policy and program effectiveness. ACSU5M estimates mandate precision well beyond what’s required to effectively target policies and programs in adults yet empirical data are scarce. Demographic and epidemiological evidence amounts to the conclusion that child cause of death is not uniform in the 1-59-month period. National empirical data at levels of specificity below 1-59 months are often not available in low resource settings with limited civil registration systems. Such data and estimates bear considerable scientific value to inform the development and impact evaluation of age-specific childhood interventions and their scale-up. Previous research has suffered from four main drawbacks: (i) using custom-collected data to understand age dynamics in a single cause; (ii) estimating ACSU5M only in broad age groups; (iii) ignoring uncertainty that arises from the empirical measurements of ACSU5M, such as prevalence measurement errors from routine household surveys; and (iv) failing to address cost effectiveness in data collection strategies. We leverage a team with extensive experience in both cause-specific and under-five mortality measurement and estimation to propose a series of Aims targeted at these drawbacks by specifically assessing and accounting for measurement errors to improve ACSU5M estimation in low-income countries. Our proposal evaluates data collection strategies through validation studies, focus group discussions and cluster randomized trials, and develops state-of-the-art statistical methodology to improve both the inputs into and the methodology behind ACSU5M estimation. Our statistical work builds on our ongoing NICHD R21HD095451 to develop a flexible Bayesian model which incorporates multiple sources of uncertainty using partial registration data. Partnerships with Country wide Mortality Surveillance for Action in Mozambique (COMSA-Mozambique) and the Matlab, Bangladesh Health and Demographic Surveillance System (HDSS) provide both infrastructure to evaluate and innovate on data collection strategies, high quality data for methodology development, and target end users for dissemination. If successful, the proposed study will further improve understanding of measurement errors in ACSU5M originated from major data collection strategies and significantly advance ACSU5M estimation to systematically address data scarcity issue so as to inform decision making to better child survival in low-income countries.