SHARE: Simplifying Human Subject Data Sharing for Alcohol Researchers

摘要 国家精神卫生研究所(NIMH)数据档案(NDA)收集并共享未识别身份的人 受试者来自NIH资助的数百个研究项目的数据，这些项目涉及许多与精神健康相关的科学 拥有合格研究人员的领域。为国家酒精研究所创建了一个新的NDA存储库 滥用和酗酒数据档案(NIAAADA)，并将作为NIAAA相关数据提交的门户 和访问权限。2019年7月发布的NIAAA数据共享政策(非-AA-19-020)要求- 从2019年开始-所有涉及人类受试者的NIAAA拨款申请必须包括 向NIAAADA提交研究数据。向NIAAADA提交的第一波申请预计将于2021年1月完成。 酒精研究人员要遵守这一政策存在许多障碍，特别是对于那些预算有限的人 和信息技术支持。为了提交研究数据，研究人员必须梳理大量和 复杂的NDA数据定义文件，通常充满了不一致和不准确，以查找测量项目 然后将他们的数据以正确的格式映射到给定数据模板中的正确字段。这 过程通常需要研究人员手动或使用复杂的脚本来操作他们的数据，这是错误的 倾向，耗时，并需要一定的技术技能。我们的最终目标是提供近乎自动化的 向NIAAADA提交酒精研究数据的程序，以便数据提交可以 由酒精研究人员用最少的IT知识和资源准确而高效地执行。 我们建议开发和验证SHARE人类研究(SHARE)平台来解决未满足的问题 需要协助酒精研究人员向NIAAADA提交研究数据。Share通过以下方式实现 提供已映射到NDA数据的预定义标准度量库的共享度量库 词典。酒精研究人员只需要选择他们想要用于研究的指标，并 让Share完成向NIAAADA提交数据的其余工作。 第二阶段的目标包括：1)平台和措施的开发。在第二阶段，我们将继续使用以用户为中心的 设计流程和最新的电子技术开发共享，我们将在这两个方面扩展到200项措施 英语和西班牙语(如果可用)。2)试点评估。期间将有4次NIAAADA提交浪潮 第二阶段为期两年。我们计划将目标1和目标2的活动与这4个提交浪潮保持一致 并在每一波中招募五个项目，以比较Share和NDA的表现。共有20名飞行员 项目将被招募并随机使用SHARE或NDA提交NIAAADA。时间和时间 将跟踪数据准备和提交的成本，并评估提交数据的准确性。我们 假设与NDA相比，Share在减少时间和成本方面的表现将明显更好 提高了数据的准确性。

Abstract The National Institute of Mental Health (NIMH) Data Archive (NDA) collects and shares de-identified human subjects data from hundreds of NIH-funded research projects across many mental health-related scientific domains with qualified researchers. A new NDA repository is created for the National Institute on Alcohol Abuse and Alcoholism Data Archive (NIAAADA) and will serve as the portal for NIAAA-related data submissions and access. The NIAAA data-sharing policy (NOT-AA-19-020) released in July of 2019, requires that— beginning in 2019—all NIAAA grant applications involving human subjects must include plans for the submission of study data to NIAAADA. The first wave of submission to NIAAADA is expected in January of 2021. Many obstacles exist for alcohol researchers to comply with this policy, especially for those with limited budget and information technology support. To submit the study data, researchers have to comb through large and complex NDA data definition files, often filled with inconsistencies and inaccuracies, to find the measure items that they need, and then map their data to the right fields in the right format in a given data template. This process often requires researchers to manipulate their data by-hand or with complicated scripts, and is error prone, time-consuming, and requires certain technical skills. Our ultimate goal is to provide a nearly-automated process for the submission of alcohol research data into the NIAAADA so that the data submission can be performed accurately and efficiently by alcohol researchers with minimum IT knowledge and resources. We propose to develop and validate the Share HumAn REsearch (SHARE) platform to address the unmet need for assisting alcohol researchers with the submission of study data to NIAAADA. SHARE does so by offering the SHARE Measure Library of pre-defined standard measures that are already mapped to NDA data dictionaries. Alcohol researchers need to only select the measures that they want to use for their studies and let SHARE do the rest for data submission to NIAAADA. Phase II aims include: 1) Platform and measure development. In Phase II we will continue using a user-centric design process and the latest e-technologies to develop SHARE, and we will extend to 200 measures in both English and Spanish (when available). 2) Pilot evaluation. There will be 4 NIAAADA submission waves during the two-year Phase II period. We plan to align both Aim 1 and Aim 2 activities to these 4 submission waves and recruit five projects in each wave to compare the performance of SHARE vs. NDA. A total of 20 pilot projects will be recruited and randomized to use either SHARE or NDA for NIAAADA submission. The time and cost for data preparation and submission will be tracked, and the accuracy of submitted data assessed. We hypothesize that compared to NDA, SHARE will perform significantly better in terms of reduced time and cost improved data accuracy.