caCDE-QA: A Quality Assurance Platform for Cancer Study Common Data Elements

描述（由申请人提供）：领域特定的公共数据元素（CDEs）正在成为基于标准的临床研究数据存储和检索的有效方法，并已被广泛采用。例如，国家癌症研究所（NCI）基于元数据存储库的ISO/IEC 11179标准创建了癌症数据标准存储库（caDSR）。然而，癌症临床研究界面临着与CDE建模的可扩展性、治理和数据质量相关的重大挑战。特别是，缺乏健壮的、有原则的和自动化的QA算法会导致CDE内容错误，这可能对下游的CDE使用产生重大的负面影响。我们的总体目标是建立一个新的质量保证（QA）框架，以克服在公共数据元素（CDEs）建模中的错误检测、重复或相似CDEs的识别以及CDE可用性方面的方法学和计算挑战，从而为癌症临床研究生产高质量的CDE。我们提出的方法是设计、开发和评估一个被称为cde -QA的集成平台，该平台实现了一套QA工具来审计语义web框架中表示的实验性癌症研究cde，并部署了一个带有标准语义服务的QA门户网站，用于社区协作。我们的具体目标是：(1)开发一套用于验证和协调癌症研究cde的QA工具。(2)将QA工具应用于以语义web框架表示的实验性癌症研究cde的审计。我们还将通过与NCI caDSR和CIMI社区中存在的基线工具进行比较，从效率、准确性和可用性方面评估QA工具的性能。(3)部署和评估用于协作CDE审查和协调的QA门户网站。我们将协调以社区为基础的努力，征求有关癌症研究CDE发现和协调的要求，并促进公共数据元素服务（CDES）标准的规范。我们将与临床数据交换标准联盟（CDISC）和CIMI社区合作，传播和测试新开发的质量保证方法和工具。该项目将为癌症研究cde的验证和语义协调提供新的QA方法和工具。这将使高效的CDE建模和生产高质量的可重用CDE具有重要意义，这对于促进癌症临床研究数据共享和加速系统的临床结果捕获至关重要。

DESCRIPTION (provided by applicant): Domain-specific common data elements (CDEs) are emerging as an effective approach to standards-based clinical research data storage and retrieval and have been broadly adopted. For example, the National Cancer Institute (NCI) created the Cancer Data Standards Repository (caDSR) based on the ISO/IEC 11179 standard for metadata repositories. However, cancer clinical research community faces significant challenges related to scalability, governance, and data quality for CDE modeling. In particular, the lack of robust, principled and automated QA algorithms contributes to CDE content errors that can have a significant negative impact on downstream CDE uses. Our overall goal is to build a novel quality assurance (QA) framework to overcome methodological and computational challenges with respect to error detection in the modeling of common data elements (CDEs), recognition of duplicates or similar CDEs, and CDE usability, thereby producing high-quality CDEs for cancer clinical research studies. Our proposed approach is to design, develop and evaluate an integrative platform known as caCDE-QA that implements a suite of QA tools to audit experimental cancer study CDEs represented in a semantic web framework, deploying a QA web-portal with standard semantic services for community collaboration. Our specific aims are: (1) To develop a suite of QA tools for validation and harmonization of cancer study CDEs. (2) To apply the QA tools to audit experimental cancer study CDEs represented in a semantic web framework. We will also evaluate the performance of the QA tools in terms of efficiency, accuracy and usability by comparing with the baseline tools that exist in the NCI caDSR and CIMI communities. (3) To deploy and evaluate a QA web-portal for collaborative CDE review and harmonization. We will coordinate community-based efforts soliciting requirements regarding cancer study CDE discovery and harmonization and fostering a specification of the common data element services (CDES) standard. We will disseminate and test the newly developed QA methods and tools in collaboration with the Clinical Data Interchange Standards Consortium (CDISC) and CIMI Communities. This project will contribute novel QA methods and tools for validation and semantic harmonization of cancer study CDEs. This is of great significance in that it will be enabling efficient CDE modeling and producing high-quality reusable CDEs, which are critical for facilitating cancer clinical research data sharing and accelerating systematic clinical outcomes capturing.