Szirbik N B, Pelletier C, Chaussalet T
Information System Cluster, Faculty of Management and Organization, Rijksuniversiteit Groningen, Landleven 5, Postbus 800, 9700 AV, Groningen, The Netherlands.
Int J Med Inform. 2006 Sep;75(9):683-91. doi: 10.1016/j.ijmedinf.2006.04.003. Epub 2006 Jun 9.
We propose a simple methodology for heterogeneous data collection and central repository-style database design in healthcare. Our method can be used with or without other software development frameworks, and we argue that its application can save a relevant amount of implementation effort. Also, we believe that the method can be used in other fields of research, especially those that have a strong interdisciplinary nature.
The idea emerged during a healthcare research project, which consisted among others in grouping information from heterogeneous and distributed information sources. We developed this methodology by the lessons learned when we had to build a data repository, containing information about elderly patients flows in the UK's long-term care system (LTC).
We explain thoroughly those aspects that influenced the methodology building. The methodology is defined by six steps, which can be aligned with various iterative development frameworks. We describe here the alignment of our methodology with the RUP (rational unified process) framework. The methodology emphasizes current trends, as early identification of critical requirements, data modelling, close and timely interaction with users and stakeholders, ontology building, quality management, and exception handling.
Of a special interest is the ontological engineering aspect, which had the effects with the highest impact after the project. That is, it helped stakeholders to perform better collaborative negotiations that brought better solutions for the overall system investigated. An insight into the problems faced by others helps to lead the negotiators to win-win situations. We consider that this should be the social result of any project that collects data for better decision making that leads finally to enhanced global outcomes.
我们提出一种用于医疗保健领域异构数据收集和中央存储库式数据库设计的简单方法。我们的方法可以与其他软件开发框架结合使用，也可以独立使用，并且我们认为其应用可以节省大量的实施工作。此外，我们相信该方法可用于其他研究领域，尤其是那些具有很强跨学科性质的领域。
这个想法出现在一个医疗保健研究项目中，该项目包括将来自异构和分布式信息源的信息进行分组等工作。我们通过在构建一个包含英国长期护理系统（LTC）中老年患者流动信息的数据存储库时所学到的经验教训，开发了这种方法。
我们详细解释了影响该方法构建的各个方面。该方法由六个步骤定义，这些步骤可以与各种迭代开发框架相契合。我们在此描述我们的方法与RUP（理性统一过程）框架的契合情况。该方法强调当前的趋势，如关键需求的早期识别、数据建模、与用户和利益相关者的密切及时互动、本体构建、质量管理和异常处理。
特别值得关注的是本体工程方面，在项目之后它产生了影响最大的效果。也就是说，它帮助利益相关者进行更好的协作谈判，为所研究的整个系统带来了更好的解决方案。深入了解其他人所面临的问题有助于引导谈判者达成双赢局面。我们认为这应该是任何为了更好地决策而收集数据最终实现全球成果提升的项目的社会成果。
