The digitization of the medical record has opened the door to a rapidly increasing number of secondary uses of medical data. Epidemiology, pay-for-performance, automated decision support, and other secondary uses of medical information are all dependent on the quality existing medical data. Thanks to advances in natural language processing and information extraction technologies, the use of existing data should increase exponentially in the near future as researchers delve into the unstructured free text of the medical record for useful information. Despite this, there is empirical evidence suggesting that data captured in the context of patient care is not well-suited for unintended uses for a variety of reasons. Unfortunately, this evidence is spread widely across the medical literature and the implications and causes of poor data quality are not well understood. This wiki and it's related mailing list were created to in an effort to bring the together the research and people trying to understand these issues in order to improve the secondary use of clinical data.
The Quality of Medical Data wiki will gather and discuss antecdotal and empirical examples of issues related to the quality of medical data gathered from peer-reviewed journals, popular media, and author contributions. Issues of importance include understanding and measuring medical data quality, contributing factors, best practices, and the effects of data quality on secondary uses of medical data.
This site is created by Leonard D'Avolio, a Ph.D. Candidate in UCLA's Department of Information Studies & Medical Imaging Informatics Group, and a National Library of Medicine Medical Informatics Fellow.