The adoption and use of Electronic Medical Records (EMRs) and Electronic Health Records (EHRs) is continuing to rise in North America. These systems contain data of varying degrees of quality, including poor quality or "dirty" data. Data entered into EMRs need to be clean or of high quality for them to be useful for a variety of reasons, including quality improvement, clinical decision support, population management, research and system management. There are two potential solutions to obtaining clean data from EMRs: data discipline and data cleansing. Data discipline focuses on ensuring that entry of data into EMRs is of high quality, while data cleansing focuses on cleaning data in the database. Clean data are necessary for healthcare providers to effectively manage chronic diseases and should lead to a reduction in the costs associated with those diseases. The objective of this paper is to compare the costs involved in implementing the two different data cleaning approaches by performing a Budget Impact Analysis (BIA) using diabetes as the exemplar in Canada. The BIA revealed that the cost to implement data discipline is $65 million whereas the cost to implement the data cleansing approach would be $21 million. Even though the cost may seem high, the cost of dirty data is even higher. Data discipline, data cleansing, or a combination of both approaches should be considered going forward.
Keywords: Electronic medical record; budget impact analysis; data quality; diabetes; diabetes management; electronic health record; quality improvement.