February 2012

Document Type


Degree Name



Dept. of Medical Informatics and Clinical Epidemiology


Oregon Health & Science University


Objective: To assess the impact of a Breast Imaging Results And Data System (BI-RADS) and electronic health record (EHR) interface on the quality of mammogram results data stored in an EHR database. Setting: Ambulatory internal medicine faculty practice affiliated with a large academic community hospital. Methods: The quality of mammogram result entries in an EHR for women between 40-65 years of age during a fifteen-month period prior to the creation of a BI-RADS-EHR interface and fifteen months after inception of the interface was analyzed. Using a list of mammogram result terms as a standard, the data quality was rated as high, medium, or low. A high-quality entry required the use of a correctly spelled standard term. Medium-quality entries used a correctly spelled abbreviated or non-standard term and low quality entries were either misspelled, blank, or correctly spelled terms not related to mammography. For null entries in the post-interface period, it was determined whether the woman had a mammogram at our facility and if she had, radiology registration data were analyzed to determine whether the primary care physician had been identified correctly. Results: There were 843 entries during the pre-interface period and 861 entries during the post-interface period. High quality entries were present in 350 (41.6%) and 552 (63.8%), medium quality in 4 (0.5%) and 1 (0.1%), low quality in 488 (58.0%) and 308 (35.8%) entries during the pre-interface and post-interface periods, respectively. Blank entries decreased from 468 (55.5%) to 308 (35.8%) after creation of the interface. In the post-interface period, in 90 of 308 (29.2%) of the blank entries, the patient had had a mammogram performed at our facility; in 100% of these 90 entries, the mammogram had been ordered by a provider from another practice. In these instances, either no primary care provider or an incorrect primary care provider was identified at the time of registration in the radiology department. Conclusions: A BIRADS-EHR interface significantly improved the quality of mammogram result entries in an electronic health record with high reliability. However, radiology registration processes ultimately determine whether a result is interfaced to a primary care physician. Assuring high quality mammogram result data in EHRs requires attention to both technical and process details.




School of Medicine



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.