With the increasing adoption of health information technology and the growth in the resulting electronic repositories of clinical data, the secondary use of electronic health record data has become one of the most promising approaches to enabling and speeding clinical research. Unfortunately, electronic health record data are known to suffer from significant data quality problems. Awareness of the problem of electronic health record data quality is growing, but methods for measuring data quality remain ad hoc. Clinical researchers must handle this complicated problem without systematic or validated methods. The lack of appropriate or trustworthy electronic health record data quality assessment methodology limits the validity of research performed with electronic health record data.
This dissertation documents the development of a data quality assessment framework and guideline for clinical researchers engaged in the secondary use of electronic health record data for retrospective research. Through a systematic literature review and interviews with key stakeholders, we identified core constructs of data quality, as well as priorities for future approaches to electronic health record data quality assessment. We used a data-driven approach to demonstrate that data quality is task-dependent, indicating that appropriate data quality measures must be selected, applied, and interpreted within the context of a specific study. On the basis of these results, we developed and evaluated a dynamic guideline for data quality measures in order to help researchers choose data quality measures and methods appropriately within the context of reusing electronic health record data for research.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D8RF5SS2 |
| Date | January 2015 |
| Creators | Weiskopf, Nicole Gray |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0018 seconds