BEST PRACTICES FOR DATA QUALITY ASSURANCE FOR HOSPITAL ELECTRONIC MEDICAL RECORD RESEARCH PLATFORMS
Bibliographic record
Abstract
Background Electronic medical records (EMRs) are a rich source of clinical data across many patients. However, the data must be high quality to be used for research. There is a paucity of information on the quality of Canadian hospital EMRs for research, as well as comprehensive EMR data quality assessment checklists. Purpose This thesis aims to validate data from a leading Canadian hospital EMR then use a scoping review to develop a survey for experts to rate items for inclusion in a checklist on assessing EMR data quality for research. Methods An entity relationship diagram (ERD) for key research data was created by navigating over 20,000 data tables. Data validation was completed iteratively by manual chart review or comparing it to Canadian Institute for Health Information Discharge Abstract Database (CIHI-DAD) data for agreement. A scoping review was conducted to identify data quality checklists or frameworks potentially relevant to EMR data quality to be summarized and included in an online survey. Survey items will be rated on a 3-point scale and content validity ratios will be calculated for inclusion in the resulting expert opinion checklist. Results The ERD showed 43 tables were used for key research data. We validated data across 5 themes: Demographics, Exposures, Outcomes, Potential Confounders, and Timestamping. Most items validated with over 95% agreement, but some diagnoses for Outcomes and Potential Confounders performed poorly necessitating the use of linked CIHI-DAD data. For survey development, 533 potentially relevant checklist items were identified and summarized as 42 data quality items in the survey. Conclusions EMR data validation took many iterations to create an accurate ERD. Most key research data in the EMR had high agreement but linked coded data is required for some diagnoses. Our survey will result in a comprehensive, expert opinion checklist of EMR data quality for research.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.785 | 0.850 |
| Meta-epidemiology (narrow) | 0.003 | 0.005 |
| Meta-epidemiology (broad) | 0.006 | 0.012 |
| Bibliometrics | 0.028 | 0.033 |
| Science and technology studies | 0.011 | 0.017 |
| Scholarly communication | 0.037 | 0.024 |
| Open science | 0.016 | 0.023 |
| Research integrity | 0.011 | 0.020 |
| Insufficient payload (model declined to judge) | 0.008 | 0.008 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".