Considerations for Creating a Restricted Data Environment with Complete Primary Care Electronic Medical Record Data
Bibliographic record
Abstract
Background: Historically, primary care databases have been limited to subsets of the full electronic medical record (EMR) data to maintain privacy. With the progression of artificial intelligence (AI) techniques (i.e., machine learning, natural language processing, and deep learning), practice-based research networks (PBRNs) have an opportunity to utilize previously difficult to access data to conduct essential primary care research and quality improvement. However, to ensure patient privacy and data security, novel infrastructure and processes are required. We describe the considerations related to accessing complete EMR data on a large-scale within a Canadian PBRN. Setting: Queen's Family Medicine Restricted Data EnviroNment (QFAMR), Department of Family Medicine (DFM), Queen's University, Canada Methods: QFAMR is a central holding repository hosted at the Centre for Advanced Computing at Queen's University. Complete, de-identified EMR records (e.g., full chart notes, PDFs, and free text) from approximately 18,000 patients from Queen's DFM can be accessed. An iterative process over 2021-2022 was used to develop QFAMR infrastructure in collaboration with Queen's DFM members and stakeholders. Results: In May 2021, the QFAMR standing research committee was established for review and approval of all potential projects. DFM members consulted with Queen's University computing, privacy, legal, and ethics experts to develop data access processes, policies and governance, agreements, and associated documents. Initial QFAMR projects involved applying and improving de-identification processes for DFM-specific full-chart notes. Five major elements were recurrent throughout the QFAMR development process: data and technology, privacy, legal documentation, decision-making frameworks, and ethics and consent. Conclusion: Overall, the development of the QFAMR has provided a secure platform to successfully access data-rich primary care EMR records without data ever leaving Queen's University. Although accessing complete primary care EMR records has certain technological, privacy, legal, and ethical considerations and challenges, QFAMR is a significant opportunity to conduct novel and innovative primary care 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 | 0.235 | 0.354 |
| Meta-epidemiology (narrow) | 0.001 | 0.002 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.005 | 0.007 |
| Science and technology studies | 0.011 | 0.010 |
| Scholarly communication | 0.026 | 0.037 |
| Open science | 0.013 | 0.031 |
| Research integrity | 0.007 | 0.008 |
| Insufficient payload (model declined to judge) | 0.017 | 0.009 |
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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".