Deidentifying Narrative Assessments to Facilitate Data Sharing in Medical Education
Bibliographic record
Abstract
PROBLEM: Narrative assessments are commonly incorporated into competency-based medical education programs. However, efforts to share competency-based medical education assessment data among programs to support the evaluation and improvement of assessment systems have been limited in part because of security concerns. Deidentifying assessment data mitigates these concerns, but deidentifying narrative assessments is time-consuming, resource intensive, and error prone. The authors developed and tested a tool to automate the deidentification of narrative assessments and facilitate their review. APPROACH: The authors met throughout 2021 and 2022 to iteratively design, test, and refine the deidentification algorithm and data review interface. Preliminary testing of the prototype deidentification algorithm was performed using narrative assessments from the University of Saskatchewan emergency medicine program. The algorithm's accuracy was assessed by the authors using the review interface designed for this purpose. Formal testing included 2 rounds of deidentification and review by members of the authorship team. Both the algorithm and data review interface were refined during the testing process. OUTCOMES: Authors from 3 institutions, including 3 emergency medicine programs, an anesthesia program, and a surgical program, participated in formal testing. In the final round of review, 99.4% of the narrative assessments were fully deidentified (names, nicknames, and pronouns removed). The results were comparable for each institution and specialty. The data review interface was improved with feedback obtained after each round of review and found to be intuitive. NEXT STEPS: This innovation has demonstrated viability evidence of an algorithmic approach to the deidentification of assessment narratives while reinforcing that a small number of errors are likely to persist. Future steps include the refinement of both the algorithm to improve its accuracy and the data review interface to support additional data set formats.
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 | 0.377 | 0.693 |
| Meta-epidemiology (narrow) | 0.002 | 0.003 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.011 | 0.006 |
| Science and technology studies | 0.004 | 0.006 |
| Scholarly communication | 0.013 | 0.023 |
| Open science | 0.006 | 0.015 |
| Research integrity | 0.004 | 0.006 |
| Insufficient payload (model declined to judge) | 0.009 | 0.007 |
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".