What's fair is… fair? Presenting JustEFAB, an ethical framework for operationalizing medical ethics and social justice in the integration of clinical machine learning
Bibliographic record
Abstract
The problem of algorithmic bias represents an ethical threat to the fair treatment of patients when their care involves machine learning (ML) models informing clinical decision-making. The design, development, testing, and integration of ML models therefore require a lifecycle approach to bias identification and mitigation efforts. Presently, most work focuses on the ML tool alone, neglecting the larger sociotechnical context in which these models operate. Moreover, the narrow focus on technical definitions of fairness must be integrated within the larger context of medical ethics in order to facilitate equitable care with ML. Drawing from principles of medical ethics, research ethics, feminist philosophy of science, and justice-based theories, we describe the Justice, Equity, Fairness, and Anti-Bias (JustEFAB) guideline intended to support the design, testing, validation, and clinical evaluation of ML models with respect to algorithmic fairness. This paper describes JustEFAB's development and vetting through multiple advisory groups and the lifecycle approach to addressing fairness in clinical ML tools. We present an ethical decision-making framework to support design and development, adjudication between ethical values as design choices, silent trial evaluation, and prospective clinical evaluation guided by medical ethics and social justice principles. We provide some preliminary considerations for oversight and safety to support ongoing attention to fairness issues. We envision this guideline as useful to many stakeholders, including ML developers, healthcare decision-makers, research ethics committees, regulators, and other parties who have interest in the fair and judicious use of clinical ML tools.
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 | 0.172 | 0.280 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.008 | 0.042 |
| Scholarly communication | 0.023 | 0.018 |
| Open science | 0.004 | 0.012 |
| Research integrity | 0.019 | 0.016 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".