018 A conceptual framework for predictive algorithm impact studies with focus on shared decision making
Bibliographic record
Abstract
Introduction There is a proliferation of predictive algorithm development studies but a need for real- world effectiveness evaluation. Current predictive algorithm reporting guidelines and frameworks focus on technical performance. We propose the development of a Predictive algorithm Impact Evaluation (PIE) framework to address the impact of predictive algorithms in the clinical setting while prioritizing patient-centred outcomes and shared-decision making. Methods We used a flexible approach to synthesize knowledge related to the evaluation of predictive algorithms and enable exploration and mapping of key evaluation criteria. The following steps were taken: 1) defining key concepts, 2) conducting a preliminary exploration of the literature, and 3) extracting and synthesising the data into themes. The framework was developed iteratively by a multi- disciplinary group of researchers (expertise in artificial intelligence in healthcare, implementation science, learning health systems, and patient-oriented research) and our Patient and Family Advisory Committee. Results There are limited real-world evaluative studies, but a wide range of scope, approaches, and outcomes. Preliminary domains include: uses (e.g., risk stratification, risk communication); target population (e.g., patient-centred, health systems, clinical or clinician-centred); development stage (pilot, scaling); and, cross-cutting themes (e.g., equity, data privacy, bioethics, transparency). Discussion Concepts of implementation science, learning health systems, patient-centred uses are not well-established in impact or real-world evaluative studies of predictive algorithms. The iterative, consensus driven nature of the process and involvement of diverse stakeholders, including patients and families, will help ensure a comprehensive and inclusive evaluation framework. Conclusion The imperative for evaluating predictive algorithms in clinical practice is rooted in commitment to providing safe, effective, and patient-centered healthcare. A comprehensive evaluation framework can help ensure that these algorithms contribute positively to patient and clinical decision- making while upholding ethical standards and promoting equity in healthcare.
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 | 0.170 | 0.119 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.004 |
| Bibliometrics | 0.015 | 0.011 |
| Science and technology studies | 0.005 | 0.036 |
| Scholarly communication | 0.021 | 0.024 |
| Open science | 0.007 | 0.011 |
| Research integrity | 0.006 | 0.007 |
| Insufficient payload (model declined to judge) | 0.012 | 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".