Identifying adverse events: reflections on an imperfect gold standard after 20 years of patient safety research
Bibliographic record
Abstract
In ancient Roman religion, Janus was the god of gates and doorways, but also beginnings, endings, transitions, passages, time and duality. Usually depicted as having two faces, Janus looks at the past with one face and to the future with the other. Why mention Janus in an editorial about patient safety? Partly because the 20-year anniversary of To Err is Human 1 marks a transition—from the beginnings of patient safety as a fledgling field to a more mature research endeavour. Beyond this symbolism of a transition period, Janus’s past and future looking faces bear another connection to patient safety. The ‘gold standard’ research method in patient safety, record review to look for ‘adverse events’ (AEs), defined as harms from medical care, has taken two forms. The more common method, famously used in the Harvard Medical Practice Study (HMPS)2 and other studies which have emulated it,3–9 involves retrospective (‘backwards looking’) record review. An initial review looks for signs of possible harms from medical care, which, when present, trigger more detailed review to adjudicate the presence of AEs and judge the degree to which adhering to accepted standards of care could have prevented them. More recently, some investigators have conducted prospective (‘forward looking’) surveillance to identify AEs in near-real time.10–12 These forward-looking and backward-looking AE studies have succeeded in showing the scope of many safety problems. But, after 20 years of research, can we continue to use the same metric for both measuring safety and monitoring its improvement over time? Though widely attributed to the HMPS,2 retrospective record review to identify AEs originally came from a much less well-known study.13 Carried out by the California Medical Association and California Hospital Association, this study sought to explore alternate models for compensating patients harmed by their medical care. This …
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.014 | 0.015 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.006 |
| Insufficient payload (model declined to judge) | 0.001 | 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; both teacher heads 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".