Consensus Conference Follow-up: Inter-rater Reliability Assessment of the Best Evidence in Emergency Medicine (BEEM) Rater Scale, a Medical Literature Rating Tool for Emergency Physicians
Bibliographic record
Abstract
BACKGROUND: Studies published in general and specialty medical journals have the potential to improve emergency medicine (EM) practice, but there can be delayed awareness of this evidence because emergency physicians (EPs) are unlikely to read most of these journals. Also, not all published studies are intended for or ready for clinical practice application. The authors developed "Best Evidence in Emergency Medicine" (BEEM) to ameliorate these problems by searching for, identifying, appraising, and translating potentially practice-changing studies for EPs. An initial step in the BEEM process is the BEEM rater scale, a novel tool for EPs to collectively evaluate the relative clinical relevance of EM-related studies found in more than 120 journals. The BEEM rater process was designed to serve as a clinical relevance filter to identify those studies with the greatest potential to affect EM practice. Therefore, only those studies identified by BEEM raters as having the highest clinical relevance are selected for the subsequent critical appraisal process and, if found methodologically sound, are promoted as the best evidence in EM. OBJECTIVES: The primary objective was to measure inter-rater reliability (IRR) of the BEEM rater scale. Secondary objectives were to determine the minimum number of EP raters needed for the BEEM rater scale to achieve acceptable reliability and to compare performance of the scale against a previously published evidence rating system, the McMaster Online Rating of Evidence (MORE), in an EP population. METHODS: The authors electronically distributed the title, conclusion, and a PubMed link for 23 recently published studies related to EM to a volunteer group of 134 EPs. The volunteers answered two demographic questions and rated the articles using one of two randomly assigned seven-point Likert scales, the BEEM rater scale (n = 68) or the MORE scale (n = 66), over two separate administrations. The IRR of each scale was measured using generalizability theory. RESULTS: The IRR of the BEEM rater scale ranged between 0.90 (95% confidence interval [CI] = 0.86 to 0.93) to 0.92 (95% CI = 0.89 to 0.94) across administrations. Decision studies showed a minimum of 12 raters is required for acceptable reliability of the BEEM rater scale. The IRR of the MORE scale was 0.82 to 0.84. CONCLUSIONS: The BEEM rater scale is a highly reliable, single-question tool for a small number of EPs to collectively rate the relative clinical relevance within the specialty of EM of recently published studies from a variety of medical journals. It compares favorably with the MORE system because it achieves a high IRR despite simply requiring raters to read each article's title and conclusion.
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 | 0.242 | 0.462 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.005 | 0.008 |
| Bibliometrics | 0.015 | 0.009 |
| Science and technology studies | 0.004 | 0.002 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.004 | 0.006 |
| Research integrity | 0.003 | 0.005 |
| Insufficient payload (model declined to judge) | 0.005 | 0.002 |
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".