The devil is in the details...or not? A primer on individual patient data meta-analysis
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
A systematic review is the process by which primary studies are identified, critically appraised, and interpreted according to a predefined plan to answer clinically important questions with minimal bias and random error. When the results are quantitatively combined, the review is referred to as a “meta-analysis.” Meta-analysis provides more precise estimates of treatment benefits and harms, may reveal treatment effects that would otherwise go undetected in individual trials, and provides a succinct “bottom line” for a clinical question based on the best available evidence.1 A variation of this method is individual patient data (IPD) meta-analysis where analyses are done using original data and outcomes for each person enrolled in relevant studies; patient databases from each study are combined into a single large database, and analysed using methods that account for variation both within studies and between studies. What is the difference between conventional and IPD meta-analysis? In conventional meta-analysis, aggregated data are extracted from published and unpublished reports according to a predetermined protocol. Analysis is performed by calculating a weighted average for effect (eg, relative risk) across randomised trials.2 Limitations of this approach include risk of publication bias,3 heterogeneity in trial results,4 inability to perform intention-to-treat analyses when relevant patient data are excluded or missing,5 and limited methodological quality of source studies.6 Because each randomised trial in a …
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.
Direct model labels (unvalidated)
Per-model category and study-design labels from the labeling rounds. They are machine output, unvalidated, and the disagreement between models ships as data. No study design here is MEDLINE-validated yet.
| Model arm | Categories | Study design | Confidence |
|---|---|---|---|
| gemma | no category Domain: not available · Genre: Methods About the Canadian research system: no · About a Canadian topic: no | Not applicable | low |
| gpt | no category Domain: not available · Genre: Methods About the Canadian research system: no · About a Canadian topic: no | Theoretical or conceptual | high |
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 | 0.229 | 0.128 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.005 | 0.003 |
| Bibliometrics | 0.001 | 0.007 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.009 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.014 | 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 it