Methodological review to develop a list of bias items used to assess reviews incorporating network meta-analysis: protocol and rationale
Bibliographic record
Abstract
INTRODUCTION: Systematic reviews with network meta-analysis (NMA; ie, multiple treatment comparisons, indirect comparisons) have gained popularity and grown in number due to their ability to provide comparative effectiveness of multiple treatments for the same condition. The methodological review aims to develop a list of items relating to biases in reviews with NMA. Such a list will inform a new tool to assess the risk of bias in NMAs, and potentially other reporting or quality checklists for NMAs which are being updated. METHODS AND ANALYSIS: We will include articles that present items related to bias, reporting or methodological quality, articles assessing the methodological quality of reviews with NMA, or papers presenting methods for NMAs. We will search Ovid MEDLINE, the Cochrane library and difficult to locate/unpublished literature. Once all items have been extracted, we will combine conceptually similar items, classifying them as referring to bias or to other aspects of quality (eg, reporting). When relevant, reporting items will be reworded into items related to bias in NMA review conclusions, and then reworded as signalling questions. ETHICS AND DISSEMINATION: No ethics approval was required. We plan to publish the full study open access in a peer-reviewed journal, and disseminate the findings via social media (Twitter, Facebook and author affiliated websites). Patients, healthcare providers and policy-makers need the highest quality evidence to make decisions about which treatments should be used in healthcare practice. Being able to critically appraise the findings of systematic reviews that include NMA is central to informed decision-making in patient care.
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
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 | Metaresearch Domain: Methods · Genre: Protocol About the Canadian research system: no · About a Canadian topic: no | Not applicable | low |
| gpt | Metaresearch Domain: Methods · Genre: Protocol About the Canadian research system: no · About a Canadian topic: no | Not applicable | low |
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 | 0.624 | 0.404 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.080 | 0.015 |
| Bibliometrics | 0.001 | 0.027 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.003 | 0.000 |
| Open science | 0.007 | 0.004 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.017 | 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, unvalidatedLabeled directly by 2 models reading the full record.
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".