Methodological review to develop a list of bias items for adaptive clinical trials: Protocol and rationale
Bibliographic record
Abstract
ABSTRACT Background Randomized-clinical trials (RCTs) are the gold-standard for comparing health care interventions, but can be limited by early termination, feasibility issues, and prolonged time to trial reporting. In contrast, adaptive clinical trials (ACTs), which are defined by pre-planned modifications and analyses that occur after starting patient recruitment, are gaining popularity as they can streamline trial design and time to reporting. As adaptive methodologies continue to be adopted by researchers, it will be critical to develop a risk-of-bias tool that evaluates the unique methodological features of ACTs so that their quality can be improved and standardized for the future. In our proposed methodological review, we plan to develop a list of risk-of-bias items for ACTs to develop a candidate instrument. Methods and analysis We will perform a systematic database search to capture: (1) ACTs published in any discipline of medicine and/or surgery; and (2) studies that have proposed or reviewed items pertaining to methodological risk, bias, and/or quality in ACTs. We will perform a comprehensive search of citation databases, such as Ovid MEDLINE, EMBASE, CENTRAL, the Cochrane library, and Web of Science, in addition to multiple grey literature sources to capture published and unpublished literature related to ACTs and studies evaluating the methodological quality of ACTs. We will also search methodological registries for any risk of bias tools for ACTs. All screening and review stages will be performed in duplicate with a third senior author serving as arbitrator for any discrepancies. Included ACTs will be analyzed in a descriptive manner, and we will perform regression analysis to identify factors associated with poor reporting quality and high risk of bias. We will also perform a risk of bias assessment of ACTs using the Cochrane risk-of-bias 2.0 tool and we will assess reporting quality using the CONSORT-ACE tool. These assessments will be performed independently and in duplicate. This will be done to help generate risk of bias concepts, themes, and items that can be included in the candidate tool. For all studies of methodological quality and risk of bias, we will extract all pertinent bias items and/or tools. We will combine conceptually similar items in a descriptive manner and classify them as referring to bias or to other aspects of methodological quality, such as reporting. We will plan to generate pertinent risk of bias items and fields and finally, a candidate tool that will undergo further refinement, testing, and validation in future development stages. Ethics and dissemination This review does not require ethics approval as human subjects are not involved. As mentioned previously, this study is the first step in developing a tool to evaluate the risk of bias and methodological quality of ACTs.
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 | Theoretical or conceptual | medium |
| gpt | Metaresearch Domain: Methods · Genre: Protocol About the Canadian research system: no · About a Canadian topic: no | Systematic review | medium |
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 | 0.398 | 0.607 |
| Meta-epidemiology (narrow) | 0.008 | 0.008 |
| Meta-epidemiology (broad) | 0.016 | 0.029 |
| Bibliometrics | 0.025 | 0.024 |
| Science and technology studies | 0.006 | 0.009 |
| Scholarly communication | 0.011 | 0.011 |
| Open science | 0.010 | 0.011 |
| Research integrity | 0.016 | 0.015 |
| Insufficient payload (model declined to judge) | 0.059 | 0.022 |
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.
The models disagree on parts of this classification; every voice is preserved in the section at the end of the page.
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".