Abstract 17059: Re-Use of Clinical Trial Data From the NHLBI Data Repository (BioLINCC) for Patient-Level Meta-Analyses of Cardiovascular Outcomes: Challenges and Opportunities
Bibliographic record
Abstract
Introduction: An objective of the Meta-AnalyTical Interagency Group (MATIG) is to conduct patient-level meta-analyses of cardiovascular outcomes using data from publicly available repositories. We describe challenges with data re-use from a seminal trial, provide a systematic approach to identify and curate data elements for hypothesis generation, and establish stackable trials to support these analyses. Methods: We used data from the ACCORD trial to assess risk factors and their gender specific differences for the event of hospitalization or death due to heart failure (hdHF), in patients with type 2 diabetes*. We identified the data elements needed to answer the research questions, reviewed the trial protocol to verify definitions, extracted patient-level data, performed quality assessment and statistical analysis. The results showed a gender difference in the effect of intensive vs. standard glucose-lowering therapy on hdHF. To validate the findings, we sought additional trials in BioLINCC to develop a compendium for meta-analysis, and repeated these steps for each trial. Results: Challenges for reusing the ACCORD trial included access to complete patient-level data and metadata. The compendium, developed to evaluate the stackability** of data across trials, identified differences in trial designs, patient populations, study interventions, and data elements that may impact the feasibility and interpretation of meta-analysis. An example of compendium components is shown in Table 1. Conclusion: High-quality metadata facilitate re-use of trial repository data. This compendium standardizes common data elements for gender, racial and age-group specific outcome assessment in major clinical trials. It provides the framework to assess the fitness of trials for patient-level meta-analyses. Efforts are underway by MATIG to expand the compendium to include risk factors and major cardiovascular outcomes across multiple large trials for meta-analysis.
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 | 0.600 | 0.820 |
| Meta-epidemiology (narrow) | 0.003 | 0.005 |
| Meta-epidemiology (broad) | 0.007 | 0.010 |
| Bibliometrics | 0.022 | 0.029 |
| Science and technology studies | 0.002 | 0.006 |
| Scholarly communication | 0.019 | 0.013 |
| Open science | 0.012 | 0.014 |
| Research integrity | 0.004 | 0.010 |
| Insufficient payload (model declined to judge) | 0.011 | 0.005 |
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".