Masking in Active Comparator Designs in Pharmacovigilance: A Retrospective Bias Analysis on the Spontaneous Reporting of Thiazolidinediones and Cardiovascular Events
Bibliographic record
Abstract
INTRODUCTION: Masking is a reporting bias where drug safety signals are muffled by elevated reporting of other medications in spontaneous reporting databases. While the impact of masking is often limited, its effect when using restricted designs, such as active comparators, can be consequential. METHODS: We used data from the US Food and Drugs Administration Adverse Event Reporting System (1999Q3-2013Q3) to study masking in a real-world example. Rosiglitazone, a thiazolidinedione with elevated reporting after safety concerns over cardiovascular risks, was the masking candidate. We hypothesized that stimulated reporting masked signals for another thiazolidinedione, pioglitazone. We computed estimates of proportional reporting ratios and information components, using the Bayesian confidence propagation neural network, for pioglitazone-myocardial infarction and pioglitazone-cardiac failure under unrestricted and active comparator designs, with and without the mask, before (1999Q3-2007Q1) and after (2007Q2-2013Q3) safety concerns. Relative change-in-estimates were computed to compare results with and without rosiglitazone. RESULTS: From 1999Q3-2007Q1, relative change-in-estimates of proportional reporting ratio for pioglitazone-myocardial infarction was 0.00 in unrestricted design and 0.10 in active comparator, and for pioglitazone-cardiac failure, the change was 0.01 and 0.62, respectively. From 2007Q2-2013Q3, relative change-in-estimates for pioglitazone-myocardial infarction was 0.41 in unrestricted design and 18.00 in active comparator; the change for pioglitazone-cardiac failure was 0.04 and 1.03, respectively. Relative changes in estimates of information component mirrored these trends. CONCLUSIONS: Masking can influence signal detection in active comparator designs where external events impact reporting rates in reference sets. Evaluating masking in related contexts is essential for drug safety monitoring and resource allocation for follow-up studies.
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 | 0.249 | 0.432 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.010 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
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".