Mahmood Ahmad
Tahir Heart Institute
author@example.com

CT.gov Sponsor-Class Hiddenness

Which sponsor classes account for the biggest and worst ClinicalTrials.gov disclosure failures? We analysed 578,109 registry records captured on March 29, 2026, with particular attention to 290,524 closed interventional studies and 249,507 eligible older studies. We summarized two-year no-results gaps, structural missingness, and composite hiddenness scores by sponsor class using the full flattened study-level feature set, deliberately separating rates, stocks, and missing-field patterns instead of collapsing everything into one composite leaderboard for public interpretation and oversight. OTHER_GOV had the worst eligible two-year no-results rate at 95.7 percent, whereas OTHER held the largest absolute stock at 127,704 missing-results studies. Industry remained too large to dismiss, contributing 44,007 two-year no-results studies, while NIH had the highest average hiddenness score among named sponsor classes. The class pattern therefore changes depending on whether one prioritizes rates, absolute stock, or structural sparsity, which means a single leaderboard is misleading. These estimates capture observable registry omission rather than motive or legal culpability.

Outside Notes

Type: methods
Primary estimand: Sponsor-class comparison of eligible 2-year no-results rate and absolute missing-results stock
App: CT.gov Sponsor-Class Hiddenness dashboard
Data: Full March 29, 2026 ClinicalTrials.gov snapshot grouped by sponsor class
Code: https://github.com/mahmood726-cyber/ctgov-sponsor-class-hiddenness
Version: 1.0.0
Validation: FULL REGISTRY RUN

References

1. ClinicalTrials.gov API v2. National Library of Medicine. Accessed March 29, 2026.
2. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement. BMJ. 2021;372:n71.
3. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to Meta-Analysis. 2nd ed. Wiley; 2021.

AI Disclosure

This work represents a compiler-generated evidence micro-publication built from structured registry data and deterministic summary code. AI was used as a constrained coding and drafting assistant for interface generation, packaging, and prose refinement, not as an autonomous author. The analytical choices, interpretation, and final outputs were reviewed by the author, who takes responsibility for the content.