CT.gov Design-Purpose Hiddenness
2026-03-29 | full-registry ct.gov audit | plots, figures, and e156 bundle
Series
IndustrySponsor ClassesPhasesStructuralVisibilityCohortsConditionsConcentrationRule ErasPubMed AuditOncologyCardiovascularMetabolicSizeGeographyPurposeDelayArchitecture
Purpose Dashboard

Design intent and missing design labels split the registry into different hiddenness regimes

Older CT.gov studies differ sharply by sponsor-entered purpose fields, and blank purpose labels are among the quietest segments in the registry.

Open dashboardRead E156 paperProtocolSeries hub
Purpose rates
Ghosts
Allocation
Treatment scale

Dashboard

Open dashboard at rootRead E156 micro-paperOpen protocolOpen series hub

Purpose labels are not just metadata; they split the registry into visibly different transparency regimes.

How to read the dashboard
Treatment no results
68.3%
Largest purpose group
NA no results
86.4%
No recorded purpose
Non-randomized ghosts
47.3%
Allocation contrast
Randomized visible
14.6%
Results plus publication
No-results by purpose
PRIMARY PURPOSE2-year no-results rate by sponsor-entered primary purposeTREATMENT68.3%PREVENTION77.1%OTHER79.1%SUPPORTIVE_CARE83.2%BASIC_SCIENCE80.4%DIAGNOSTIC80.8%HEALTH_SERVICES_RESEARCH79.2%SCREENING84.1%DEVICE_FEASIBILITY84.1%NA86.4%
The treatment bucket is large and still poor, but the unlabeled NA group is quieter still.
This page is designed to separate common high-stock purposes from smaller but more extreme purpose buckets.
Ghost protocols by purpose
PURPOSE GHOSTSGhost-protocol rate by primary purposeTREATMENT40.5%PREVENTION38.7%OTHER50.3%SUPPORTIVE_CARE47.9%BASIC_SCIENCE51.2%DIAGNOSTIC50.9%HEALTH_SERVICES_RESEARCH38.5%SCREENING46.6%DEVICE_FEASIBILITY59.4%NA59.2%
Diagnostic, basic-science, device-feasibility, and unlabeled studies sit high on the ghost metric.
The ghost view matters because a purpose bucket can have some publication rescue without looking truly transparent.
Allocation and treatment scale
ALLOCATION AND SIZEAllocation labels and treatment-size contrastsRandomized73.8%Non-randomized73.0%Allocation NA69.4%Treatment | 1-5069.6%Treatment | 1001-500048.0%
Non-randomized studies and missing allocation labels are both quieter than randomized studies, while treatment trials still improve with size.
This keeps the purpose page tied to a concrete within-purpose structural gradient rather than a flat list of labels.
Read Across Projects

Across The Series

Each project isolates a different dimension of registry opacity, but the point is the contrast between them, not a single leaderboard.

Industry
CT.gov Industry Disclosure Gap

Industry-focused missing-results stock, sponsor backlogs, and structural sparsity inside CT.gov.

Sponsor Classes
CT.gov Sponsor-Class Hiddenness

Sponsor-class comparisons on rate, stock, and structural hiddenness rather than one flattened ranking.

Phases
CT.gov Phase Reporting Gap

Phase-by-phase disclosure gaps showing how silence changes along the development pathway.

Structural
CT.gov Structural Missingness

Field-level missingness across publication links, IPD statements, descriptions, and locations.

Visibility
CT.gov Evidence Visibility Gap

Results-plus-publication visibility states showing how many older trials are fully visible, partly visible, or ghosted.

Cohorts
CT.gov Completion Cohort Debt

Completion-era reporting debt showing how older eligible cohorts drift on no-results and ghost-protocol rates.

Conditions
CT.gov Condition Hiddenness Map

Keyword-classified therapeutic-area hiddenness mapping across common condition families.

Concentration
CT.gov Sponsor Backlog Concentration

Concentration and inequality analysis showing how much unresolved stock sits inside a thin sponsor slice.

Rule Eras
CT.gov Rule-Era Reporting Gap

Policy-era comparisons across pre-FDAAA, FDAAA, and later CT.gov completion cohorts.

PubMed Audit
CT.gov Publication Undercount Audit

Sample-based external PubMed NCT audit testing how often CT.gov no-link records hide an external paper trail.

Oncology
CT.gov Oncology Hiddenness

Oncology-specific CT.gov hiddenness showing where cancer-trial stock, phases, and sponsors still go quiet.

Cardiovascular
CT.gov Cardiovascular Hiddenness

Cardiovascular CT.gov hiddenness showing how heart and vascular studies remain quiet across major phases and sponsors.

Metabolic
CT.gov Metabolic Hiddenness

Metabolic CT.gov hiddenness across obesity, diabetes, and related trial portfolios with large late-phase and NA stock.

Size
CT.gov Enrollment-Size Gap

Enrollment-size gradients showing how older small trials remain much quieter than larger registered studies.

Geography
CT.gov Geography-Scale Visibility

Site and country footprint analysis showing how larger trial geographies map onto much better public visibility.

Delay
CT.gov Completion-Delay Debt

Registration-to-completion delay analysis showing short-cycle studies carry the heaviest reporting debt.

Architecture
CT.gov Trial-Architecture Gap

Arm-count and intervention-count analysis showing simpler trial architectures are often the quietest.

Source on GitHub | Project home | Dashboard