Research note

We tested TrialForge's coverage of cardiorenalmetabolic assets. Here's what we found.

TrialForge Research, June 2026.

We continuously audit our own database against other datasets: company pipelines, industry reports, analyst coverage, and trial registries. And we publish what we find, including the gaps. Here is one of those audits, on cardiorenalmetabolic assets and their trials, prompted by a team in that space who wanted confidence that new assets were fully captured.

The assets: how wide is the net?

• We assembled cardiorenalmetabolic assets named across public company pipelines, industry reports, analyst coverage, conference abstracts, and pipeline scans, then audited each against our database. In total, 499 cardiorenalmetabolic assets were identified.
• 484 (~97%) of the 499 total assets were verified to be clinical stage. The rest are preclinical or discovery programs.
• 452 (~93%) of the 484 verified clinical stage have a registered ClinicalTrials.gov trial, indicating that ClinicalTrials.gov has strong coverage of the assembled cardiorenalmetabolic clinical stage assets. The outside set was verified by code against live ClinicalTrials.gov, not by asset name, so trials registered under a development code are caught (for example amycretin under NNC0519-0130 and ofirnoflast under HT-6184, both held and counted here).
  • 14 (~3%) of the 484 clinical stage assets are brand new Phase 1 programs, newly registered and may not yet be on ClinicalTrials.gov, suggesting the above coverage gap could close to ~96% coverage.
  • 18 (~4%) of the 484 clinical stage assets sit genuinely outside ClinicalTrials.gov today, registered only outside the US (China, Japan, the EU, the UK, Denmark, Australia, Argentina, Indonesia, and the Netherlands). This is the real, evidenced gap, and our fall rollout of additional registries (the WHO ICTRP, EU CTIS, ISRCTN, jRCT, the Chinese registries, and ANZCTR) is intended to close it (and we will audit it as such).
• 452 of the 452 (100%) found on ClinicalTrials.gov were also found in TrialForge, supporting the notion that TrialForge is a high fidelity mirror of ClinicalTrials.gov data (and soon to be more registries).

In short: of clinical stage assets with a ClinicalTrials.gov trial, TrialForge holds 100 percent.

Sources: Public company pipelines from 100+ PharmaCos; IQVIA Global Trends in R&D Report 2023-2026; Citeline Pharma Annual R&D Review 2023-2026; 58 four- and five-star analyst reports (e.g., Jefferies, JPMorgan, William Blair, Leerink); Claude Fable 5 web scan; ClinicalTrials.gov; TrialForge database.

The trials: how deep does it go?

Coverage at the asset level is necessary but not sufficient. What TrialForge is good at is replicating the entire ClinicalTrials.gov database with cleaner mapping so the data can be analyzed and searched more easily. To prove our high fidelity coverage of ClinicalTrials.gov, we went one level deeper, to the individual trial.

• We enumerated every registered ClinicalTrials.gov trial run by those 452 assets, across every name form (INN, brand, and development code), and deduplicated the results to 6,942 unique trials. Marketed drugs contribute their full histories, so this is a complete intervention level count, not a lead program subset. Dapagliflozin, semaglutide, and empagliflozin each bring hundreds.
• At the initial audit under a weekly mirror, TrialForge held 6,938 of the 6,942; the four newest trials, posted June 10 to 12, lagged on recency.

With daily reconciliation now live, all 6,942 are held, with a live lag of at most one day.

697 of these NCTs are shared by more than one asset; each asset carries all of its trials, so nothing is double counted. One record, NCT06736536, was dropped as a confirmed typo from transposed digits; the real trial, NCT06737536, is present.

Source: ClinicalTrials.gov; TrialForge database.

Why this matters

• In this audit, TrialForge held ~100 percent of clinical stage assets with a registered ClinicalTrials.gov trial and ~100 percent of the trials those assets have run. With daily reconciliation now live, the live lag is at most one day.
• Thus, for pipeline analysis, TrialForge captures ClinicalTrials.gov with high fidelity and recency.
• Additionally, TrialForge offers a cleaner, faster, and more analyzable version of ClinicalTrials.gov. For example, the platform maps every asset across its names (INN, code, and brand) and its conditions and therapeutic areas through leading ontologies and guidelines. This makes drug and clinical landscapes much more searchable, comparable, and quantifiable both by the database's in-app features and also for LLMs reading it.
• TrialForge, and the team behind it, is built on transparency, trust, and analytical rigor, continually pressure testing and updating the database to bring you a better version of global clinical trials data.

Where the boundary is

ClinicalTrials.gov is our base today, and the gap this audit found is small and specific: 18 up to 32 clinical stage assets outside ClinicalTrials.gov. Our development plan is to map in all available global registries by fall. Thus, TrialForge will bring you the best version of global clinical trials data possible, translated into English, mapped, and pressure tested to the same standard.