Hospitals across your network can use the same code and mean different patients. Vercori finds that problem before your study runs.
OMOP standardizes how data is structured. It does not standardize what clinicians mean when they record a diagnosis.
One hospital requires a confirmed test before coding a condition. Another codes it on clinical judgment alone. A third applies a different threshold entirely. All use the same concept ID. All pass every standard data quality check.
But when you combine them in a study, you are combining different patient populations and you will not know it from the data.
The result is a study that looks clean and produces a number that is wrong.
Illustrative scenario · HFrEF · Based on Gluckman et al., Clinical Cardiology 2024
No software to install on your end. Each site in your network analyzes its own OMOP instance locally. No patient records go anywhere.
Each site sends Vercori a statistical summary, a semantic fingerprint, of how it defines each concept. That is all we receive.
Vercori compares fingerprints across your entire network, across five measurable dimensions per concept, and identifies where sites diverge.
Every concept is rated, every divergence is explained, every reviewer decision is recorded in a sealed audit log. Ready for your submission package.
A report for each concept in your study. Each one is rated: consistent across sites, divergent with a known explanation, or divergent and needing clinical review.
Every decision is documented by a qualified reviewer and recorded in a sealed audit log. Ready for your submission package.
Attach it to your submission. Reference it in your methods. Use it to answer the question: how do you know your sites were measuring the same thing?
A documented, tamper-evident record of every assessment, reviewer decision, and rationale. When an FDA reviewer asks how you know your sites were measuring the same thing, you hand them this.
Problems are identified before the study runs, not after results are in. At the point where they are still fixable — not after you have invested months in an analysis built on mismatched data.
A network-level picture of how every site defines every concept. A cross-site view that does not exist anywhere else. Attach it to your publication. Reference it in your methods.
Semantic mismatches discovered after results are in mean re-running the analysis, adding caveats to your findings, or defending a conclusion you are no longer confident in. Vercori finds them at the point where they are still fixable.
How do you know your sites are equivalent? Today, most teams answer: they all passed DQD. That checks structure. It does not check meaning. Vercori gives you the second answer.
When an FDA reviewer challenges your data quality narrative, you have a documented answer. Not a narrative, not an assumption. A quantified assessment for every concept in your study.
The output maps directly to what regulators expect: which concepts are consistent, which diverge, what was done about it, and who made the decision. Attach it to your evidence package. It was designed for that purpose.
Find out if your sites are actually measuring the same thing before you combine their data.
Give your sponsors documented proof that semantic consistency was checked across every site, not assumed.
Strengthen your network's research credibility with a quantified trust layer between sites.
The DQD checks whether your data is correct within a site. We check whether your sites mean the same thing. Both matter. We pick up where DQD stops.
A study that combines non-equivalent patient populations does not fail. It produces a number. The number just is not the number you think it is.
FDA guidance increasingly expects sponsors to show that data elements were captured consistently across sites. Saying all sites passed structural quality checks is no longer a complete answer.
Finding a semantic mismatch after a study has run means starting over or publishing with a caveat. Finding it before means fixing it before it matters.
FDA and EMA guidance now expects documented evidence that data sources are semantically consistent, not just structurally correct.
With over one billion patient records across 34 countries in OMOP alone, the probability that all sites define concepts identically is effectively zero.
A single post-submission data quality challenge can delay approval, trigger re-analysis, or undermine a study that took years to complete.
Cross-site coding variation in federated OMOP networks is documented, discussed, and widely acknowledged. What has been missing is a practical way to measure it, document it, and act on it before a study runs.
That is what Vercori does.
Vercori was founded by Sandra Estremera-Zink, J.D.
The name comes from two words: verdict and core. The product renders a verdict on the core concepts driving your study before the study runs.
The methodology is patent pending and has been submitted to the 2026 OHDSI Global Symposium. It is grounded in published research on cross-site concept variation in OMOP networks.
We are accepting pilot partners now. Book a demo to see the product, then let us assess one study at no cost in exchange for sharing what we find.
If you run multi-site OMOP studies, operate a network site, or advise pharma sponsors on real-world evidence, we want to hear from you.
Pilot terms
One study assessed at no cost or reduced cost. In exchange, we share what we find. If divergence is detected, findings may be co-published as a methods note.