EHR to EDC integration takes data a research site already has in its electronic health record and writes it into the sponsor's electronic data capture system, so a coordinator stops retyping lab values and vital signs into case report forms. On October 5, 2026 Mount Sinai's Tisch Cancer Center said it had moved its EHR to EDC system from pilot to routine use across all 16 cancer disease groups, with about 70 percent of eCRF data now filled automatically at each visit. For a site director, the questions are practical: which of your fields can move, what you still own as the investigator, and what to ask before a sponsor or your health system switches it on.
In short
- Mount Sinai reports about 70% of the clinical trial data required on eCRFs is populated automatically at each patient visit (Mount Sinai, 2026).
- Average data entry time fell from 5.5 to 2.5 minutes per visit, with a 0% re-query rate in the data evaluated (Mount Sinai, 2026).
- A German feasibility study found only 24% of all eCRF variables automatically transferable, 36% of eligible ones (Mueller et al., JMIR, 2023).
- FDA's 2018 guidance supports EHR and EDC interoperability but leaves source record retention with the investigator (FDA, 2018).
What EHR to EDC changes for a research site
The manual way is familiar. A participant comes in, the visit is documented in the EHR, and later someone at the site opens the sponsor's EDC and types the same blood pressure, the same lab panel and the same medication list into the eCRF. Mount Sinai's release puts it plainly: "Traditional clinical trial workflows often require research teams to manually enter and transfer information from the EHR into separate EDC systems." Every transcription error becomes a query, and every query comes back to the coordinator.
With an integration, structured EHR data is mapped to eCRF fields and sent across. Mount Sinai uses IgniteData's Archer platform with Epic. It reports the result in numbers a site can measure itself against: about 70 percent of required eCRF data populated automatically, entry time down from 5.5 to 2.5 minutes per visit, and "100 percent data accuracy, with a 0 percent re-query rate" in the data evaluated so far. It has also built the system into startup for new trials, and is extending it to medical history, concomitant medications and efficacy lab data.
How much of your eCRF can actually move
Not all of it, and the share varies by study. A 2023 feasibility study in the Journal of Medical Internet Research tested automated transfer in German hospitals and found that "5 forms with 185 variables contain 67 automatically transferable variables (67/274, 24% of all variables and 67/185, 36% of eligible variables)." Mount Sinai's 70 percent sits at the other end, in oncology trials with heavy lab and vital sign content and a single EHR.
The pattern behind the spread is simple. Structured, coded data moves: labs with units, vitals, medication orders, demographics. Data that lives in physician notes, rating scales done on paper, adverse event narratives and anything the protocol defines differently from routine care does not. Before you assume a number, take one active study's eCRF and mark each field as "in the EHR as a structured value," "in the EHR as text" or "not in the EHR." That one exercise tells you more than any vendor estimate.
What FDA expects when EHR data flows into EDC
FDA's guidance on this is from July 2018 and is still the reference: Use of Electronic Health Record Data in Clinical Investigations. It wants sponsors to "promote the interoperability of EHR and EDC systems" and notes that interoperable systems "may also reduce errors in data transcription." Four points matter at the site.
- Part 11 applies to the EDC, not your EHR. The guidance says "FDA does not intend to assess compliance of EHR systems with 21 CFR part 11." The EDC and the transfer process are the sponsor's to validate.
- The transfer has to be shown to work. Sponsors should ensure the interoperability "functions in the manner intended" and that "data are transmitted accurately, consistently, and completely." Ask the sponsor for that documentation and file it.
- Source stays yours. "Clinical investigators must retain all paper and electronic source documents (e.g., originals or certified copies) and records" as 21 CFR 312.62(c) requires. An integration does not move that duty.
- FDA can see the EHR. "All relevant information in the EHR pertaining to the clinical investigation must be made available to FDA for review upon request." Your inspection plan needs a way to give an investigator access to the relevant record without opening the whole hospital system.
The guidance also asks that consent documents identify who may gain access to the participant's electronic health record, including sponsors, monitors and regulators. Check your current consent language says so.
Questions to ask before a sponsor or your health system turns it on
- Which fields are mapped, by name? Get the mapping list for your study's eCRF, not a percentage.
- Who reviews auto-filled values before submission? Mount Sinai's release does not describe its review step, and your SOP has to. Decide whether the coordinator confirms each form or reviews by exception.
- How do corrections flow? If a lab value is corrected in the EHR after transfer, does the EDC update, and is the change visible in the audit trail?
- What happens to units and reference ranges? A lab reported in different units at two hospitals is a query waiting to happen.
- What is still manual? List it. That list is your coordinator's real workload, and the right basis for the visit budget.
The mistake sites make at question 5
They take the headline percentage as the time saving. If 70 percent of fields fill automatically, the 30 percent left is usually the hard part: assessments, narratives, protocol-specific scales. Visit time does not fall by 70 percent. Mount Sinai's own number is from 5.5 to 2.5 minutes per visit, which is real and worth having, and still leaves a person doing the rest. Measure your own before and after, the way our post on clinical trial site burden suggests, before you promise anyone capacity.
Where we fit
EHR to EDC interfaces belong to your EHR, your EDC and the sponsor. What we build sits beside them, on our clinical research systems page: source document digitization for the paper the integration cannot read, with every value confirmed by a person against its source image before it goes into the EDC through the EDC's own interface. The time a site gets back is best spent where recruitment is won, which is answering a referral in the first hour, covered in clinical trial recruitment.
Count your fields before you count your savings.
