What is eSource and why does it matter?

The US Food and Drug Administration (FDA) defines eSource, or electronic source data, as data that are initially recorded in electronic format. It can refer to “information in original records and certified copies of original records of clinical findings, observations, or other activities captured prior to or during a clinical investigation used for reconstructing and evaluating the investigation”.² eSource dataflow is the electronic transfer of data after it is initially entered at a single point of entry into an electronic health record (EHR) or similar source system. There is no subsequent manual data entry involved post-original eSource entry.

Examples include direct from device data capture, such as via wearables or sensors; direct from patient or clinician, such as an electronic clinical outcome assessment (eCOA); or direct from an EHR or electronic data capture (EDC) system used as an eSource system.³ It reduces the need for source data verification (SDV), minimizing the need for transcription, and provides real-time guidance on illogical or inconsistent data. It can reduce the volume of automated and manual queries, which has the potential to decrease the workload of data review and circular back and forth reconciliation activities with sites. If implemented correctly, it can reduce the burden on sites and sponsors, boost patient centricity, and improve data quality.⁴

With new International Council for Harmonisation (ICH) Good Clinical Practice (GCP) guidelines recommending that researchers avoid unnecessary transcription of data, the drive to eSource is accelerating.⁵

eSource Implementation Consortium

eSource is an evolving space. While the FDA has published a general framework for its use, highlighting its benefits and the body’s commitment to Health Level-7 (HL7®) Fast Healthcare Interoperability Resources (FHIR®) standards,⁶ there are currently no standard guidelines regarding implementation.

The 2024 SCDM working group meetings and roundtable aimed to accelerate the adoption and scaling of eSource clinical trial data collection and transfer by building the foundations of a consensus for implementation. The program of work tasked key stakeholders from across the research ecosystem—including biopharmaceutical firms, sites, and technology vendors—with sharing their knowledge and best practice examples, in a bid to provide best practices for the implementation and operation of HL7® FHIR® data transfer. See Author Contributions section for participants.

This report summarizes those discussions and best practices in three key areas:

• The feasibility of existing and theoretical eSource approaches

• Contracting considerations by eSource approach

• Technological readiness

eSource approaches

Site readiness is a challenge, in no small part, because eSource is not a single pathway. Over the last five to ten years, there has been an influx of solutions, vendors, technologies, and methods all aimed at reducing or eliminating dual data entry between site EHR and sponsor/clinical research organization (CROs) EDC systems. With such a wide range of different approaches, navigating this new, complex landscape is difficult.

The eSource Implementation Consortium identified seven different eSource approaches, or scenarios, currently being used in industry:

• Site identifies preferred EHR-to-EDC technology

• Sponsor identifies preferred EHR-to-EDC technology

• Clinical research-focused eSource application

• Sponsor and site each use their own technology

• Direct data capture (DDC)

• EHR research module

• Mixed use of technology within same study

Of the seven identified approaches, two will be described in detail with case studies. These were site-provided and sponsor-provided preferred EHR-to-EDC technology, using a mutually agreed upon, system-agnostic technology vendor, and with HL7® FHIR® as the data exchange standard.

Currently, the Consortium are aligned that these two are the most widely used approaches for eSource dataflow using EHR systems meaning they have the richest base of use case examples, and for eSource EHR dataflow approaches are the least challenging to implement. As such, the eSource Implementation Consortium believe they have the greatest potential for scalability.

Approach 1: Site identifies the preferred technology

When sites establish technical connections with the EHR-to-EDC technology of their choice, the sponsor’s EDC system receives data from multiple site technology systems (see Figure 1).

Figure 1

Sites identify their preferred technology eSource approach.

For sites, it means all security and privacy documentation and approvals, system/vendor qualifications, contracts (e.g., site-vendor agreements, Clinical Trial Agreements), and other agreements are done upfront, before study set-up, and assigned study staff are already trained on the system. As the technology vendor already has experience of the site’s EHR set-up, there can be some reuse of mapping across sponsors and/or studies. In addition, the single set-up can enhance efficiency by minimizing the need for on-site IT (Information Technology) expertise and infrastructure. Technology vendors should share appropriate validation reports as required by sponsor and regulators.

Furthermore, once the solution has been established, the site can then expand its use to any or all of its pharma-sponsored studies, as applicable.

Sponsors tend to work with hundreds of sites, and this model means they will need to have relationships and mappings with many vendors. However, not all technologies will integrate well or seamlessly with the sponsor’s EDC system, and some sites may not have the resources necessary to maintain their own technology solution.

One proven option is that sponsors look to work with vendors that have relationships with high-volume preferred study sites. The large number of studies, patients and data elements available for exchange at such sites can contribute to economies of scale, thereby accelerating adoption. In addition, it does not put pressure on sites to implement a vendor they may only use once.

Case study: Site-identified technology in action

The Memorial Sloan Kettering (MSK) Cancer Center, USA, has made scaling eSource EHR-to-EDC a top priority. It has been working with a system-agnostic technology vendor to deploy a cloud-based virtual research assistant that provides interoperability between EHR and key research systems, such as EDC. It provides seamless, secure transfer of high volumes of structured clinical trial data. The two started working together in May 2023, with implementation taking place in September 2023, and the first studies going live two months later, in November 2023. As of January 2025, MSK is using this technology with four sponsors and 15 protocols and continues to scale.

MSK analysis of the first five live studies in the fourth quarter of 2023 found the system reduced the time and effort of data entry, based on data point per second, by 49% in commercially sponsored studies, and 68% in investigator-initiated trials. A validation study found the platform increased data entry throughout by 55%, from 10 data points/min with manual entry to 15 data points/min with the vendor technology platform and reduced data entry error by 99%.

The team also carried out a user satisfaction survey. They found all users preferred using the platform over manual data entry. It was easy to learn, easy to use, and resulted in fewer queries from sponsor/site monitors.

Case Study: Site-identified technology in action

Johnson & Johnson Innovative Medicine (J&J) conducted a comprehensive global analysis to identify site preferences for eSource vendors. As part of our commitment to enhancing clinical trial efficiency, J&J undertook an extensive evaluation of eSource solutions focusing on those typically favored by our site partners. Through this exercise and aligning with our internal business requirements, we ensured that the selected vendors met the diverse needs of multiple sites. The initiative included a request for proposals and subsequent pilot programs with several vendors preferred by our site partners. Importantly, our vendor selection criteria include the ability to be agnostic to both Electronic Health Record (EHR) and Electronic Data Capture (EDC) systems, which is vital for operational flexibility and integration. Our requirements were based on The European Institute for Innovation through Health Data (i-HD) vendor selection criteria.⁷

Following the alignment process, J&J decided to proceed with a vendor that garnered preference from multiple sites and was committed to partnering with us to meet our business requirements. In 2023, we initiated our first live study at one site and in 2024 we successfully scaled operations to six sites in two countries—specifically the United States and the United Kingdom. Looking forward, rollover work from 2024 provides a solid foundation to scale operations to an additional 14 sites, bringing the total to 20 sites with 56 integrations across countries in North America, the European Union, and Asia Pacific, as well as the United Kingdom. We plan to continue to pragmatically scale up in line with the maturity of the solution in this still emerging space with a goal that all studies that are using sites that are eSource enabled will source data using this solution.

The strategic expansion of eSource integration at J&J represents our dedication to innovation in clinical trials and our commitment to our critical site partners to reduce the burden of data collection. By leveraging site preferences and fostering a collaborative environment with our sites and vendors, we aim to enhance the quality and efficiency of our research initiatives globally advocating for greater data interoperability to benefit clinical research.

Approach 2: Sponsor identifies preferred technology

In this scenario, sites establish technical connections to each sponsor’s preferred technologies via middleware (see Figure 2). If a site is working with more than one sponsor, which is often the case, they will have to make multiple connections. Sponsors or their contracted vendors work with sites to establish mappings between site EHR data outputs and the sponsor’s standard or study-specific case report forms (CRFs) entered into their EDC system.

Figure 2

Sponsor identifies preferred technology eSource approach.

All EHR data coming into a sponsor’s EDC system will be via a single vendor. This provides simplified, more consistent data flow management—for example, for snapshots or locks—than when working with each site’s individual vendor. It also means the sponsor can be confident that the technology will work with their chosen EDC solution. Once the connection has been established with a site, the sponsor can then extend the solution to be used on any or all studies being conducted at that site.

However, this can lead to a longer start-up time for the first study. Each vendor will need to be assessed by the site’s security, privacy, and compliance processes before the program can be initiated. In addition, the sponsors technology vendors may not have experience with each site’s EHR set-up and security/privacy expectations and protocols.

It can also place additional burden on sites as it requires staff and resources to work with each sponsor and vendor. Site users also need to learn, adopt, and use multiple systems.

Case study: Sponsor-identified technology in action

Since 2018, Pfizer has been working with a technology vendor to develop middleware to enable EHR to EDC data transfers.⁸ The approach has proven successful from a data flow perspective.

The team started small, first focusing on demographic and vitals data with a small number of sites before moving on to include laboratory data to enable developing the technology solution further.

When scaling up the number of sites, there were several considerations. Firstly, not all sites across a large study were open to electronic data transfer and sites had variations in EHR structures. This led to scalability challenges with initial set-ups on a site-by-site basis taking a significant amount of time. Once a site was configured, however, the data transfer was successful with production data transfers enabled using this solution.

Case study: Sponsor-identified technology in action

Quantum Leap Healthcare used a hybrid model when implementing a vendor technology solution of an EHR-EDC integration framework. Sixteen of 35 sites working on a Phase II platform trial adopted the system, which allows patient health record data to flow directly from site EHR systems into electronic case report forms (eCRFs), while those without integration readiness continued to enter data manually.

After eight of 16 site deployments were activated, the team studied the impacts of EHR to EDC integration and recorded:

• A 16-minute per form saving for elapsed data entry time on the primary daily eCRF – a 61% reduction compared to manual entry;

• For just 52 patients studied at one site with a 15-day treatment course, 208 hours of clinical research coordinator time was saved;

• The elimination of data errors by the automated, validated EHR-loaded data and the downstream costs associated with cleaning and SDV;

• Low implementation costs

• Reusability across sites

Quantum Leap Healthcare has since deployed the integration to three trials, across 24 sites, and continues to scale.

Contract readiness

Contracting was identified as a challenge to EHR-to-EDC transfer adoption and scalability, representing a significant bottleneck in moving eSource from theory into practice.

The eSource Implementation Consortium recommend understanding an organization’s contracting requirements at the outset. This can ease the journey and shorten the time to study set-up.

Overall:

• most contract groups will need to confirm that all security/compliance requirements have been met;

• most contract groups will require confirmation that the technology needed to support the validated software can be provided, both from a functionality and a resource perspective, and the validation documentation should be available for audit/inspection;

• each contracts office will have boilerplate language related to data, security, privacy and Intellectual Property that the vendor will need to agree to, as well as standard implementation, change management/amendment policies;

• contracts cannot be executed if the required functionality is not present; and

• every implementation of eSource has a cost, and contracts cannot be executed until budget feasibility has been confirmed.

It is also worth noting that navigating regional variations in regulations can take time. While this part of the process does not tend to impact contracts with sites, it does require different procedures and, potentially, different language within vendor contracts.

As execution of the contract is dependent on all the listed components, the timeline is largely driven by their complexity and achievability.

Technological readiness

The numerous considerations when approaching technological readiness can be grouped into the following categories:

• Implementation: Includes installation, testing, validation, performance and re-verification triggers

• Protocol: Includes evaluation/feasibility; schedule of activities; change management/amendments

• Data standards and mapping: Includes transfer specifications

• Data cleaning: Data integrity checks, data reconciliation with different sources, or data consistency within a source. Includes SDV, source data review (SDR), data changes, queries, audit trail review, investigator sign off

• Compliance and provenance: Includes ensuring data origin/authenticity

The eSource Implementation Consortium recommends that these categories be considered as early as possible in your planning.

Conclusions and next steps

Overall, the working groups concluded that approaches one (site identifies the preferred technology) and two (sponsor identifies the preferred technology) are currently operationalized and are the most scalable eSource EHR dataflows. Approach one is the most commonly used of the two. These two approaches are widely used, meaning they have the richest base of use case examples, and, importantly, are the most feasible for implementation. While DDC into EDC is arguably the least challenging to adopt from a technological perspective, the approach although manageable adds complexity elsewhere, particularly around the potential threat to contemporaneous updates to EHR data as applicable.

When the sponsor or site identifies the preferred technology, they do so in partnership with a vendor. These vendors have extensive experience and expertise in areas such as data standards, mapping, and transfer processes—experience and expertise that sites or sponsors may not necessarily have in-house. They will also have access to a wide range of use case examples to help guide best practice implementation. While there is still a question regarding which of the two is preferable, these options, which narrow down the wide range of potential starting points, are an important step in the direction of scalability.

There is also now a consensus, built on the FDA’s support, surrounding the use of HL7® FHIR®, and its use as standard for US trials is incontrovertible. However, there are still issues to be fleshed out, particularity in the context of global trials. There is a need for clear regulatory guidelines that consider privacy and security laws across countries. Currently, such documents tend to be FDA-generated and US focused, but trials are often multinational. A global Office of the National Coordinator for Health Information Technology (ONC) Certification could help teams to roll out eSource more widely.

One large area of discussion was a lack of standardized mapping, leading to repeated work and wasted time and effort at every study set up. There are often, for example, hundreds of LOINC codes for glucose measurements, and understanding which should be used for each trial can be time consuming and resource draining. Creating a library of common mappings, with collaboration between sites, vendors, standards organizations and sponsors, could help to overcome this barrier, but regulatory guidance is required. Adverse event domains need particular attention, as definitions, workflows, and data requirements vary across patient monitoring, clinical intervention, FDA safety reporting requirements, and clinical trials monitoring use cases. Fortunately, work is ongoing on this area. The United States Core Data for Interoperability (USCDI+) project, for example, has seen the ASTP/ONC, National Cancer Institute (NCI), and FDA working with EHR vendors to identify a minimum dataset and deploy the HL7® Vulcan Adverse Events (AE) Implementation Guide. Moving forward, a project to acquire more information on AEs, and how sites and EHRs are capturing them, would be useful. The SCDM will continue to support work to improve domain mapping availability.

In addition, the metrics required to measure success are still under discussion, though the roundtable did reveal a general trend toward focusing on site satisfaction.

Developing a standard process for archiving EHR data, which is needed to ensure inspection readiness, is the last piece of the puzzle. Organizations cannot, after all, use clinical research data for its intended purpose if they are unable to demonstrate its provenance. Such a process could be based on the historical transition from paper to EDC, which led to eCRF, eCOA, and other third-party data being returned to site in archive packages, and sites being required to keep a certified copy of any eSource data as a document noted in the sponsor Data Management Plan (DMP). The FDA’s EHR eSource guidance (https://www.fda.gov/regulatory-information/search-fda-guidance-documents/use-electronic-health-record-data-clinical-investigations-guidance-industry) includes principles for tracking data originators and data element identifiers in the audit trail to demonstrate data provenance and chain of custody. To build recommendations in this area, the community needs to work with regulatory agencies to understand how they access submitted EHR data.

In conclusion, there are ongoing projects and initiatives aimed at improving data standards and interoperability, and new technologies and approaches to enhance data integration and improve efficiency in clinical research. There is still much work to be done as a hybrid world of traditional data collection and modern eSource dataflows compete today, but the SCDM hopes the best practice recommendations set out in this paper will help the sector to continue to move toward wide-scale eSource adoption.

Competing Interests

Linda S. King – employee of Sponsor company, Astellas Pharma

Michael Buckley – none

Amy Cramer – employee of Sponsor company, J&J Innovative Medicine

Muzafar Mirza – employee of Sponsor company, Pfizer

Cory Ennis- none

Lauren McCabe – previous employee of Sponsor company, Pfizer, current employee of CRO, Everest

Cal Collins – employee and shareholder of OpenClinica LLC

Aruna Vattikola – employee of Sponsor company, Merck &; Co., Inc.

Author Contributions

Working Group Leads: Lauren McCabe, Muzafar Mirza, Cory Ennis, Cal Collins, and Michael Buckley; Roundtable Leads: Aruna Vattikola and Michael Buckley.

Working Groups and Roundtable Participants: Ben Baumann, OpenClinica; Amy Cramer, Johnson & Johnson Innovative Medicine; Christopher Dunn, Pfizer; Linda S. King, Astellas Pharma; Joseph Lengfellner, Memorial Sloan Kettering Cancer Center; Dalia Mack, Duke Medicine; Rakesh Maniar, PKR Health and Life Sciences LLC; Catee Mullen, Duke Medicine; Dhwani Patel, Bristol Myers Squibb; Anna Patruno, Memorial Sloan Kettering Cancer Center; Denise Snyder, Duke Office of Clinical Research; Zach Taft, IgniteData; Ira Togo, Duke Medicine; Thanh Tran, Astellas Pharma; Barry Treichel, Advarra; Chris Weiss, OpenClinica; Veronique Wilson, Premier Research.

1. Cramer AE, King LS, Buckley MT, et al. Defining methods to improve eSource site start-up practices. Contemp Clin Trials Commun. 2024; 42:101391. Published 2024 Nov 14.

2. US Food and Drug Administration (FDA). Guidance for Industry Electronic Source Data in Clinical Investigations. Published 2013 Sep. Available at: https://www.fda.gov/media/85183/download Last accessed: January 17 2025.

3. Wang B, Lai J, Jin F, Liao X, Zhu H, Yao C. Clinical Source Data Production and Quality Control in Real-world Studies: Proposal for Development of the eSource Record System. JMIR Res Protoc. 2022; 11(12):e42754. Published 2022 Dec 23.

4. Parab AA, Mehta P, Vattikola A, et al. Accelerating the Adoption of eSource in Clinical Research: A Transcelerate Point of View. Ther Innov Regul Sci. 2020; 54(5):1141–1151.

5. The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). (2023) ICH Harmonised Guideline Good Clinical Practice (GCP) E6(R3). Available at: https://database.ich.org/sites/default/files/ICH_E6%28R3%29_Step4_FinalGuideline_2025_0106.pdf Last accessed March 5 2025.

6. US Food and Drug Administration (FDA). Role of CDER in the Health Level Seven (HL7) Vulcan Accelerator. Published 2024 May. Available at: https://www.fda.gov/drugs/electronic-regulatory-submission-and-review/role-cder-health-level-seven-hl7-vulcan-accelerator Last accessed: January 17 2025.

7. The European Institute for Innovation through Health Data. Early adopters’ minimum criteria for success. Available at: https://www.i-hd.eu/wp-content/uploads/2024/09/D1_Early-adopters-minimum-criteria-for-success_Download-ver-004.pdf Last accessed: January 30 2025.

8. Ochsner Health. Ochsner Health and Pfizer Partner to Develop Innovative Models for Clinical Trials; 2019. Available at: https://news.ochsner.org/news-releases/ochsner-and-pfizer-partner-to-develop-innovative-models-for-clinical-trials Last accessed: January 30 2025

9. European Medicines Agency (EMA). Qualification opinion on eSource Direct Data Capture (DDC). Available at: https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/qualification-opinion-esource-direct-data-capture-ddc_en.pdf. Published 2019. Last accessed: January 17 2025.

10. US Federal Register. Exploration of Health Level Seven Fast Healthcare Interoperability Resources for Use in Study Data Created From Real-World Data Sources for Submission to the Food and Drug Administration; Establishment of a Public Docket; Request for Comments. Available at https://www.federalregister.gov/documents/2025/04/23/2025-06967/exploration-of-health-level-seven-fast-healthcare-interoperability-resources-for-use-in-study-data Last accessed: June 4 2025.

11. US Food and Drug Administration. Real-World Evidence. Available at: https://www.fda.gov/science-research/science-and-research-special-topics/real-world-evidence Last accessed: June 4 2025.

12. eClinical Forum. The eSource Readiness Assessment Tool (eSRA). Available at: https://eclinicalforum.org/site-system-assessments Last accessed: 17^th January 2025