Clinical trials are all about the data:
Why ensuring stakeholder alignment on eCOA data management roles and responsibilities is critical
By: Steve Lyons, Vice President, Data Delivery and Integration Engineering, uMotif
Every aspect of a clinical trial, from demonstrating therapeutic benefit to monitoring participant safety, depends on the collection of accurate, reliable, and complete data. High-quality data enable both sponsors and regulators to make evidence-based decisions about the efficacy and safety of new treatments, as well as supporting decision-making by investigators during trial conduct.
When clinical outcome assessments are being captured electronically (eCOA), including participant-reported outcome measures (ePROMs), clinician-reported outcomes (eClinROs), observer-reported outcomes (eObsROs), and performance outcome measures (ePerfOs), ensuring clear stakeholder alignment throughout the lifecycle of a study with regard to data management activities is critical to streamline activities and study success.
Defining Roles and Responsibilities
Successful eCOA data management begins with the very basic task of clearly defined roles and responsibilities among all stakeholders. These stakeholders typically include the sponsor, the eCOA technology provider, contract research organizations (CROs).
Clearly outlining responsibilities ensures that all parties understand who is accountable for tasks such as database configuration, data review, issue resolution, and system maintenance – especially in instances where it could be either party that could conduct an activity. Establishing governance structures and communication pathways early in the study will help prevent delays and ensure that any data-related issues are addressed efficiently and in a timely manner.
Data management activities for eCOA systems span the full lifecycle of a clinical trial, from system set-up through study close-out.
1. Set-Up and Go-Live
Effective data management begins long before data collection actually starts, and the set-up phase lays the foundation for successful data collection.
When the eCOA system is being designed and configured, including questionnaire design, visit schedules, reminders and alerts, reports, system integrations, data managers should be actively involved to ensure efficient downstream data handling. Ensuring the right people from all Organizations (sponsor, CRO, eCOA vendor) are represented during design meetings is key.
Alignment between data managers from all parties on elements such as:
- variable naming conventions
- scoring and calculations
- validation/ edit check rules
- data formats and metadata standards
- analysis requirements
- data change processes
- data capture back-up methods
- data access
- data transfer frequency
It is also important that during user acceptance testing (UAT), where sponsors confirm that the system functions as expected (in line with the design specifications), that data exports and reports are tested. This ensures that the data outputs from the eCOA are submitted and stored correctly [1].
At uMotif, we provide role-based access to self-service data downloads, allowing data to be downloaded at any time.
Table 1. Data management activities and responsibilities during eCOA Set-up
| Data Management Organization | ||
|---|---|---|
| Activity | eCOA Provider | Sponsor (and/or designated third party such as CRO) |
| Create data management plan (including data flows and data change requirements) | ||
| Create eCOA data transfer specification | ||
| Review and approve eCOA data transfer specification | ||
| Conduct UAT | ||
2. Ongoing Study Conduct
Once the study is live, and data starts to be captured, ongoing data monitoring becomes essential to ensure continuous data quality and system performance.
Powerful reporting and dashboards, such as those at uMotif, support Sponsors and CROs, as well as sites, to visualize the data and quickly identify issues.
Despite carefully designed protocols, systems, and site and participant training, mistakes happen, and situations can arise during a trial where a data change is required. Any requests to change data during a trial must have sound justification and follow a controlled data change request process, as outlined in the data management plan and agreed upon during set-up.[2] All changes made should be clearly captured in the audit trail.
Regular meetings between data management teams should occur, and cover elements such as:
- compliance
- missing data
- technical issues
- data formatting
- any integrations are functioning as expected
- data change requests
Table 2. Data management activities and responsibilities during eCOA data collection
| Data Manager Organization | |||
|---|---|---|---|
| Activity | eCOA Provider | Sponsor (and/or designated third party such as CRO) | Investigator |
| Monitor incoming data and reports for completeness and anomalies | |||
| Request data changes | (if this happens, investigator authorization is required) |
||
| Approve data changes | |||
| Implement approved data changes | |||
| Data cleaning | |||
| Export data to Sponsor | |||
| Review data exports | |||
3. Close-Out Activities
At the end of the study, close-out activities focus on ensuring the completeness and accuracy of the collected data before database lock. Key tasks include comprehensive data cleaning and the resolution of outstanding queries or discrepancies.
Data correction processes must follow defined procedures to maintain audit trails and regulatory compliance. Once all issues are resolved and the data are verified, the dataset can be finalized for statistical analysis and regulatory submission.
Table 3. Data management activities and responsibilities during eCOA close-out
| Data Manager Organization | ||
|---|---|---|
| Activity | eCOA Provider | Sponsor (and/or designated third party such as CRO) |
| Data cleaning | ||
| Data reconciliation | ||
| Resolve open queries | ||
| Lock database | ||
| Final data transfer | ||
| Statistical Analyses | ||
| Archiving | ||
The future…What can AI do?
Artificial intelligence is beginning to play a practical role in supporting eCOA data management by helping study teams identify issues earlier and work more efficiently. By reducing the time needed to review and interpret routine study data, AI has the potential to accelerate issue detection and improve overall data oversight, and help teams manage growing data volumes more effectively while maintaining high standards of data quality and integrity.
At uMotif, our MotifAI Assistant is designed to help teams interact with study data more easily and surface insights more quickly. For example, AI can help summarize study metrics, highlight emerging trends in participant compliance, and support faster review of data quality signals across sites and participants.
AI models can also help predict the likelihood of certain outcomes, such as which participants are at risk of missing diary entries or dropping out of the study. These insights can allow study teams to intervene earlier, improving compliance and reducing missing data.
Another emerging application is automated anomaly detection across integrated study datasets. Because eCOA data often interact with other clinical data sources such as EDC, safety systems, and wearable device platforms, AI tools can help identify inconsistencies between systems more quickly than traditional rule-based checks alone.
Importantly, AI acts as a support tool for study teams rather than replacing established data management processes. Human oversight remains essential for interpreting findings and making decisions that affect study data, trial conduct, and can impact participant safety.
Conclusion
As the use of digital tools continues to expand within clinical research, effective data management for eCOA systems is more important than ever. From initial design and system setup to ongoing monitoring and final data cleaning, each stage of the study lifecycle plays a role in ensuring the reliability and usability of collected outcomes data.
By clearly defining responsibilities, rigorously testing systems and data flows, and maintaining continuous oversight throughout the study, organizations can ensure that eCOA data contribute meaningfully to the scientific and regulatory evaluation of new therapies.
References:
- Gordon S, Crager J, Howry C, Barsdorf AI, Cohen J, Crescioni M, Dahya B, Delong P, Knaus C, Reasner DS, Vallow S, Zarzar K, Eremenco S; Electronic Patient-Reported (ePRO) Consortium, PRO Consortium. Best Practice Recommendations: User Acceptance Testing for Systems Designed to Collect Clinical Outcome Assessment Data Electronically. Ther Innov Regul Sci. 2022 May;56(3):442-453. doi: 10.1007/s43441-021-00363-z. Epub 2022 Mar 1. PMID: 35233726; PMCID: PMC8964567.
- Delong PS, Humler D, Haag T, Yeomans, A, Andrus J, Eremenco S, Finan A, Gable J, Gilfillan D, Howry C, Kern S, Lesniewski SJ, Simpliciano K, Staunton H, Turnbull J, Workman C, Raymond S, on behalf of the Electronic Clinical Outcome Assessment (eCOA) Consortium and Patient-Reported Outcome (PRO) Consortium, and eClinical Forum. Best Practice Recommendations for Electronic Clinical Outcome Assessment Data Changes. Journal of the Society for Clinical Data Management. 2023; 2(1): 12, pp. 1–13. DOI: https://doi.org/10.47912/jscdm.249