Clinical data management (CDM) is the work of building a study database, validating it, reviewing every data point as it arrives, and delivering a clean, locked dataset for statistical analysis. In a medical device trial, careful CDM cannot change the results. It decides whether those results hold up when a Notified Body, an ethics committee, or the FDA looks at them.
In clinical research, data is everything. Every decision, analysis, and conclusion depends on the quality of the data captured during a study. Results can never be predicted, but one thing can be controlled: how carefully data is managed from the first patient in to the final database lock. At Qserve, we combine thorough preparation with continuous review. That keeps unpleasant surprises to a minimum and keeps studies moving forward with confidence.
Clinical data management in a medical device trial runs through five phases: planning, database build, validation, live data review, and close-out. A data manager writes the Data Management Plan, builds the eCRFs and edit checks in the Electronic Data Capture (EDC) system, tests the system before go-live, reviews and queries data throughout the study, and locks and archives the final dataset.
|
Phase |
What happens |
What it delivers |
|
1. Planning |
Data Management Plan and Data Review Plan, aligned with the clinical investigation plan |
An agreed approach to how data is collected, checked, and corrected |
|
2. Database build |
eCRF design, database administration, and programming of validation (edit) checks |
A database that captures every protocol-required data point |
|
3. Validation |
User acceptance testing (UAT) under every user role |
Evidence that the EDC works as specified before real patient data enters it |
|
4. Study conduct |
Site access and data entry instructions, daily review of new and modified data, query management, SAE follow-up, and exports for the data monitoring committee |
A dataset that stays current, with no backlog at interim analyses |
|
5. Close-out |
Final query resolution, database lock, and archiving |
A clean, locked dataset ready for statistical analysis and reporting |
The scale differs between a pivotal pre-market clinical trial and a PMCF study, but the five phases are the same.
ISO 14155 is the international good clinical practice (GCP) standard for clinical investigations of medical devices. Its current edition, ISO 14155:2026, replaced the 2020 edition in March 2026. It requires the sponsor to ensure that clinical data are accurate, complete, and verifiable against source documents, and that electronic data systems are validated, access-controlled, and keep an audit trail.
For data management, the practical consequence is simple: a validated EDC system is not optional, and the validation needs evidence a reviewer can follow. Other frameworks sit alongside the standard. Under the EU MDR, Annex XV sets out what a clinical investigation plan must contain, including how data will be handled and monitored. Health data from EU participants falls under the General Data Protection Regulation (GDPR), which is why a study database holds pseudonymized subject IDs rather than names. If the data will support an FDA submission, 21 CFR Part 11 sets the requirements for electronic records and signatures.
For a fuller walkthrough of the standard itself, read ISO 14155 explained.
User acceptance testing (UAT) is the final check of a clinical trial database before go-live. During UAT, the data management team tests every field, edit check, and EDC workflow under each user role, using both valid and deliberately wrong data. It is the only point at which a sponsor can confirm the database works as specified before real patient data depends on it.
Before a single patient is enrolled, a lot happens behind the scenes. Qserve supports sponsors with database design, database administration, and programming of validation checks, together with a clear Data Management Plan, a Data Review Plan, practical data entry instructions, and live access and an introduction for each site.
UAT is where all of that comes together, and it is not just a technical exercise. We quite literally test every field, feature, and EDC behavior. By logging in as each user role (Investigator, Site Staff, Monitor, Data Manager, and Sponsor), we test the system from every angle. We enter correct data, but also deliberately ridiculous data, to push the boundaries of the EDC. Four questions guide every test:
You can’t see whether code is neatly written or extensively tested. You can validate it through UAT.
User acceptance testing
Clinical trial database
Daily clinical data review
Clinical data management at Qserve: database setup and user acceptance testing before the first patient is enrolled, then daily data review through to database lock and archiving.
Daily data review means a data manager checks all new and modified clinical data every working day, from first patient in until the last query is resolved. Reviewing continuously prevents query backlogs, so the dataset is current at every interim analysis, data monitoring committee meeting, and export, and database lock is not held up by a large end-of-study cleanup.
Once the database is live and the first patient is in, the real work begins. Instead of facing a big cleanup at an interim analysis or at the end of the study, the data stays up to date throughout the trial. Serious adverse events (SAEs), data monitoring committee requirements, data exports, and the eventual database lock and archiving are all handled in a controlled and timely way.
That control pays off when a case gets complicated. An ambiguous device deficiency or a challenging SAE needs time and attention, and because the routine data is under control, there is room to dig deeper and really understand what is happening with the patient.
The most common clinical data management mistakes in medical device trials happen before go-live: an eCRF that doesn’t match the final protocol, edit checks that are missing or overly aggressive, and UAT that is rushed or skipped. During the study, the classic mistake is letting queries pile up until an interim analysis or database lock forces a cleanup.
A medical device sponsor typically outsources clinical data management when it has no in-house data manager, runs only a few studies a year, or needs a validated EDC setup faster than it can build one. A medical device CRO brings tested processes, trained data managers, and established UAT routines, so the sponsor keeps oversight without building the function from scratch.
Outsourcing doesn’t move responsibility. Under ISO 14155, the sponsor remains accountable for data quality even when tasks are delegated, so look for a partner who documents decisions and gives you direct access to the data as it develops. At Qserve, clinical data management is part of our medical device CRO services, either as a standalone service or within a full study. Documentation is the other half of inspection readiness; our guide to choosing the right eTMF covers that side.
Clinical study results cannot be predicted. Being prepared is entirely possible. By staying continuously up to date with clinical data, each result rests on a solid, reliable foundation. Careful data management is not just a process; it is peace of mind. And if you are still deciding when to start generating clinical evidence, read why you should start collecting clinical data early.
Whether you are preparing a new study or want more control over an ongoing one, Qserve’s data managers support you with structured, hands-on clinical data management, from database design to final lock. Get in touch with our team to discuss your study.