End of Study Data Migration

We often see that organisations fear migrating data at the end of the study. Concerns include a heavy burden on resources, and that the data is at risk during the migration process.

In the previous section, we explored the benefits of transferring essential GCP records into a consolidated archive. An additional benefit of migrating the data is that it provides sponsors with the opportunity to QC the data as well. In Arkivum’s experience, it is often the case that we find missing or incorrect documents during the migration process. If kept within the source systems, or not properly checked at the end of the study, this could cause major issues later down the line.

Arkivum has led many end of study migrations, transferring data from various stakeholders into our customers' digital archives. In this section, we’ll explore some of our key learnings and tips for a successful end of study migration into a dedicated archive.

A validated migration process?

Study Data Migration is the process of permanently moving existing data from one system into another. This includes metadata, audit trails and any other information and records that permit and contribute to the evaluation of the conduct of a trial and the reliability of the results produced.

As described in the 2023 EMA Guideline on computerised systems and electronic data in clinical trials, “In the course of the design or purchase of a new system and of subsequent data migration from an old system, validation of the data migration process should have no less focus than the validation of the system itself”.

The EMA 2023 guidelines describe requirements for performing data migrations (Section 6.10 Migration of Data). These guidelines summarise much of the good practice described in the 2022 GAMP5 2nd Edition of A Risk-Based Approach to Compliant GxP Computerised Systems (Appendix D7 Data Migration).

"As one of APLs requirements was that records remain easily searchable in the event of an inspection by the competent authority, having meaningful metadata was key.

An initial step in migrating data into the system is called “mapping”...the team assessed and organised all metadata to reflect the structure of the TMF while also identifying gaps to be recovered. Through organised and complete metadata, documents remain easy to access."

Ten steps for a successful migration

Given the importance of data migrations, in particular, the need to maintain and validate data integrity during the migration process, we recommend the following steps as good practice and in line with the regulatory guidelines.

Data migrations should be properly planned. A documented Data Migration Plan (DMP) should be produced and agreed upon with all stakeholders in advance of the migration commencing.
The DMP should include a risk assessment, especially around data integrity, and ways to mitigate risks, including the use of mock data and test migrations to expose unforeseen problems.
The DMP should define the roles and responsibilities of all involved, especially if data migration is being delegated to a third-party supplier.
Data migration should not just focus on data such as documents and databases, but should also include all relevant metadata, and, most importantly, migration of audit trails. Audit trails are a go-to for inspectors when evaluating the conduct of a trial and the reliability of the results. Moving data between systems, including into an archive, doesn’t remove the need for ongoing access to audit trails for the whole data lifecycle, including when the data used to be live.
Particular attention should be given to exports from source systems to ensure everything that needs to be migrated is exported correctly and that data integrity is maintained. It is common for organisations to spend a lot of effort validating how they can get data into source systems, but a lot less effort validating they can get it out again! Export issues are the number 1 problem we see when doing data migrations.
Data, metadata and audit trails should remain connected, directly accessible, and searchable/navigable after the migration, for example, to support ongoing Audit Trail Reviews and regulatory inspections of data in the target system.
Data verification tests should be done on the target system to confirm the complete and correct migration of data from the source system. Verification activities should be proportionate to the criticality of the various types of data being migrated, for example using sampling strategies for QC that focus on the essential records under ICH E6 (R3).
The tools and systems used to perform the migration should be fit for intended use, i.e. validated or qualified, and should be under configuration management and change control. Inspectors like to see evidence that systems are ‘fit for purpose’.
Simple data checks go a long way, especially if used at each stage of the process. For example, make inventories of all the source data to be migrated and its locations, make inventories of data after each transfer, for example after transfer from CRO to Sponsor, and make inventories post migration after data is loaded into the target system. These can then be compared and reconciled at each step to show that nothing has been missed or altered.
Document all the steps and activities above in a Data Migration Report as evidence that the data migration was conducted as planned, risks were successfully managed, and the migration has been validated. This provides valuable evidence for inspectors that Data Integrity has been managed appropriately during the migration.

Outsourcing your migration process

Outsourcing the migration process is a genuine option for sponsors. Arkivum has managed many migrations for its customers. Our service (as I’m sure others like it) dramatically reduces the effort to migrate study data into their archive at the end of a trial, while still crucially maintaining oversight of the process.

This includes working with both internal and external stakeholders to accept the data, ensuring the process is captured within the audit trail and that the data is quality checked before signing off on receipt.

Leveraging the experience of external experts who conduct regular migrations can ease pressure on organisations, enabling them to focus on other end-of-study priorities.