Due to the high demand, on the 9th of March we repeated our complimentary webinar “How to Perform a Successful Data Migration in Life Sciences”. In this webinar, we focused on the best practices of building a lean and structured approach to data migration in Life Sciences.
Together with the webinar participants, we discussed the main risks and challenges of data migration and ways to minimise the data migration impact on day-by-day business operations.
Continue reading to discover the questions from the Q&A session from the webinar.
In data mapping, the use of Excel helps to find the items that need to be checked. Any suggestions on how to make this document highly readable considering that writing a more discursive document is time-consuming and not really useful for easy control of the items?
If you are forced to use an Excel document for data mapping specification, our first suggestion is to identify the people that will have to review the document. If end users have to go through continuous revisions, it will be useful to use the format that they are used to.
Here are some tips:
- Present the data in the same way the application interface shows it instead of using database tables
- Structure the Excel in a way that allows you to easily consult the documents
- Make sure you can quickly find the information in your document
- Group information by colour to identify pending and completed tasks, or topics that are important to be completed such as dependencies with other topics
Any additional tips on measures to be taken for a file system migration where files also have to undergo AIX to LINUX transformation?
The only way to migrate properly is to use scp, sftp, rsync, and samba (all software) to copy from one computer to the other file by file (most of the tools allow copying of folders). The software takes care to change from Big Endian to Small Endian. Another option would be tar on AIX and to incar on LINUX as the incar does the conversion automatically. The above mentioned software is GAMP 5 Category 1, thus, no validation is required. We suggest doing a spot check for one file and count the files on both sides.
Which quality documents a pharma company should prepare internally to properly document the migration?
Each company should prepare the documents in compliance with its internal procedures.
How legacy systems should be managed and discontinued? / Once the migration is over, how should I manage the decommission of the legacy system?
The management of the discontinuation process is called decommissioning. It is a final stage in the lifecycle management of any application or product.
In order to decommission an application, you have to do a validation exercise like you have done for data migration. This means, that you have to create a Decommissioning Plan in order to:
- cover rules and responsibilities like data migration plan
- identify and correlate the types of services and the types of users that use the service
- identify alternative services if available
- develop a communication plan for each type of user and sub-service linked to the dismissed system
Once the plan is completed and the decommission of the service approved, the service can go offline.
When setting the decommissioning date, you have to consider the impact on your users because they will not be able to access the service after this milestone.
Once the service is offline, the decommissioning is completed. After that you have to consider SLAs, licenses, and warranties. Keep in mind that the service is committed to these agreements until this milestone.
Please note that decommissioning process is a part of a Data Migration strategy. This means that some data in the source system could not be migrated to the new system, maybe for technology constraints, but it still has to be consulted after the migration.
Archiving and making data read-only are the options that you should consider as part of the strategy and planning phase of your data migration projects.
It is important to say that keeping the data read-only also requires consideration of some factors, such as:
- Cost associated with maintenance and licenses
- Limited access to data
- No data changes
- Data record retention requirements
We advise to make sure you evaluate well a strategy for archiving and consulting of the data that was not migrated.
How long did it take to perform this pharmacovigilance data migration you have just described?
The automatic migration with ETL tool took about 6/7 months for the headquarter and 5/6 months for the affiliate that have the same source database.
The migration with XML takes about 3 months.
In which cases should I do automated data migration and in which – semi-automated?
There are several aspects to take in consideration while choosing the type of data migration:
- Type of data and the amount of data that you need to migrate
- Potential data loss:
- In some case, a semi-automatic migration uses export features available out of the box in the solution that could cause loss of data. For example, in pharmacovigilance, if you export a case data using E2B XML R2 format, you might lose some of the data of the original case, such as note fields that are not included in the standard E2B R2 format;
- With automatic migration, you have the possibility to control all the fields and decide where and if migrate the data;
- Internal resource effort:
- In a semi-automatic data migration, your team has to perform some of the activities manually;
- Automating the process will allow your team to do more valuable tasks instead of moving data from one place to another;
- Costs:
- The complexity of automated migration might require support of the specialised consultants and specific tools;
- In some cases, it might be more efficient and cheaper to perform manual activities even if they require the allocation of external resources;
- Downtown period:
- maybe required to perform the data migration that could lead to a too long unavailability of the system.
As we see, there are numerous aspects to take in consideration while choosing the type of data migration. Following a structured methodology presented during our webinar will help you to create a consistent Data Migration Plan and to take the decisions that work the best for your project.
Do regulatory agencies look at data migration process during audits?
Absolutely yes, agencies always verify if data integrity has been maintained in a data migration project especially with critical data migrations like the pharmacovigilance one.
They also always look at the migration strategy and the migration verification approach.
Do you have any questions about data migration that you did not manage to ask during the webinar? Contact us to talk to our experts!
