Three key takeaways from the TMF Summit 2024

Last month, Arkivum exhibited for the third time at the TMF Summit (this time taking place in Savannah, Georgia). For those of you who follow me on LinkedIn, I’ve already shared my top three takeaways from the conference. You can have a look at that post here if you’re interested!

I wanted to take the chance though to expand on those points, as given the type of platform LinkedIn is, I wanted to try and make the post as brief as I could.

So let’s dive into my top three takeaways from the TMF Summit’s 2024 edition.

The evolving mindset towards TMF retention

In previous years much of our conversation at the TMF Summit has been focused on answering questions along the lines of ‘Why do I need to put long-term records into a separate archive?’.

This year I felt a shift in those conversations.

The TMF community is much more aware of the need to manage long-term digital records, such as the TMF, differently to that of historic paper records. With evolving regulations such as EU Reg 536/CTR and ICH E6 R3 there is growing focus on sponsors to improve their approach to data archiving, to ensure that digital records endure and are legible for the entire retention period.

That said, there is still more that can be done. Views still exist that locking a study down at the end of a trial for a couple of years is a good approach. I would like to challenge that view.

Locking down a study simply delays the process of migrating that data into a fit for purpose long-term digital archive. When it comes to migrating that data two or three years later, it will be a much greater challenge as staff will have moved on to other projects or organisations, and those managing the migration will be unfamiliar with the records.

Moving the data into a centralised long-term repository which mirrors the structure that the data is organised in at the end of the trial, will significantly reduce the risk of any data integrity issues (and ultimately compliance issues) later down the line.

From Arkivum’s experience we see far more issues with locked studies being migrated years after close out when compared to completing that process near to the end of the study phase. Additionally, for sponsors accepting documents back from CROs, this migration process can act as an excellent opportunity to fully quality check the documents before signing off on receipt.

AI is very much on the TMF agenda

I may be wrong in saying this but I don’t remember ever seeing or hearing AI discussed at previous TMF Summits. This year it was not only covered in headline presentations but made up the topic of several conversations on the Arkivum stand.

The community is very interested in ways in which AI can be leveraged to improve document sorting/indexing, improve quality and add value to other operational tasks.

Arkivum currently doesn’t leverage generative AI within our solution; what we do have is a lot of automation to complete large scale operations such as regularly checking every record for data integrity issues or digitally preserving each file within the archive.

Looking ahead though (and speaking broader than just within the Arkivum solution), there is huge potential for AI across TMF documentation and clinical data in general. Having a large database of trial information which is well indexed and has a significant amount of metadata associated to each record, will enable the possibility to query and interrogate that data. ‘How can I improve study design?’ is just one small example of the sort of question you could put to this data.

Crucial to this though is having that database in a healthy and clean state. Without it, current AI will largely struggle to extract value from it. Regardless, it will be fascinating to see where this leads in the future.

Risk-based approaches: The good and the bad!

This might be a controversial section for some, but despite being a hot topic right now I think that in some cases risk-based approaches are being used to the detriment of sponsors.

Traditionally a risk-based approach could be used to say check 10% of paper records for any issues, and to then extrapolate the number of errors across the entire document set. From there you would take appropriate action. In this instance, this approach makes sense given the time and resource required to manually check physical documentation.

My challenge comes when I see examples of where this concept has been transferred to managing digital documents and records. This shouldn’t be the case. Having the right tools (Arkivum’s solution or otherwise) means that you can automate tasks such as checking every record within a fraction of the time.

And this leads me onto what I would call the right approach to risk-based approaches. From a long-term data management perspective we would recommend understanding the overall retention objective (e.g. in this instance retaining the TMF in line with EU Reg 536 for +25 years), identify and understand the risks to achieving that objective, and then building an appropriate data archiving strategy in line with your organisations risk appetite…and ideally do much of this thinking/planning before the study starts.

This frames a risk-based approach in a much more strategic light, and I would argue will enable TMF managers, document and records professionals a much easier time in communicating the value of data archiving to the wider organisation.

Taking a strategic risk-based approach can be a powerful tool in transforming the way an organisation works.

To conclude…

I’ve obviously only picked three topics from the tens, if not hundreds covered at the TMF Summit this year. I hope though this has provided a useful snapshot of the event for those unable to attend and food for thought in your planning and approach to TMF retention and archiving.

For those who attended the conference, I’d be really interested in hearing what your takeaways were, whether via email (tom.lynam@arkivum.com) or start a discussion on LinkedIn.