Will anyone still be doing SDV in 2035?

Duncan Scattergood | | 3 minute read

No.

But actually, if I am being honest, that answer relies on a very specific interpretation of the word “anyone”.

Let’s take a step back and think about the problem that we are trying to solve with SDV: we need to know that the clinical data collected in the course of a study is accurate. If it is not, then we might be concluding the wrong things about the investigational product and render the whole study pointless. High quality clinical data is absolutely mission critical.

Regulators share this concern — guidance such as ICH E6(R3) and FDA risk-based monitoring frameworks emphasize that critical data should be verified and monitored in a way that balances quality, patient safety, and operational efficiency.

The current situation

Setting aside more modern approaches such as ePRO (electronic Patient Reported Outcomes) or eSource, the typical process today is that when a medical professional sees a patient, information from the visit is recorded in the site’s EHR system. Later, a member of the site team manually re-enters the relevant data into the sponsor’s or CRO’s EDC system.

Manual re-entry is an opportunity for error. Hence, the idea of SDV where a CRA checks that the data matches across the two systems. But there are problems that we are all familiar with:

• SDV can consume a lot of expensive CRA time, say 30–40% of total monitoring effort in many studies – time that could potentially be better used on other tasks that would benefit from human judgment
• Other initiatives, such as improved data validation in the EDC forms and consistency checks at the point of ingesting the data into the study database, make it less likely that anything is missed and further reduce the value of CRA efforts
• SDV doesn’t discover data inaccuracies in the EHR system itself – inaccuracies that might be picked up by alternative processes such as SDR
• In truth, some data points matter much more in determining conclusions from the study than others.

Promised solutions didn't work

In theory, we can eliminate the need for SDV with eSource (i.e., an automated interface from the EHR to the EDC). But, over the years, this has been a pipe dream because the cost of implementation is too high: there are many EHR systems and different implementations at different sites, so a standardised approach is difficult. Moreover, in most studies, there are only a few subjects at every site. This means that the economics simply don’t add up.

So, is there a way that we can eliminate manual SDV without the challenges of a traditional eSource architecture?

A new era

As with so many things at the moment, the use of AI in a targeted, task-oriented way offers a way forward.

With only limited instructions, it will take the two data sources and identify discrepancies. Even if APIs are not available for the EHR system, AI will be capable of looking at and potentially automating the UI to capture images and use those images as one of its two sources. And moving forward, if we use AI to reconcile the data in this way, the next step is to convert that reconciliation into a process that makes the eSource dream a reality.

As an industry, we will need to think about validation, but validation of non-deterministic AI processes is going to have to be dealt with in many contexts.

Ultimately, the goal isn’t to eliminate SDV for its own sake, but to ensure that every check we perform in a trial adds measurable value. As the tools and technologies of Clinical Operations evolve, the most successful organizations will be those that align their people and processes around the same principle: data quality built in from the start. Whether it’s through AI-assisted reconciliation or smarter monitoring workflows, the opportunity is to let humans focus on the work that truly benefits from judgment, empathy, and context.

At OnTheMove, we’ve seen first-hand how much value is released when CRAs are equipped with the right information at the right time. The path beyond SDV isn’t about removing people from the process — it’s about empowering them to do the work that technology can’t.

About the author

Duncan Scattergood is Managing Director at OnTheMove Software and has worked with Clinical Trial Management Systems (CTMS) for over 10 years. OnTheMove for Veeva enhances the Site Monitoring process by presenting the CRA with the information they need, when and where they need it. This improves monitoring quality and reduces the time spent navigating multiple systems and performing report write-up.