Solving the Data Management Challenge in Digital Clinical Trials

Richard Young previously explained on ACT why sponsors need a strong data foundation for successful digital clinical trials. Now he focuses on how to address some of the associated data management challenges.

Improving the use of technology in clinical trials is now a prime opportunity for the life sciences industry. With the right tools, sponsors can enhance the efficiency of trial execution, provide patients with new ways to participate in studies, and simplify collaboration with clinical sites. An integrated digital strategy that is centered around these distinct user experiences could exceed the sum of the individual parts.

Many pharma companies introduced virtual elements to clinical studies to maintain patient access during the COVID-19 pandemic, from remote consent and patient monitoring tools to video assessments. This decentralized approach proved popular with patients and is likely to expand. Research indicates that 95% of companies could use decentralized trials in future, compared to just 28% before COVID-19.1 Boosting patient convenience and improving the site experience are clear advantages (see Figure 1 below).

But decentralized trials also create significant challenges for sites around data privacy, collection, and reporting (see Figure 2 below). Site users, who are already shouldering heavy workloads, are under intense pressure as they adopt new technologies to manage data collected from multiple offsite sources.

Addressing shortcomings in the data journey is key to unlocking better patient- and site-centricity. The data challenges will increase as decentralized trials become more prevalent and require oversight of a myriad of overlapping sources. Without a way to aggregate and harmonize data in one place, the pace and accuracy of decision-making during studies are likely to deteriorate.

‘Decentralized’ trials: New definition needed

If we want to create the conditions for enduring change in our industry, we must first agree on a definition that stretches us to deliver value to every consumer and contributor of data. The FDA’s definition of decentralized clinical trials is commonly used and high level: “Those [trials] executed through telemedicine and other mobile/local health providers that use technology and processes different to traditional clinical trial models.”

I do not believe that this definition sets the right standard for our industry. Following its logic, using electronic consent or electronic Patient Recorded Outcomes (ePRO)—now standard in many clinical studies—is sufficient to meet the criteria, even though these solutions do little to drive real change when deployed in isolation. We need to set more ambitious goals for patient- and site-centricity.

There is also the risk of innovation fatigue, which could result in the phrase ‘decentralized trials’ being devalued through overuse. Recall the industry’s journey with ‘adaptive’ trials since the 1980s. Although the concept was well documented, its practical use took some time to land effectively. With the release of the PhRMA and FDA definitions, and with regulators actively encouraging the application of non-traditional designs, we saw more and more protocols use the word ‘adaptive’ in their titles, irrespective of protocol content.

The same journey is underway with ‘decentralized,’ and it is important we recognize that now and hit the reset button. It is simply not enough to insert the word ‘decentralized’ into the description (or add one tech component to a project) for a trial to be considered fully decentralized. Truly adaptive and decentralized designs are created before protocols are even authored.

An alternative interpretation is to consider ‘decentralized’ as where clinical data is collected, for example, in a caregiving setting, a patient’s home, or a remote location. ‘Digital’ is better understood as how data is gathered and managed. If clinical data can be collected in decentralized, independent silos, but then managed centrally, sponsors and CROs will not have to compromise on the speed or accuracy of decision-making.

A more ambitious and patient-centric definition could then be: “Digital trials are those that are executed according to the wishes of the patient, from one day to the next.” This sets the standard higher, as such flexibility is only possible within a context of operational excellence.

Building the foundation for digital trials

When data management is undertaken in silos, it is done manually. Conversely, centralizing data management creates opportunities to automate the acquisition, cleaning, and reconciliation processes—which could be considered working digitally.

The right digital strategy can empower genuine site- and patient-centricity. But its development requires us to address three challenges: localizing to patient and regulatory requirements, delivering complete and concurrent clinical data cleaning, and automating critical operations.

Sponsors can expect local rules around decentralized capabilities at sites and need to incorporate these into their trial designs. For instance, not all countries accept eSignatures, so a mix-and-match approach will be essential when designing a clinical study to avoid geographical limitations on patient recruitment. The therapeutic area and patient profile are also highly relevant to which technology to use: an acne study with tech-savvy teenagers is different from a Stage 4 pancreatic cancer study with patients over 70. The latter cohort may require a hybrid approach to connecting with their caregivers, depending on how they feel each day.

Second, decentralized trials mean data will be collected from sources that did not exist (nor were accepted by regulators) a few years ago. The traditional model, in which most data came from electronic data capture (EDC), is shifting. Think of the millions of health data points now captured on Fitbits and Apple Watches, or other e-data sources. As technology advances, trial designs need to keep pace. However, as some data points may be inconsistent or invalid, sites must retain the ability to view and respond to incoming queries, outside of EDC.

Digital trial processes enable us to automate operational decisions, which in turn can improve efficiency and site-centricity. A typical data journey could begin with data being captured on paper at a site, before entry into EDC, streaming into a safety system, and then being included in a monitoring report on adverse events, for example. Technology should be able to identify which scenarios require source verification, almost as soon as the data enters the system. By connecting user experiences to data flows, different teams can come together at the source, creating new opportunities for trial acceleration.

Digital trials need a comprehensive data strategy

Data is the thread that links the disparate elements of a digital clinical trial. The volume of data cleaning and reviews is substantial, ranging from patient recruitment and retention to health data collection and analytics. A decentralized approach raises the risks of siloed and disconnected decision-making, making it critical to have central oversight.

To be effective, we need to define a data strategy upfront. The aspiration should be for all clinical data to be aggregated in one platform, so every consumer and contributor of data can fulfill their roles in close to real-time. For example, if your role in a clinical study is to predict the likelihood of a serious adverse event, you will likely need patient data from multiple sources (e.g., labs, imaging, sensors) to accumulate in one central repository to be usable. Applications need to be connected by design if they are to solve this challenge.

Typical clinical trials use dozens of systems across data management, clinical operations, and biostatistics, which need to be configured, maintained, and integrated. To remain focused on the core objectives of optimizing the site and patient experiences, we will need to map out the data flow against the user experience for different functions (e.g., clinical operations, safety, supplies).

The industry is attracted to the ‘decentralized’ label today. That’s fine, as long as we remember it references a specific transition: from centralized to distributed data collection (the where), made possible by digital clinical trials (the how). It will be crucial to have a central platform that can facilitate rapid decision-making as trial complexity grows, and data flows in from multiple locations. Data collection will become decentralized, but data management should not.

Fewer applications, better connections

Realizing the potential of digital clinical trials requires clearer definitions for industry terminology. Otherwise, the conversation will be at cross-purposes.

Despite the potential of decentralized trials to improve site- and patient-centricity, there’s a long way to go before studies are executed according to the wishes of the patient. We are in a transition phase today, where the next steps will be to define a digital strategy that is underpinned by effective data management processes and tools.

Counter-intuitively, accelerating clinical trials may require a slower approach to integrating new technologies. To date, sites have shouldered the heavy burden of technology adoption. In future, sponsors and CROs will need to evaluate more thoroughly whether solutions optimize the user experiences of patients and investigators.

Fewer applications, with better connections between them, are good starting points both for digital clinical trials and effective data management.

Richard Young, Vice President, Vault CDMS, Veeva Systems

Read Young’s previous article on building data foundation for sponsors here.


  1. Veeva Systems, Digital Clinical Trials Survey Report, 2021