Nothing like it had ever happened before – a global slowdown in new study starts and a massive decrease in trial accruals for open studies. The turn of events was a shock for sponsors and providers. Sudden disruptions on that scale rarely hit multiple players in an ecosystem at the same time. But that’s exactly what happened at the intersection of clinical care and clinical development during the pandemic. Researchers and clinicians were forced to change how they operated trials, and technological solutions that researchers had been slowly exploring for trials were suddenly adopted at rapid speeds.
We recently spoke with Jeff Elton, CEO of ConcertAI, a healthcare and life sciences AI and real-world data (RWD) solutions provider. Elton is acutely aware of the impact that this rapid adoption of technology has had on trials through the work ConcertAI does to support and enhance trials. Here are his thoughts on what’s to come.
The pandemic dramatically impacted clinical trials around the globe. How have you seen trial operations evolve as a result?
Jeff Elton, CEO of ConcertAI: One of the biggest differences was that ‘people’ become less material to the process. Site monitors, other CRO personnel, regional medical liaisons, and other clinical operations personnel could not visit sites. As a result, many novel approaches – AI for patient identification for study eligibility, decentralized trials, etc. – were put into the foreground and taken out of ‘pilot mode.’ Complementing this was greater use of telehealth and related tools by the providers. It is exceptionally rare that a shock hits multiple players in an ecosystem at the same time – but that’s what happened. This ‘shock’ put AI and digital solutions in the foreground and that suddenly made entirely SaaS-based digital solutions less risky than conventional approaches.
As a result, there are companies that have advanced policies and mandates to be entirely ‘digital.’ This is huge – the industry is asking to transcend the legacy model and move towards new approaches that can work more efficiently and are less subject to disruptions, such as the pandemic.
It was not that everything worked as intended. Rather, the move from ‘pilot’ to ‘production’ showed us the difference between interesting concepts and new ways of working. For example, a research team can’t use multiple eScreening solutions at the same time. So having multiple vendors and solutions deployed ends up being counterproductive. Some of these early solutions are being terminated and single solutions are being put into place across all studies and sponsors. Similarly, decentralized trial solutions had the value of being deployable in remote settings, such as the patient’s own home, but patients needed to be seen for tests and imaging studies. Even more salient, patients wanted to see their physicians and did not want to be engaged in trials without that closer contact and surveillance. So, we know we can capture data remotely, but we also know that digital solutions within a facility will likely be the hub of the study and be favored by both patient and provider/researcher.
How does RWD play a role in new trial operations?
Elton: By the close of next year there will not be a pharma or biopharma organization without large-scale RWD datasets that guide and inform their trial designs and that become the basis of their sites of emphasis for a single trial and across studies in a therapeutic area. ‘Multi-confederated’ data solutions bring together clinical, medical claims, social determinants of health, and lab data in ways that allow new AI and machine learning approaches to predict trial performance, accessibility, and the likelihood of a site performing. This is also not a static process, meaning it’s not one-and-done. The standard of care and outcomes change, results at sites change, etc. Consequently, these data and tools should be used over the life of the study to make ongoing adjustments and tightly manage alternative approaches. For example, if a new entity launches during the trial conduct period a sponsor can design a parallel real-world evidence (RWE) hybrid as a complement to RCT control to better inform the interpretation of incremental therapeutic benefits and relative safety.
RWD and RWE are also informing trial designs and endpoints. The goal for trials is to reflect the relative safety and efficacy of a new therapeutic entity, versus the standard of care, and do so with as few patients as possible and as quickly as possible. Larger studies and slower completion times mean less relevance to the trial results relative to the standard of care. As a result, RWD and RWE are informing new trial endpoints that can be cross-correlated. Some of these bear the term ‘surrogate,’ not implying they are not real, but rather they are formal biomarkers that now stand in replacement of former ones. Our own belief at ConcertAI is that the new ones stand to provide more utility to sponsor trial designs, decisions to progress into the clinic, and to regulators assessing the outcomes.
How, if at all, has RWD/E’s use changed since the onset of COVID-19?
Elton: We are getting closer to the point where all studies use RWD to inform their trial design. Most of the oncology development leaders use RWD early in the process to better understand the standards of care for subpopulations, their outcomes, the deficiencies of current therapies, etc. Now there are large data science and trial analytics teams that use RWD at scale in AI SaaS solutions to optimize trial designs, assure patients can complete the study, inform which sites have the infrastructure to participate, assure the burden on sites and patients is comparable to the standard of care, among many related considerations.
RWD is also used to assure appropriate trial diversity, an FDA mandate and goal, that looks at the ethnic and racial populations most negatively impacted by disease and seeks to optimize a design and find sites that can accrue to statistically meaningful numbers for those subpopulations. Finally, it is used to validate heritage sites for their ability to perform, and more importantly, to find new sites that the sponsor has not worked with but which may play significant roles in total patient accruals. Again, this is all in service of better designs, more broadly deployed to higher-performing sites.
What kinds of technologies are more commonly used in the design and execution of trials now than were before the pandemic?
Elton: There are five new tools of the trade:
– Large RWD datasets that have been assembled, linked and engineered specifically with coverage of 20 to 50% of a country and with a clear minimization of biases in regions, racial and ethnic subpopulations, urban versus rural, etc.
– SaaS solutions with AI optimizations for study design and the ability to predict site performance with very high accuracy
– Patient matching AI tools, acting on EMR data, using NLP, connecting lab data, etc. within the workflows of the trial sites – the broader the use and the more ubiquitous, the more useful
– Novel consenting solutions and data linking tools that allow trial designs to have a follow-up period using EMR and Medical Claims data
– Digital study execution tools that can now use clinical data sources to populate a study and sponsor-specific eCRF library and write to a targeted EDC
How have you seen researchers and clinicians evolve the ways in which they run trials since 2020? Are these changes ones that you expect will remain in play post-pandemic?
Elton: There are shocks with reversion back to formal models and shocks that inexorably change models. What happens as a consequence of the pandemic is the latter. There will be no going back to the legacy models. Look at the behaviors of CROs, EMR companies, medical distribution companies, and legacy clinical trial software companies. They are pivoting their business models to digital-only solutions that work on a foundation of RWD but bring large-scale AI-enabled SaaS solutions forward as the basis. However, this is not their natural domain – this is a shift to sustain relevance. The power of the new solutions is their integrality to care institutions – they don’t need the labor and third parties that were required before.
For sponsors, they further allow model simplifications, greater productivity, and direct contracting and operational relationships with provider research enterprises.
This is why our model is to partner with two ecosystems – providers and sponsors – and evolve in lockstep with both around this new paradigm and models. There is exceptional value for these systems – greater trial accessibility for community providers; greater trial access for patients; faster trial execution for sponsors; better generalizability of trial results to the general population and for different subpopulations; the ability for biopharma innovators to fund more studies given these efficiencies; and, the ultimate value of more medicines to patients more quickly in order to assure the best possible outcomes for the greatest number of patients.