It has taken a decade of scientific integrity and rigor – and a very particular mindset – for Cognivia to develop its current suite of innovative solutions. But there is a sense that we are just scratching the surface.

Anyone who has experienced firsthand the frustrations that come from knowing a compound should work – but not being able to demonstrate efficacy – will understand the driving force behind Cognivia’s beginnings. The first attempt to solve the unsolvable.

It started with a realization and a hypothesis: Clinical trials only ever considered biological data. Patient psychology and its impact on patient behavior was never taken into account.

Could the personality of a patient be the key to understanding that patient’s response and behavior when participating in clinical trials? Could integrating this information have an impact on drug development?

As Congnivia celebrates its 10-year anniversary, we are taking a look at some of the founding principles and how these pillars continue to drive the decisions we make as a company. We’ve already talked about Cognivia’s vision. In this article, we’re taking a deeper dive into the scientific rigor that has led us to this point. How an innovation mindset, the right team, and a determination to work through challenges have powered Cognivia’s successful growth.

An Innovation Mindset

Dominique Demolle, Ph.D., Chantal Gossuin and I were overseeing drug development at a major pharmaceutical company when we came up against a common problem: placebo effect. Usually, placebo response is treated as a hurdle that decreases study power. Instead, this time, we decided to think about the entire idea in a different way.

Instead of looking at placebo response as something to work against, we decided to work with it. Instead of trying to reduce its impact by working around it or training it away as usual, why not characterize what will remain anyway by taking into account its mechanisms to improve the way clinical trials data are analyzed?

In 2013, the three colleagues launched Tools4Patient – now, Cognivia – took inspiration from scientific academia, and began a deep exploration of one of the most accepted “unsolvable” clinical trial challenges.

We then tackled the question of how to empower drug development and steer it in a new direction with new information – specifically, never-before-quantified patient data. This included conducting our own clinical trials to test our hypothesis and building the first models that could predict the placebo response, but this would require thousands of patients over multiple trials. Then inspiration hit. We could use machine learning to improve the hypothesis and models to predict the behavior of the patient.

We quickly realized that by incorporating the behavior of patients in clinical trials, we could better understand some of the mechanisms and have much more powerful tools – so we focused on leveraging disease-specific AI-based predictive models to deliver patient psychology insights.

Challenges to Innovation

One of the first steps toward realizing the vision was to convince investors and the industry that a problem historically deemed unsolvable was actually solvable using what at that time was a “controversial” new technology. AI and machine learning technology had not yet been popularized to the extent it is today. But even the best idea doesn’t make a company, so building the right infrastructure – from data management, software, quality and IT platforms – was prioritized, even on the startup budget.

And then – perhaps most importantly – the founders had to put the right group of people behind this novel idea. Cogniva was dedicated to cross pollinating mathematical engineers specialized in machine learning and drug development experts to find the most appropriate solution.

The top requirement was the ability to see not just five years down the road but twenty years – and to understand that, especially in drug development, it takes a long time to make real change. It also required risk tolerance, the ability to be controversial at times, and a willingness to set out with a small group of people to change the way drug development, medicine and patient care worked – not something many people would bet their livelihood on.

With a properly dedicated team on board who believed in the same vision and were ready to solve the unsolvable, Cognivia was ready to build their first clinical trial solution for predicting each patient’s placebo response: Placebell.

Future commercialization was always the goal, which also meant prioritizing customer experience and simplicity of use when it came to any new technology: To add value without adding burden. This continues to be the north star against which all solution development is measured.

Evolution of Ideas

By 2018, machine learning and AI technologies had significantly matured in the public eye. At the same time, trials were facing more challenges with variability than ever, and the push for more personalized medicine and decentralized clinical trials was increasingly trending.

As early adopters of machine learning and AI, Cognivia was well positioned to respond to these circumstances with well-established technology and a novel approach to enable more patient-centric clinical trials.

Placebell was successfully validated across multiple therapeutic areas and was being deployed within numerous clinical trials – and that success meant two things. First, that the hypothesis was correct: patient personality is critical to understanding variable response to treatment. Second, with this newfound understanding of the patient at an individual (as opposed to group) level, there were opportunities to integrate these insights into strategies for data adjustment, patient screening, and targeted recruitment and engagement.

The next phase of growth for Cognivia was to take the lessons from Placebell and see where else they could be applied – all while remaining true to the initial vision.

Again, we started with some critical questions and hypotheses: If we can predict placebo responsiveness, what else can we predict? Can we predict which patients in a clinical trial are actually going to adhere to the medication until the very end?

Nonadherence in clinical trials and high rates of dropout make efficacy and safety more difficult to determine, and, like placebo responsiveness, has been a notoriously difficult-to-solve issue.

Thus, was born Compl-AI, a way to predict treatment non-adherence and dropout risk. Whether a patient completes the demands of the clinical trial or not, whether they are unlikely to take their medication as prescribed, and whether they respond to placebo, are all influenced by the patient’s psychological makeup and surrounding environment or origin.

Looking Toward the Future

There are still many aspects of patient behavior data that can be explored within the framework of clinical trials. From patient/physician interactions and site variability to length of trial, patient diversity, and even health literacy. However, Cognivia sees a future that includes a way to better inform every interaction with a patient.

Though Cognivia has developed very powerful tools, the true innovation isn’t necessarily the technology. The most innovative aspect of what Cognivia has achieved is in adding value to the drug development process and helping ensure the best drug candidates are accessible to the people that need them. Technology is simply a tool. How we use the technology is where we can make a real difference.