February 13, 2019 | Data is king and Kyle Holen, head of the development design center at AbbVie, sees great promise in how it can be applied to clinical trials. But there still challenges ahead of us. Interoperability, for instance, has not been solved in clinical research, he says. And ensuring data of appropriate quantity and quality still stand in the way of the best use of machine learning and other artificial intelligence technologies.

But even as we explore exciting opportunities afforded by new technologies, Holen stresses that we mustn’t lose sight of our priorities.

Clinical research, “should be about: Number one, what’s best for patients. Number two, the quality of data that you collect. And three, I think should be cost. In that order.”

On behalf of Clinical Research News, Marina Filshtinsky spoke with Holen about new avenues of data capture and what a data-focused clinical trial means for the patient voice.

Editor’s note: Marina Filshtinsky, Executive Director of Conferences at Cambridge Healthtech Institute, is planning a track dedicated to Clinical Data Strategy and Analytics at the upcoming Summit for Clinical Ops Executives, SCOPE, in Orlando, February 18-21. Holen will be speaking on the program. Their conversation has been edited for length and clarity. 

Clinical Research News: What is required to create data-powered infrastructure for the next-generational clinical trial?

Holen: Well, I think that there are multiple opportunities that need to be explored for the ease of data transfer. One of the biggest, of course, is interoperability. And many, many people are working on this. And I think it already has been solved in some environments, but certainly needs more effort. If we have inter-operability of healthcare records, that will benefit all of us. Including those of us that are conducting in clinical trials.

We know that companies are trying to come out with a single data collection hub. What are the challenges and advantages of such an approach?

The advantages are that you have all your data aggregated in a single place. Such that you can learn across studies and within studies, and maximize your learning from the data. But, there are a lot of challenges to this; it's not easy. My understanding is that we have a long way to go to make sure that all of our data, truly, is integrated. But, we certainly are taking smaller steps, and the efforts to create these data hubs is a step in the right direction.

How is digital health changing the way clinical trials are being conducted? Does it affect the duration and the overall cost of the trial?

Oh, that's a huge topic. Digital health is integrating into so many different aspects of how we conduct clinical research. From finding the appropriate patients using big data to wearables, and sensors to better digital data transfer methods. There are tons of different aspects where digital health is making an impact. And, I think in all of these different areas of clinical trials, even including some of the analytics that we're doing on the data that's coming in, all of those can have an impact on the length of time it takes to conduct our study in different ways. Selecting the patients that would more rapidly have your desired endpoint certainly will have a huge impact in being able to close your study, lock your database, and start analyzing the data.

From the very beginning of the process, where you're selecting the patients that will more rapidly accrue endpoints. To the very end of the process where you have efficient means of transferring data. And even in the middle where you're using digital technologies to capture these endpoints; all of these processes can help to efficiently streamline our clinical trial process.

Are you seeing any direct impact on the cost?

Well, if you can conduct studies faster that will definitely have an impact on the cost. But, I'll tell you that cost shouldn't be the main driver of why we implement digital technologies into our clinical research paradigm. I think it should be about Number one, what’s best for patients. Number two, the quality of data that you collect. And three, I think should be cost. In that order.

Another huge area of interest is artificial intelligence and machine learning. What are the three challenges you see that stand in the way of harnessing the power of AI in clinical trials.

Well, clearly, the biggest problem that stands in the way machine learning and AI techniques is having the massive amount of data to feed your algorithms. Machine learning is only as good as the data that you feed into the machine. And in most cases, we don't have the millions and millions of data points that are necessary for the machine to adequately learn something and give us insights. First and foremost is the quantity of the data.

Secondly is the quality of the data. So, if you don't have good quality data to feed the machine then you might get mislead by whatever algorithms come out of the process. So, those are two biggies around the data.

I think a lesser challenge is just having the expertise. There are a lot of people who are experts in machine learning and Artificial Intelligence; those people, typically, are not people that understand the pharma industry well, or healthcare for that matter. So marrying the people who know the pharmaceutical industry with those who understand machine learning algorithms, that would be critical.

With all of this new technology, is there a space for patient’s voice? Are we moving towards more patient-centered trails or the opposite direction?

Oh gosh! Well, there's always a place for the patient voice. And I think all of us who are involved in clinical research need to make sure that the patient is first and foremost in our thoughts when we're trying something new or designing a clinical trial.

I do believe that we're designing studies that are more patient focused. And, I think that the future will allow for very direct patient involvement in the clinical trials. Either thorough the wearables and sensors that they use or through better control of their personal health data. Where patients can decide what components of their health information are shared with their doctors, or with the pharmaceutical partners who are conducting clinical research. As patients have more control over their personal health data, I think they will then have more control over how research is conducted, and how much data they're willing to share with people who are conducting those trials.