The Right Questions: Bayer’s Approach to Data Challenges and COVID-19
April 12, 2022 | TRENDS FROM THE TRENCHES | The way Jeanne Kehren sees it, pharma either has data graveyards, or “We are building data lakes without thinking about getting a boat.” Neither option advances the life sciences mission.
Kehren, SVP and CIO at Bayer, shifted into the pharma industry from veterinary medicine right after the human genome was sequenced. It was a heady time of vision and opportunity, she says, but she quickly saw the pitfalls of how data were being managed—or mismanaged.
Since then, Kehren has worked toward a new vision of how pharma uses data, what it means to ask the right question, and how the healthcare industry as a whole can serve patients. COVID-19, she said, has been both a magnifier of some of our challenges and an opportunity.
Kehren recently sat down with Stan Gloss, founding partner at BioTeam, to discuss Bayer’s approach. Bio-IT World was invited to listen in.
Editor’s Note: Trends from the Trenches is a regular column and now podcast from BioTeam, offering a peek into some of their most interesting case studies. A life science IT consulting firm at the intersection of science, data and technology, BioTeam builds innovative scientific data ecosystems that close the gap between what scientists want to do with data—and what they can do. Learn more at www.bioteam.net.
Stan Gloss: Before we talk about digital transformation, how did you get into this field?
Jeanne Kehren: I went to vet school, where my initial passion wasn’t IT but cows. There, I discovered viruses and immunology. One thing you need to know is that in terms of information strategy, viruses are masterminds. Think about how they store information in super small space and how they evolve—it's fascinating. I then found myself entering the pharma industry at the time of the first genome reading. We were all excited. We thought we had found the key to everything, to personalized medicine. I started to work on those data, metrics plus genetic and genomic data, and realized that those data would only talk to me if I was going to put them in contrast with clinically relevant outcomes. I also realized pretty quickly that we had a bit of a problem of data organization and data connection, and that it would never work if we were not thinking about our data and the entire metadata around it.
And you also need to think about your analytics. Because otherwise everybody is super excited about having data, but just pile up data. We pile up and up and up, but we never get the information out of the data. Then we would end up with data lakes without ever thinking about getting a boat.
Well, 80% of data lake projects have just been an abject failure because they were never really well defined and well designed for what they were supposed to do. Everybody thought if I told everybody we had 20 petabytes of data in a data lake, that would be impressive. But if you can't find it, it wasn't FAIR, it wasn't usable, what good is it?
Or you can have a beautiful, FAIR data lake, but if you don't have the right metadata, you can't do anything with that. And that's where I think the metadata definition, how you're going to characterize your data is essential, because what you need as metadata is defined by the questions you're asking. And that's exactly where everything we do in terms of data analytics, advanced analytics and data science, meets what we have been doing in the lab for decades. First, you need to ask the question, then you will define your experimental setting, and then you will order the lab material.
In the world of data, there persists this myth that because you've got a lot of material, the question will answer itself or will define itself. And then you come back to the famous paradigm of “The answer is 42, but what is the question?”
Or we can come back to the famous quote from Albert Einstein that says 95% of your time should be spent asking the right question.
When we went from having Sanger sequencing to high throughput DNA sequencing, it was almost a moment of punctuated equilibrium. There are series of times where we are going through evolutionary change, and then punctuated equilibria are moments where we go through revolutionary change. Could COVID be our next punctuated equilibrium moment?
Basically, COVID has been the catalyst and the magnifier, the catalyst because it led to the removal of some barriers in the adoption of technology, some of them purely psychological. You see in many countries that tele consultation, being able to reach a practitioner through video, is now common practice. It has been adopted by the entire systems, by the patients, practitioners, and payers.
It's not only a catalyst, but it has been a magnifier also showing us the pain points and inefficiencies of healthcare systems. Digital technologies are great at gaining time, removing frictions and COVID has been the magnifier pointing out frictions. First one is that we realized that it's not all about money in healthcare. It's about time. It's about time of qualifying healthcare personnel. And time is the most precious currency in our world and in the world of healthcare professionals.
We've got an increased demand in healthcare. We've got more and more people with complex conditions and with disconnected care. It's a problem of time for the patient. And it's inefficient when care is practiced in silos and the patients just pile up prescriptions. It ends up converging toward primary care affecting the practitioner. And those people are overwhelmed, completely overwhelmed. We never had that many burnouts in practitioners in the U.S. Never.
COVID has only now put the magnifying glass on that problem as well.
It's exactly that: Coming to realize that the challenge is to manage the time, manage the efficiency, tackle the silos.
Finally, in our current way of thinking about disease and health, we slice people according to therapeutic area, but a body is a system. We need to change the view. COVID helped to realize that the actual emergency is to make care more efficient, to break the silos, to change our point of view, to have high quality content in terms of medical and scientific knowledge but made accessible and actionable for the different stakeholders. And the patient is one of those stakeholders. Because if you're a patient, the first thing you want in life is stop being a patient. And if you cannot stop being a patient, what you want is to get as much as possible of your life back.
It's time, it's time, it's time, and being able to take back control and be the master of your time.
Right. But the fundamental problem is that what you design your system to do is what your system does. If you design a system to treat disease, you'll get disease, simple. If your mission is to promote wellness, you have to rethink the model. What is Bayer doing to transforming into a disease management company?
There are different things. First of all, I'm going to go back to the sustainability of health care systems. More demand, not enough supply – not enough healthcare practitioners, basically. The first thing you need to do to keep this manageable, is to empower the patient to be more a subject and less an object. The big revolution in furniture was IKEA, because by proposing that you do part of the work, Ikea allows you to get something more affordable. We have to allow the patients to do more, the people to do more, whether it's in terms of managing a chronic condition or wellness or prevention. New digital health platforms have the potential to integrate aspects of care and life, of health and disease. They can help individuals integrate that into their own life, not the life of the neighbor, but theirs specifically.
That's why we embarked on the collaboration with One Drop, knowing that they will bring the digital piece and all agility around thinking how you evolve your digital model. We bring the content, because information, valuable validated information is also very important to bring as a service to the patients. And we are working together also on analytics, how you derive insights, recommendations, personalization for people, that can help the people to stick with the system.
But think about it for us as a company, as an industry for pharma, what is this telling us? You're going to generate data where you have potentially progression or regression of disease and conditions with continuous measurements. For the moment the entire pharma industry and the entire field of medicine has been prescribing things on scarce discreet measurements.
By getting more precise, continuous data on conditions like hypertension, cardio, chronic heart failure, genetic disease, we'll get a much better understanding and potentially a re-definition of those conditions in the long term. I think this will be the second act after the leveraging of those digital health platforms.
If you want to start focusing on disease management, will you change the way you get paid?
This transformation is going to advance how we think about outcomes and lifelong burden in terms of care, consumption, and cost. We all agree that early diagnosis is key, that preventing a disease is better than treating or even curing a disease. However, we need to have other factors catalyzing the change because today the maximum funding goes to treatment, while the diagnostic and prevention businesses see a minority of the investments.
The entire healthcare industry is saying that we need to do more prevention and more diagnostics, but the entire system does not yet have the money where the mouth is.