Digital health thought leader shows how patient centricity benefits clinical trials and patient outcomes
Keynote Presentation I: Digital Health: Reimagining Healthcare with Patients at the Centre | Watch Video
It may be hard to see because we are still in the midst of it, but Pistoia Alliance conference keynote speaker Cristina Ortega Duran, Chief Digital Health Officer R&D for AstraZeneca, believes we are “living through the largest transformation in human history.”
During her speech, Duran cited statistics that predict the digital health market will reach $600 billion by 2026. “The pandemic has changed behaviors, the way billions of people interact with health care,” she said.
Instead of asking individual patients to move around to navigate health care, Duran said, with new digital technologies and tools, health care will place the patient in the center and adapt to their lifestyle. In clinical trials, 90% of data was collected from sites. Now, with clinically validated technology, it is possible to collect 70% of data at home. AstraZeneca surveys showed 75% of people, both patients and physicians, now prefer virtual health care visits for non-urgent care.
"It will be holistic, not just physical but also mental health care," Duran said. "And hopefully it will use data analytics to find the best pathway to get someone better."
Already, digital health tools are shifting the focus from late-disease management to disease prevention, she added. This has led to prolonged survival for oncology patients, as well as improved quality of life.
Clinical trials are becoming more patient-centric and efficient as well. AstraZeneca surveyed clinical trial participants to get their feedback. By doing most of the assessments at home now through a platform called Unify, patients have less travel time and less time away from work.
For the pharma industry, making clinical trials more patient-centric can have huge benefits, Duran said. "In some of the studies we are currently running, by having less physical visits you can get more patients and you can accelerate these studies by nearly 30 percent," she said. "It can reduce the cost by 20-40 percent because you're doing it in a more efficient way."
Adding in new technologies, such as artificial intelligence, contributes even more to efficiency. For example, AI can employ algorithms to adjudicate adverse events in clinical trials -- a painstakingly slow job that has historically been done manually by physicians.
Improving the patient experience also leads to better outcomes, Duran said. “In this new era of targeted precision medicine, we all play a role in creating the patient-centric future that patients deserve.”
In one case study, proactive and continuous digital monitoring of symptoms in cancer patients was used to catch early signs of lung cancer progression in patients, which allowed physicians to take preventative steps earlier, in between regular health care visits.
A few challenges remain: one is always being vigilant about patient privacy. The other is making sure clinical trials are using global data and data that is consistent across platforms and nations, Duran said: “Only by sharing will we be able to improve human health."
Keynote Presentation I: Digital Health: Reimagining Healthcare with Patients at the Centre | Watch Video
- Cristina Ortega Duran, Chief Digital Health Officer R&D, AstraZeneca
Keynote speaker outlined pandemic uses for machine learning in pharma
Keynote Presentation II: Digital Health: Discovering the Governing Equations of Medicine | Watch Video
Pistoia Alliance members who attended the Pistoia Alliance’s annual European conference learned, first-hand, how cutting-edge tools like artificial intelligence and machine learning helped a group of physicians give their Covid-19 ICU patients exceptional care in a way that has never been available in the clinical setting before.
During her keynote speech, entitled "Using ML to Discover the Underlying Models of Medicine," Mihaela van der Schaar, PhD, John Humphrey Plummer Professor of Machine Learning, AI, and Medicine, University of Cambridge, outlined a novel approach to Covid-19 patient care that brought machine learning to bear on an important challenge during the pandemic: the cytokine storm.
Covid patients with severe infections, who were hospitalized in the ICU, frequently had an overly-aggressive immune response to the virus, which caused inflammatory proteins called cytokines to flood into the bloodstream. While this immune response helped the body fight the Covid virus, it also resulted in organ damage to the kidneys and liver. Patients who survived, often ended up having chronic kidney disease and the possible need for ongoing dialysis.
Van der Schaar's lab at University of Cambridge came to the aid of physicians who were attempting to prevent this organ damage with the use of a glucocorticoid called dexamethasone. Research studies performed during the pandemic showed dexamethasone kept more Covid patients off of ventilators, by reducing their inflammatory lung infiltrates, and it also significantly reduced the mortality of Covid patients in the ICU.
By using a Latent Hybridization Model (LHM), combining a pharmacological model with additional expert clinical resources and data, LHM gave providers the ability to more precisely administer the dexamethasone doses. Machine learning was brought to the hospital bedside, to help clinicians and pharmacologists better understand the way a patient's immune system has its own unique response to the steroid.
“Better markers of the immune response, such as cytokine 1, can be measured in the research laboratory, but it is not typically measured in the clinic," van der Schaar said. The model showed clinicians the various interactions between such variables as the inflammatory marker C Reactive Protein (CRP), the dexamethasone concentration in the patient's lung, the viral load and the patient's immune response.
The additional data provided by the machine learning hybrid model gave the physicians an important new tool that helped them to better titrate the use of dexamethasone and refine their treatment plan.
In the end, the lab's use of machine learning was aligned with the growing trend of patient-centric health care. “With this model, we were helping physicians understand their patients and they were personalizing the decisions to their unique characteristics,” van der Schaar said.
Machine learning models in pharmacology and life sciences can also be used for medical risk predictions, epidemiology and to show models of disease progression, she added. They may also help scientists gain better approval from government regulatory agencies for novel biomarkers in oncology and other specialties.
Van der Schaar invited conference attendees to participate in her lab's virtual "inspiration exchange" sessions with industry partners and academics by visiting www.vanderschaar-lab.com.
Keynote Presentation II: Digital Health: Discovering the Governing Equations of Medicine | Watch Video
- Prof Mihaela van der Schaar, John Humphrey Plummer Professor of Machine Learning, AI, and Medicine, University of Cambridge; Director, Cambridge Center for AI in Medicine
Janssen's digital health catalogue helps researchers find new tools, benefits patients
Keynote Presentation III: Digitally Derived End Points and Evidence Generation | Watch Video
During his keynote speech at the Pistoia Alliance London conference, Kai Langel, Senior Director of Strategy and Innovation at Janssen, showed how his company is perfectly aligned with Pistoia's mission to collaborate and innovate.
Langel outlined a digital health initiative called DEEP (Digital Endpoints Ecosystem Protocol), which, may soon come under the Pistoia Alliance umbrella in order to expand and offer more resources to the life science and pharma industry.
The DEEP project offers a unified data platform with important information about digital health tools and ways to measure patient assessment data.
“I've always felt very passionate about collaborating together to advance things we care about," Langel said. The time is right for a project like DEEP. “There's a lot of venture capital in digital health right now. It's starting to bloom and grow," he said. “This will allow us to discover how patients function."
The project ties into cutting-edge trends, such as personalized medicine and patient-centric drug development, he added.
With more clinical trial assessments being done remotely since the pandemic, Langel said, scientists need this catalogue of different digital health tools. DEEP combines searchable information on different measuring devices with evidence-based studies on their efficacy and links to specific algorithms and data that shows how well the devices work.
“This is very useful to connect researchers with different aspects of health, so they can measure all of these different variables,” Langel said. “It's quite a rich catalogue and we're proud of what we offer here. I think this will be a fantastic tool.”
During his talk, Langel outlined a specific example of how the DEEP resource, which is currently used internally at Janssen, can be used for a scientist who is trying to gather precise data for Parkinson's disease.
In the past, Parkinson's patients were difficult to measure because they have a great deal of tremors. They also have memory and cognitive issues, so they were not the most reliable partners. By searching for Parkinson's tools in the DEEP catalogue, researchers could discover a new way to measure movement using electromyography (EMG), which measures electrical signals in muscles.
By tapping into this new, unified resource, scientists will be able to participate in the patient-centricity trend and create a new way to benefit this patient population. The EMG tool will help physicians better understand the relationship between patient tremors and the timing of their medication doses. By doing so, they can change the morning dose, for example, to minimize tremors later in the day.
"These measurements bring a ton of value," Langel said.
Looking to the future, he feels DEEP provides the perfect opportunity to bring Pistoia members around the table to collaborate and innovate by using coordinated data sets and sharing best practices and information. “There is a lot of synergy and sharing," he said. “We want to make a real impact with real value. It's time to come together and move forward.”
Keynote Presentation III: Digitally Derived End Points and Evidence Generation | Watch Video
- Kai Langel, Senior Director, Strategy & Innovation, Janssen
Keynote speaker outlined pandemic uses for machine learning in pharma
Keynote Panel Discussion: Application of Patient Centricity in Therapeutic Drug Discovery and Development | Watch Video
During a lively keynote panel discussion at the Pistoia Alliance's annual European conference, experts discussed the need for the pharma and health care industry to rethink the language used to discuss people with health conditions and how patient involvement and insight can lead to better outcomes. This thought-provoking discussion concluded with a story about how, in the end, a dog can sometimes be the best medicine of all.
The panel members, led by Pistoia Alliance Advisory Board member Bryn Roberts, Global Head of Data and Analytics, Roche Diagnostics, agreed on some new terminology. Instead of the patient, panelist Birgit Bauer, Digital Health & Social Media Expert, Manufaktur für Antworten UG, said she prefers to be called a woman living with multiple sclerosis. Birgit Bauer also admitted she dislikes the term “patient journey," and prefers to see her experience as more of a pathway.
“I'm sorry," Birgit Bauer said, “I'm not on a trip. It's not an all-inclusive holiday in the Caribbean sea." That is why she calls her popular blog, “Almost Normal: My Life with MS." “To live with a chronic disease means to live an extraordinary life," she added. “It's not a normal life, it's just almost normal because the disease affects me every day, from time to time. It's not a journey, it's a path. It's a way through my life."
Instead of patient centricity, panelist Katherine Capperella, VP, Global Patient Engagement Leader for Janssen Pharmaceuticals, Johnson & Johnson, said her company encourages staff members to use the terms patient engagement or patient involvement instead of patient centricity.
“We need to rethink a little bit about what language we use," Roberts said. “In turn, that will also transform our mindset."
Changing the language and terminology is just one way people living with diseases are having an important impact on health care, said panelist Thomas Hach, Executive Director, Patient Engagement Cardiovascular, Renal & Metabolism, Novartis, and a physician by training. Since he has also participated in clinical trials as a patient, he has a unique perspective from both sides of the pharma and health care industries.
"When I was a doctor, I thought patients didn't have a clue and they had better listen to what I say. But it's actually the other way around," Hach said. “Patients will be the biggest disrupters in health care, now and in the future especially. We better listen to them, and really listen to them."
Birgit Bauer agreed that the first priority is better communication. “Sometimes you feel like a number. You're lying in the bed and doctors are standing around you," she said. “They are discussing about you, but they don't discuss it with you."
Meanwhile, clinical trial participants willingly give the pharma industry the data they need to develop new drugs and treatments. Hach explained the need to circle back and show the participants the impact and value that the data generates. And, it's important that the information is explained in simple, understandable language, Birgit Bauer added.
"When you live with a disease, you want to learn something about that disease, you want information," she said. "We have to make decisions about treatments, we have to plan steps for our future. Every piece of information can help us to make better decisions."
By simply listening, Novartis has made its clinical trials more human-centric, Hach explained. For example, in a leukemia study, clinicians felt they needed to do five repeated bone marrow biopsies to get the most accurate data they needed. When asked for feedback, participants responded loud and clear: they said, “No way, five bone marrow punctures in one year? It's not going to happen... " Patient feedback had a positive impact on trial recruitment.
Putting the people living with the disease first is the best way to bring all of the collaborators together, focused on the right priorities, said panelist Patrick Lample, Principal Specialist Solutions Architect, Healthcare and Life Sciences, Amazon Web Services.
After discussing ideas around how the Pistoia Alliance could help the industry unlock data for the benefit of patients and collaborate even more broadly with other patient engagement initiatives, the panel discussion ended with Hach's story about how listening to a patient and the patient’s family can change a physician’s intervention.
As a physician, he informed one of his patient's insulin was the best way to treat his diabetes. The patient's wife had an alternative treatment plan in mind: adopting a dog. Three months later, the patient returned to his office, 15 pounds lighter, with a good HbA1c score. The dog provided an outlet for exercise and emotional support.
"A dog isn't in the medical textbook, but it solved his problem. The person living with the disease proved me wrong," Hach said, having learned a lot from this experience. “So, that is why we better listen to the people because sometimes they know so much more than we do."
Keynote Panel Discussion: Application of Patient Centricity in Therapeutic Drug Discovery and Development | Watch Video
- Chaired by Bryn Roberts, Global Head of Data and Analytics, Roche Diagnostics
- Birgit Bauer, Digital Health & Social Media Expert, Manufaktur für Antworten UG
- Katherine Capperella, VP, Global Patient Engagement Leader, Janssen Pharmaceutical
- Thomas Hach, Executive Director Patient Engagement Cardiovascular, Renal & Metabolism, Novartis
- Patrick Lamplé, Principal Specialist Solutions Architect, Healthcare and Life Sciences, Amazon Web Services
