Clinical Innovators: Using Mobile Tech and Artificial Intelligence to Improve Cardiac Rehab
Trishan Panch, MD, MPH, is the chief medical officer of Wellframe and clinical strategy adviser at the Commonwealth Care Alliance. He has practiced medicine for 13 years, studied health policy and management at Harvard University, and worked as a researcher and lecturer in health sciences and technology at the Massachusetts Institute of Technology. His work involves using information systems to optimize complex care processes, empower patients, and create more efficient and responsive health systems.
What is Wellframe?
Wellframe is a mobile platform for chronic disease management. We are a health IT company based in Boston. In a sense, Wellframe gives patients a GPS navigation system for their health that offers turn-by-turn guidance to help them recover from acute episodes, better manage chronic conditions, and avoid secondary complications. The Wellframe system is personalized for each and every patient based on their unique characteristics and adapts according to how they interact with their care plan. Wellframe integrates into clinical workflows and uses the data generated from a patient's interaction with their care plan to inform more intelligent human resource allocation and optimize population health.
Despite the dizzying array of health care apps available, most have limited real-world functionality in the context of care delivery. This is why we are fundamentally focused on using all our experience as clinicians and with patients to build technology that can be woven into the very fabric of care delivery to amplify existing clinical relationships. This has involved looking at the real needs of patients and clinicians, then building technology to meet these needs.
What has been your career trajectory leading up to the creation of Wellframe?
I was a primary care physician in London for 10 years, running a practice in a fully capitated environment. I've always felt that one of the many reasons that health care is interesting is that the problems are typically multidimensional and therefore require multidimensional solutions. I really strongly felt that a seamless fusion of computer science and clinical medicine could greatly improve patient care. This would require cooperation across disciplines, and in 2009 I met the group with whom I would start Wellframe.
We all come from a pretty diverse set of backgrounds and our paths collided due to our involvement in a couple of the home-brew health innovation groups at Harvard and MIT. Jacob Sattelmair (our CEO) received his PhD in epidemiology from Harvard and was a product manager at RunKeeper; Vinnie Ramesh (our CTO) is a mathematician and computer scientist from MIT; and Archit Bhise (our head of product) is also a computer scientist from MIT whom I worked with on developing the open-source Sana Mobile Health platform.
How did you get the idea to work in cardiac rehabilitation?
Our interest in cardiac rehab was the fusion of two streams of work. I had been working on ICU data mining in the Lab of Computational Physiology at MIT and Jake, my co-founder, had been working on consumer technology at RunKeeper. When we started talking as a group about trying to use a computational approach to improving preventive medicine, we quickly realized two things. Firstly, it was clear that such an approach was not possible without a structured granular data set of patient's interaction with their care plan across all the dimensions of the care plan. Such a data set did not exist and it was clear that we would need to make it.
Secondly, it was clear that mobile devices (smartphones and tablets) were the best means of delivering care plans to and collecting data from patients. Once we figured this out, we began looking for an area of preventive medicine where there was a standard of care and broad consensus around this standard. We quickly found cardiac rehabilitation.
The more we looked into cardiac rehab, the more fascinating it was. Not only was there an extensively validated standard of care, but there was clear evidence of clinical and costs benefits across populations and over long time frames. Alarmingly, given the evidence and endorsement of cardiac rehab by all the relevant professional bodies, only about 25% of eligible patients attend clinic-based rehab. Incentives are not as yet aligned around expanding cardiac rehab participation given the high cost of delivery. This makes it the perfect clinical area to work on using technology to bring what is known to work to more people.
How does Wellframe improve cardiac rehabilitation?
Wellframe improves cardiac rehabilitation by taking what we know works and using mobile devices, artificial intelligence, and process re-design to bring it to more people. This means that we have been able to drastically increase the amount of contact between patients and clinicians and make it easier for patients to take control.
There are four parts to the Wellframe platform that are optimized for cardiac rehab.
- Firstly, the protocol of cardiac rehab has been modeled and then represented in a logic structure. This means that it is machine-readable and can be used to instruct a computer program to deliver the right tasks to the right patient at the right time.
- Secondly, we have customized our mobile app that gives the patient a daily multimedia to-do list with all their cardiac rehab tasks for that day, a searchable library of approved cardiac rehab content and a secure messaging channel to connect cardiac rehab clinics to their patients.
- Thirdly, we have a HIPAA-compliant cloud data collection environment and care-routing engine.
- Finally, we integrate into provider electronic medical record systems or provide a bespoke care management dashboard for clinicians.
What we have learned is that in cardiac rehab, as in all medicine, what works is not just the protocol of care but the connection with a clinician in a cardiac rehab clinic. This will not come as a surprise to clinicians, of course, as we all know that the practice of medicine is as much about building relationships as it is about reorganizing disordered anatomy, physiology, and biochemistry. At Wellframe, we want to use technology at the service of these therapeutic relationships in cardiac rehab to amplify care.
What are some of your early results?
We recently completed a small observational trial of the Wellframe Cardiac Rehab program as an adjunct to clinic-based cardiac rehab with two academic medical centers here in Boston. Cardiac Rehab patients received a validated program on their cellphones and were given daily feedback from a cardiac rehab nurse who they also saw in clinic. Study participants were between the ages of 40 and 80 and completed a 30-day program during the second phase of their cardiac rehab. The findings were presented at the national meeting of the American Association of Cardiovascular and Pulmonary Rehab and have been submitted for publication. As the paper is under consideration for publication, I am not able to discuss the results in specific terms now but they will be public soon.
We were really pleased that patients of all ages—in particular older patients—used the app regularly and tracked physical activity, medication usage, and also engaged in regular contact with their clinic. Staff reported that via the Wellframe dashboard they had a window into patient's lives outside the clinic, and they could work with patients to make lifestyle changes in a way that they could not have before.
We appreciate that this is an early observational study but we are really encouraged by the findings. In addition to ongoing studies in other clinical verticals, we are pursuing grant opportunities for a multicenter randomized trial of Cardiac Rehab using the Wellframe system versus standard clinic based cardiac rehab. We are really excited about what is coming in 2014 from a clinical validation point of view.
Do you think the approach you have taken has value in other therapeutic areas?
Absolutely! At the same time as we are scaling our work in Cardiac Rehab, we are working with leading academic medical centers in Boston and New York to deliver programs for patients with CHF, chronic obstructive pulmonary disease, and even for the rehabilitation of patients with serious mental illness. In addition, we are due to start working with a national pharmacy benefits manager to serve a publicly-insured Spanish-speaking population with chronic disease and behavioral health comorbidities.
I realize that the needs of patients almost never dwell in discrete therapeutic categories. We had to be focused in terms of early implementations, but it is really exciting to see the programs coalesce. Wellframe is now being used to combine clinical protocols across disease verticals to meet the needs of the individual patient. We have always been and continue to be interested in using mobile technology and artificial intelligence to build the first mile of the health system for people with chronic diseases rather than just working vertically in one clinical area.
What are some of the barriers and obstacles to implementation?
We have focused on looking at our technology like any other therapeutic such as a drug or medical device. This is considered quite a conservative approach in the consumer tech world where the rhetoric seems to be about disrupting the health care system. As a clinician, I believe that whilst change is absolutely necessary, wholesale disruption of the health system is not necessarily in the best interests of patients. It is better to work with the system to bring groundbreaking technology to the frontlines of patient care. The problem is that introducing technology in this context necessarily and appropriately requires extensive evaluation prior to roll out.
Clinical trials take a lot of time and require a skill set not often found in the tech sector. For a start-up, pursuing clinical validation invariably means raising more money, earlier (i.e., more expensively) and of course incurring more risk. These are significant obstacles that can stand in the way of even the most committed tech entrepreneurs. We have deliberately chosen to take this harder path rather than releasing an app that will confuse rather than help patients and clinicians. As I said before, our aim is to weave technology into the very fabric of care delivery.
What do you see as the future of Wellframe?
I believe that now we are looking at the "Henry Ford moment" for medicine. By this I mean as vertiginous advances in computational processing, mobile devices, and artificial intelligence collide with improved understanding of the genetic basis of disease and the interaction between genes and the environment, we are faced with the prospect of an "industrial revolution" in medicine. At Wellframe we want to be part of this.
In other areas of the economy, industrialization has already taken the work of artisans and enshrined it in processes executed by humans and machines working in symphony that have been able to achieve greater productivity at higher quality and lower cost. However, even in spite of advances in medical science, the practice of medicine remains bespoke, reactive and notoriously costly. What this means is that individual clinicians, like artisans before the industrial revolution, create therapies de novo for each individual patient in response to that patient's individual complaints and largely through the medium of face-to-face consultation. On the one hand, this is Cadillac care. However, the costs of practicing in this way are dizzying and the well-documented variability in quality and safety is unacceptable.
Wellframe is a platform for mass customization. We allow scheduled care to be personalized and delivered to patients using their mobile devices. Through patient interaction with the care plan, we dynamically segment the population and deliver a customized experience for each group of patients. Some people will get video information, and some will get text. Some will get programs with lots of caregiver interaction, and some will have more social programs.
In our next phase of trials with cardiac rehab clinics we are extending the concept further. It is clear that most patients don't go to cardiac rehab, but there are a number of reasons why they don't go. We want to investigate whether we can offer a spectrum of cardiac rehab options to patients with app-based support to address their particular needs. Through the app they could gain access to cardiac rehab options through mobile delivery or in person, or through a hybrid model. Patients could have access to this rehab not just over the 4 months after they are discharged from the hospital but potentially over their remaining 10 to 15 years of secondary prevention! We are really excited about the potential we have to use this technology to improve health and quality of care, at a lower cost and over a longer period of time.
< Back to Listings