In this Lohika Re:think event, we featured a panel of digital health experts who explored innovative and emerging technologies in healthcare. With the emergence of AI, machine learning, IoT and the cloud, the way consumers interact with healthcare providers is changing.
Our panel discussed what they have seen in healthcare innovation today and what we can expect to see tomorrow.
Panelists Raman Frey (moderator) Sanjiv Narayan, Professor at Stanford University Noosheen Hashemi, Founder and CEO at January.ai Leo Janze, Head of AI and Data Engineering at Change Healthcare Anant Gupta, VP Engineering at Grand Rounds Health
Discussion Points The role of AI, machine learning and data science in digital health innovation How the pandemic has changed patient care The emerging technologies that will impact health care in the coming years How to assemble the health care innovation dream The future of digital health
Lohika and Capgemini’s Applied Innovation Exchange (AIE) co-hosted an online event titled “Leveraging Innovation and Emerging Technologies in Digital Health.”
Innovation in digital health The panel was asked to answer this question:
*“What does ‘innovation in health care’ mean to you?”
For Anant, a relative newcomer to the health care industry, innovation is about bringing familiar features from consumer technology into health care. Anant gave the example of getting pizza or buying sneakers. He can quickly search online and find thousands of reviews from hundreds of customers.
After reading those reviews, he can make decisions on which pizzeria to choose or which sneakers to buy. If he needs to find a doctor, Anant says that the experience is completely different. Rather than finding him the doctor who’s most qualified to address his condition, the health care system routes him to doctors based on insurance coverage, zip code or availability.
Why has health care lagged behind on information accessibility? In response to Anant’s comment, Raman asked the panel why the health care industry has lagged behind on information accessibility.
Sanjiv chalks it up to “medical conservatism,” or the need to avoid mistakes. If software fails, the application crashes. With health care, on the other hand, a mistake can cost someone their life. As a result, the industry is forced to put limitations and guardrails in place and that includes access to information.
In response, Raman asked if this aversion to risk is antithetical to rapid innovation.
Nosheen says “yes,” in a sense, but blames the system. She says that what seems like inefficiencies to some are profits to other companies. Some companies will intentionally propose custom one-off implementations that take years to build and cost billions of dollars. The custom nature of the project is intentional, as customization is profitable.
Further discussion: innovation in health care For Noosheen, innovation in health care means the complete and total consumerization of the system. Instead of treating people as patients, we should treat them as users, members or consumers, with 100% access and transparency with their personal health care data. Noosheen says that people should see and own their health care data and become invested in it.
Noosheen notes that the financial markets have done a good job with data transparency, but health care has not. The prices of drugs, procedures, and services are opaque and not publicly available or disclosed.
Leo notes that health care data is a mess. With patient records in a multitude of PDF files and photocopies, Leo notes that 80% of patient data is unstructured. For Leo, innovation in health care is using technology to turn unstructured data into structured data that’s understood by computers and applications, something he’s working on at Change Healthcare.
Sanjiv notes that nobody knows what an early warning system for cancer looks like. Nobody knows who’s going to get a stroke or a heart attack in the next month or next year. Sanjiv notes that you can combine wearables with algorithms and AI to find early warning signals for patients. He notes that this is what Noosheen is working on at January AI.
Raman asked the panel about data sovereignty and the portability of personal health care data. Should consumers be able to provide their data to a clinician, then revoke that access as needed? Anant says that yes, the system should work this way. He can get his credit history from five different companies who will sell him access, but can’t figure out the last time he got a flu shot.
Leo had a slightly different take. While he agrees that patients should have access to their health records, he thinks it’s best in the hands of a clinician. Leo says that he has access to his financial data, but has a professional accountant do his tax returns. In the same way, he wants his health records to be handed to a clinician, who uses that data to provide decision support.
Leo says that when you see a new clinician today, the intake form lists 10-20 questions — the same questions you answered for a different clinician last month. The health care system should only ask you to provide this data once. From there, you grant certain clinicians access to it.
Technologies on the horizon The panel was asked what emerging technologies will impact health care in the next 1-3 years. Sanjiv spoke about computational analytics and modeling, two things he’s actively working on to study cardiac health. In addition, Sanjiv believes that augmented reality and biometric sensors will become a core part of the virtual visit experience.
In a video-based virtual visit, the clinician will be able to see a patient’s heart rate, breathing rate, and temperature. The clinician can see if the patient is stressed by a change in their vascular perfusion. To make this scale, the industry will need to agree upon standard metrics and platforms, says Sanjiv.
Noosheen says the use of wearables for personal health is exploding. Apple Watch is more and more a health device. Garmin integrated with Dexcom to allow users to monitor their blood glucose levels. By measuring things like VO2 max, blood pressure, caloric burn, and resting heart rate, wearables are quickly becoming the future of health care and preventive care.
Noosheen notes that the key is the proper synthesis and curation of data from wearables. AI can help with the interpolation of the data and trend analysis, something that humans are incapable of doing at this volume of data.
~~Current projects from the panelists
Raman asked the panelists to describe the projects they’re currently working on.
Noosheen said that January AI is enabling its users to take off their Continuous Glucose Monitoring (CGM) device and continue to receive predictions on how they’ll respond to different foods. This new feature is based on having the AI learn from the data it’s modeled from the user.
Sanjiv is using a combination of AI and modelling from data they collect in the clinic to identify “subpartitions” that create personalized models for patients. The result is that each patient has a personalized recommendation on the best way to treat their condition.
Anant is using data to navigate patients through health care to improve their clinical outcomes. Grand Rounds Health helps people get to the right health care outcomes and reduce unnecessary medical visits by 25%. In addition, Grand Rounds Health helps change the treatment plan that was initially recommended, which leads to better outcomes 70% of the time.
Leo is taking PDF charts and running them through Optical Character Recognition (OCR). The data is then run through a machine learning model to predict diagnoses for a patient. Leo gave the example of a patient experiencing tremors in the upper body for the past four months. His team’s software used machine learning to generate five predictions of diagnoses.