Robin Farmanfarmaian is a Silicon Valley-based entrepreneur working in technology and artificial intelligence. She has been involved with more than 20 early-stage biotech and healthcare startups, including ones working on medical devices and digital health.
With more than 180 speaking engagements in 15 countries, she has educated audiences on many aspects of technology intersecting healthcare, including artificial intelligence and the shift in healthcare delivery to the patient’s home.
She has written four books, including “The Patient as CEO: How Technology Empowers the Healthcare Consumer” and, most recently, “How AI Can Democratize Healthcare: The Rise in Digital Care.”
Healthcare IT News spoke with Farmanfarmaian to discuss where AI is impacting remote patient monitoring today and how AI can democratize healthcare.
- Where is remote patient monitoring today? Where do you see RPM five and 10 years from now?
- Remote patient monitoring is still in the first five years of adoption and integration into the healthcare system, and the pandemic accelerated this trend by illustrating the need and value of RPM. There are many clinical-grade devices now that patients can buy or use to measure and monitor various vital signs, including EKG, heart rate, heart rate variability, blood pressure and blood oxygen level.
The Centers for Medicare and Medicaid Services is one of the organizations that sets the standard of care in the U.S. healthcare system, and CMS launched CPT codes for remote physiological monitoring more than four years ago. CMS has expanded coverage and specificity over the past few years with additional and updated CPT codes.
In 2022, CMS launched CPT codes for remote therapeutic monitoring (RTM). These codes cover RTM for respiratory and musculoskeletal (MSK) conditions, such as remote physical therapy and COPD inhaler tracking. Considering that most of healthcare happens in a patient’s daily life, not the occasional clinic visit, this is a big step forward toward helping patients use their treatments in the best possible way on a daily basis.
Many mainstream corporations have launched their own wearables that have cleared the FDA, blurring the lines between healthcare companies and consumer-facing tech companies. Apple, Amazon, Google and Samsung are some of the giants that can shift consumer habits on a national scale, and they all have launched mainstream wearables.
For instance, the Apple Watch has outsold the entire Swiss watch industry multiple years in a row, and the device has an EKG monitor that has cleared the FDA for use with people over the age of 22 and with no history of arrhythmia.
This trend is great news because many people may already be tracking something about their health, whether that’s blood pressure monitoring, continuous glucose monitoring or even a simple accelerometer for step count. That makes it significantly more likely a patient will continue to use the device if their healthcare professional recommends it and has access to the data.
In 10 years, remote patient monitoring will be mainstream, and likely reimbursed by all the major payers. We’re already seeing that RPM has the ability to catch hospital readmissions days before they happen. The healthcare industry is experiencing a revolution in vital-sign measurement devices, with many companies innovating on ways to collect vital signs.
New innovations include taking vital signs using a smartwatch, using just a smartphone or laptop camera, breathalyzer devices for standard vital signs like BP and Sp02, sensors in clothing, epidermal sensors and subcutaneous sensors.
Within 10 years, tracking vital signs will be done in ways that are more seamless and effortless for the patient, such as subcutaneous sensors that last five years. Eversense already has an FDA-cleared implantable sensor for continuous glucose monitoring that passively records glucose levels 24/7.
- How did artificial intelligence first come into the picture with RPM? What was the connection?
- Some of these new FDA-cleared devices measure vital signs continuously, which means they are collecting thousands of data points a day on each patient. BiolntelliSense has a medical-grade rechargeable sensor that sticks to the chest and passively measures more than 20 vital signs, recording 1,440 measurements a day.
Humans don’t have the ability to analyze and interpret thousands of data points every day for every patient – which is why these clinical-grade wearables and sensors have an AI software component to manage, monitor, analyze and interpret the thousands of daily data points per patient. The AI software typically flags or alerts the healthcare team and patient when the vital signs are outside predetermined ranges, personalized to the individual.
While it is still early in this trend, there are examples of new innovations that only exist because of continuous, personalized data collection. January AI uses the previous three days of data from a continuous glucose monitor, combined with vital-sign data, to predict glucose response in real time to individual foods, educating the patient at the point of the decision-making.
This helps manage diabetes in a more personalized and predictive way, instead of the standard reactive way diabetes is currently treated. But January AI isn’t just for people with diabetes. They work with athletes, people with pre-diabetes or metabolic syndrome, and people who just want to be as healthy as they can be.
This education in real time doesn’t just assume the standard diet for diabetes is right for every individual or that there is any one healthy diet that works for everyone. People don’t react the same way to food as others, or even to themselves.
Everyone has a unique glucose response to food based on many factors, including that day’s activity level, sleep, amount of fiber, stress, weight, age and many more data points. AI-based software, combined with RPM, allows personalized care 24/7.
- Today, how does AI work with RPM to improve patient care and outcomes?
- When RPM is used for serious conditions, it can be the difference between life and death. VitalConnect ran a study on their single-lead EKG VitalPatch and was able to predict hospital readmission for cardiac patients 6.5 days in advance.
Alacrity Care is working on RPM for oncology that combines vital signs taken with FDA-cleared devices including the Omron blood pressure watch and the Oxitone pulse-oximeter watch with a daily oncology practitioner check-in and blood labs taken in the home.
This is to catch serious, life-threatening problems such as neutropenia, sepsis and cytokine storm days before a patient with cancer is in serious medical trouble. Catching these three conditions early can be the difference between life and death.
New AI-based software tools are clearing the FDA, including one earlier this year for TytoCare that analyzes lung sounds for the patient and the remote clinician using a connected stethoscope in the home.
There are other companies working on sensors in clothing that are covered by Medicare. SirenCare has socks available by prescription that monitor the temperature on the bottom of the foot.
For patients with diabetes, a hotspot on the bottom of the foot could lead to a skin ulcer, which could eventually lead to an amputation if the wound doesn’t heal. With access to the continuous data, the software can alert the patient and clinician when there is a problem so it can be treated before the skin breaks.
The promise and goal of RPM is to keep patients safely in their homes and catch problems early, before they become serious or emergency issues.
- You have a new book out with Michael Ferro, “How AI Can Democratize Healthcare.” How does that theme fit in with the combination of AI and RPM?
- When dealing with AI, life begins at 1 billion data points. There are some major problems with traditional healthcare datasets that exist today to train software. Most healthcare data is locked into silos, whether that is the EHR, faxes, the payer or in clinical notes.
In fact, when I get lab results from my physician through the hospital’s patient portal, it is uploaded as a scanned fax and saved as a PDF that is not machine readable, and sometimes, not even human readable. While we are seeing interoperability move forward, there is still a long way to go.
The typical healthcare data is collected on people at one point in time, such as their annual physical or if they are hospitalized. Frequently, that means the data doesn’t include an individual’s baseline, taken in their daily environment. It also means that most of the clinical-grade vital-sign data is on people who are already sick enough to be in a hospital.
By shifting the data collection to the patient’s daily life, RPM has the ability to collect clinical-grade data when people are in all stages of health and at all ages. When collected continuously in machine-readable databases, once RPM is more fully adopted, those databases have the ability to dwarf EHR data from a hospital or health system.
That is the type of training data that can give healthcare a much deeper look and understanding of normal vital signs across ages, genders and genetics.
RPM helps democratize healthcare in a way never before possible. Many people don’t live within easy access to a doctor or clinic. Trying to get to a clinic during their open hours can be next to impossible for some people due to many factors – from not being able to take off work, school, finding transportation, distance, childcare and other barriers, to traveling to a physical clinic.
Even for established patients, specialist doctors are frequently booked out one to three months in advance, which gives a medical problem time to advance and potentially get much worse. That, in turn, lowers the odds of a successful outcome when and if that patient is ever seen and treated by a healthcare professional.
Instead of trying to physically get to a clinic, RPM can be used to determine when someone needs to see a healthcare professional and can make a virtual care visit much more effective.
The best healthcare is the healthcare that actually gets done. RPM enables passive healthcare in someone’s daily environment, 24/7.
Original post: https://www.healthcareitnews.com/news/intersection-remote-patient-monitoring-and-ai