Can wearable sensors save the most vulnerable in a climate disaster?
Part II in a series of opinion articles about what is still missing in healthcare for it to become truly patient-centered.
It’s August 2021 in Greece — two weeks into a dire heatwave and wildfire emergency. While reading the news, I stumble onto an article written by a woman pontificating over why she won’t turn on the air conditioning unit when the temperature is 40 degrees Celsius (104 in Fahrenheit). Then she reminds us of the deadly heatwave of 1987 in Greece, during which more than 1300 people died in Athens. I was nine years old that summer. It was memorable, indeed.
Summers in Greece have always included several days over 35 °C for as long as I have been alive. The number of hot days, the temperature, and the duration have notably increased. And, while this story is not about how you feel about what is causing climate change, it is definitely about how extreme temperature fluctuations make it imperative that technology adapts to help vulnerable people manage their health.
Post-1970’s building materials in Greek cities were concrete walls and marble or ceramic floors tiles — perhaps a little wood in the bedroom, if you could afford it. In the summer of 1987, I remember the marble tiles in our apartment feeling warm under my feet. Athenian rooftops were flat and not frequently insulated and the only “ecological” upside to them was the solar water heaters. Everything else was, and still is, a climate change disaster in progress.
We didn’t have an air conditioning unit in our somewhat middle-class apartment, so sleeping was impossible in our fourplex, which absorbed the heat all day and blanketed us with it all night. Every hour, I would go to the freezer, fetch crushed ice cubes, sprinkle them all over my bed, and sleep on damp sheets with a fan. During the day, I was dizzy and lethargic, and I fainted frequently. The “summer fainting” continued for years since we didn’t have air conditioning at school or in most hospitals, either. At school, I was a regular at the nurse’s station after gym class.
Hydration, potato chips, and rest were all that were available to me, and I lost countless hours just lying down waiting for my blood pressure to normalize. For many years, the “diagnosis” I received was “heat-sensitive,” and several of my gym teachers didn’t hesitate calling me weak-willed and lazy. Sometimes, I would need three bags of intravenous fluids until I managed to urinate and I was constantly advised to “drink more water and salt.” The suffering went on for years, but it wasn’t just heat that tormented me — my blood pressure would crash after any kind of stress: extreme temperature, prolonged illness, or intense emotional situations.
It wasn’t until 2008 when I was living in Los Angeles that I received a diagnosis and a treatment regimen to help me manage my condition. My endocrinologist was a lifesaver, but he had no answers on how to titrate my medication other than “just take more as needed.” Easier said than done, but I had to figure out how exactly to do it after I spent a frustrating Saturday night at a bustling Cedars-Sinai emergency room. I was severely bradycardic so they did an ECG, yellow-tagged me, and sent me back to the waiting room. They told me that it would be at least four hours before someone could even stick an IV in me.
An hour of waiting later, I decided that dying at home would be a better option and so I walked three blocks back, determined to fix this or faint in my bed. “Just take more,” echoed in my head, so I started with my steroid. I was surprised that I needed a triple dose, and within an hour, it was like nothing had ever been wrong with me. Until my potassium dropped, that is, and I became extremely orthostatic. Then it was back to potato chips, and that created a cycle of other problems too painful to describe. “Just take more” was not the correct remedy.
I always wondered why my case was so complex to manage. As I searched online, I realized that millions of people with autoimmune, endocrine, cardiovascular, renal, and autonomic problems were suffering from similar issues. Titrating medication is an art and a science, and it can only work if it’s personalized. In 2008, one would never tell a person with diabetes to “take more insulin as needed.” We had better ways to manage making calculations, but the trouble is that making calculations, even simple ones while you are not feeling well, is a recipe for disaster.
I won’t go into the muddy details about the greater endocrine disaster that is my life, but suffice it to say, I survive by tweaking a mineralocorticoid, potassium, and fluids 24-hours a day. For almost a decade after my diagnosis, I had my blood pressure cuff on-hand at all times to verify what I was feeling. The analog sphygmomanometer was eventually replaced by a digital one, but my challenges remained the same. While I was grateful for any device that could tell me what was wrong, it was not enough to help me titrate. Moreover, I had a deep-seated fear of becoming hypertensive or hyperkalemic by overmedicating.
By 2014, single wire electrocardiogram (ECG) sensors for consumers entered the market. I was intrigued. I purchased a unit for almost $200 but reading an ECG was very new to me, so I spent some additional money on books and around 100 hours reading online to learn how to decipher an ECG. The training was fun, but somewhat generic because I needed to know how all this was relevant to me, instead of learning about rare arrhythmias that would never happen. If they did, of course, I would obviously be horizontal and not fretting over how to use my device!
After six years with the AliveCor unit, I’ve slowly but surely learned a few things about managing my condition. I know all of my daily patterns, and experiment with my medication, fluid, and electrolytes. While managing my situation is difficult, I have developed a personal heuristic to troubleshoot my needs, and I can assure you it is more detailed than “just take more as needed.” The first thing I discovered was how to space out my medication. This resulted in the near-elimination of my migraines and premature ventricular and atrial contractions. Second, I found that I had a very high need for potassium and calculated how much I needed over what period based on the ST curve elevation of the ECG.
Finally, my fear of becoming hypertensive or hyperkalemic was assuaged. The only time that happened was when several kidney stones blocked my ureter on their way to causing a one-week colic episode (a situation far more memorable than the Greek heatwave of 1987!) I felt empowered to manage my health and kept myself out of the emergency room — except for the kidney stones, of course. Still, it was not enough. In the summer of 2018, I was on sabbatical in Crete and melting on my bed again without air conditioning and every morning, I awoke severely dehydrated with a migraine and had to slowly nurse myself back to life with the aid of all my devices and pills.
In my desperation, I took so much of my steroid that I started experiencing peripheral neuropathy in my legs. What I needed was something to wake me up when my blood pressure and heart rate dropped so that I could take my pills on demand — not pre-emptively and not in retrospect. There was no such device on the market. I complained to my very geeky cardiologist, who told me that what I was doing was pretty much all that was possible at the time. As a technologist myself, I felt betrayed by all the hype of digital health and sensor innovation. Where were the wearables for the sick, compromised, and disabled?
The consumer market is geared toward improving athletic performance, of course, and the most that consumer-grade wearables could capture was atrial fibrillation (A-Fib) via the inclusion of new sensor technology in the Apple Watch. While A-Fib detection is a precious tool, it is a tiny drop in the bucket. And if we can detect A-Fib, why not detect other things like ST-segment changes? It is all coming to market, but oh-so-slowly. The patent wars have already begun.
The truth is that health professionals and health systems are not prepared for the deluge of data. Researchers are wary of invisible algorithm changes. Many of these ECG changes are very personal, and a fair amount of patient education is needed to make sense of the data. While those aiming to improve their athletic performance are highly motivated to figure out their patterns, millions of people suffering from diminished health are left behind.
User-centered design practitioners are not strangers to developing personalized heuristics for individuals, even those with cognitive impairments or low literacy. There is an obsession with big data and digital health, but the real secret is closely observing small numbers of people. Doctors have done this since the beginning of time and effective health professionals hold in their memory dozens of heuristic patterns evolved from their practice or learned from research so they may respond in real-time to what is happening. These patterns shift with experience, and they can be biased depending on one’s patient population sample (and other reasons).
This is where more extensive trials with greater patient diversity can help (plus training reform). But, big data alone is no more or less prone to bias, just like machine learning. We are failing to use mixed-method designs in healthcare. Quantitative methods are great, but they tell us little about context transfer or about lived experience. And most importantly, they tell us nothing about the priorities and needs of an individual. Not unless the goal of the trial is to gather those needs and organize them into patterns.
In 2020, Samsung released the Watch 3, a watch that continuously measures blood pressure and performs an ECG reading. I ordered the watch in the middle of the pandemic with renewed hope and eagerly unboxed it. The first thing I tried to do was to customize the heart rate alerts. My core needs are banal: alert me if my heart rate is over 80 for longer than 10 minutes while idle, and wake me up if the heart rate drops below 60 while I am sleeping. I was disappointed that Samsung has fixed heart rate endpoints, so they cannot be fully customized, and customizations are mostly for aerobic training, of course. It felt like I was being insulted by my gym teachers all over again.
The final nail in the coffin was that alert types could not truly be customized in a meaningful way. Even the most potent buzz the watch produces is not enough to wake me up when my heart rate is very low. I need the watch to call my phone or someone who can wake me up, or be as loud as possible. Police-raid loud. My heart rate may slip as low as 40 in the middle of the night, and I frequently wonder if one day I just won’t wake up.
While this sounds dramatic, it is a reality I have to live with. If the watch cannot wake me up, I have little hope that it can also do other banal things like detect orthostatic hypotension or tell me how to manage my electrolytes. I tried downloading custom apps from the store, but none could fulfill even my most minimal needs. It seems I’ll have to code an app myself or pay someone to do it because what I need can apparently only be achieved by spending a lot of time or money on top of what I have spent already.
For the moment, the solution to the heatwave has been to buy an air conditioning unit and not leave the house at all. And you better believe that I keep my unit on. It’s set to 27 degrees Celsius — always. And while I experience guilt for having it on, I do feel grateful that at age 43, I have the privilege of affording air conditioning, a smartwatch, and my medication. Despite all that, last week, as the temperatures reached 42 degrees (108 Fahrenheit), I fainted from the heat after walking 50 meters outside for the first time in nearly a decade.
In the next 20 years, life on this planet is predicted to be quite dire. For people like me, hope for a better quality of life entirely relies on having electricity, climate control, and continuous sensing technology that can help us titrate. If the electricity is out, having a wearable technology that is more efficient may be the only chance I have to survive, and one day, technology may work similarly to a continuous glucose monitoring device to measure and adjust hormones and electrolytes. I don’t know if that will happen in my lifetime, but I often say that endocrinology is the final frontier of medicine. And the needs of people like me are not unique to endocrine disorders.
The pandemic has brought increased awareness to people with autonomic dysfunctions in long-COVID. Now lumped in with POTS, a syndrome hard to diagnose because patients were often treated like they were mentally ill, there is a renewed focus on managing these complex conditions. POTS patients need personalized assessment and treatment and these sufferers experience worse symptoms during temperature fluctuations and in extreme conditions. It took the demise of thousands from COVID for the so-called “potsies” to find some hope. Why is more suffering needed to fix suffering that is already there?
Unfortunately, in most of the world, air conditioning is still impossible, and it may never be possible, nor is it very sustainable technology in the long run. Wearables and computing technologies, in general, are not carbon neutral either. But, selfishly, I ask for a chance to survive. Whether you have a utopian, dystopian, or apathetic view toward the future, there is one truth that’s hard to dispute: personalized wearables can be lifesaving and improve quality of life. After all, I’m not asking for the ability to become an Olympic athlete. I’d just like to make it through the night. And, if I can regain some stamina, I promise to “athletically” plant more trees to help cool the planet.
Marientina Gotsis is an artist, professor, and academic researcher who has been consulting as a design and information technology expert in the non-profit and corporate sector since 1995. She has not received any compensation in kind or otherwise from companies that produce the products mentioned in this article.
Edited by Stacy Davies.