The smart everything
Convenience, optimization and personalization are a huge part of our lives now, mostly to the better.
Our phones and all the smart devices around us are portals to convenience, instant satisfaction and a personalized experience which we grew to expect, always. I can upload my own art that will be printed on my physical credit card issued by a large bank. I have an app that allows me to adjust the temperature between shelves in my fridge and my smart thermostat knows when I am 10 minutes away from home and starts heating up the house just the way I like it. The lights in my living room adjust to my mood and I have a bicycle that was painted by a large manufacturer to my very specifications.
Yet, the moment I get sick I am back to the last century: I must first find a doctor (usually on a static database of doctors in my area), whom I need to call for an appointment. On this call I awkwardly dictate my name and details to an assistant. At last I get an appointment in three weeks. On the day of my appointment, I am using a pen to fill in onto a paper the very details I gave over the phone three weeks ago. Eventually, I get seen by a very busy human who asks me a bunch of standard questions (“How bad is your cough? “)and sends me off with a prescription – “take one of these after each meal and call me in 3 weeks if nothing changes”.
Why do I get a more customized experience from my fridge than from my doctor? Why is my vacuum cleaner “smarter” than my healthcare experience? Should we not expect a more personal touch with our health?
Longitudinal data
People all over the world are using wearables and apps on their smartphones to optimize sleep, diet and exercise. Smart watches are packed with sensors and they continuously track streams of data about biomarkers that used to be the exclusive territory of doctors’ offices and laboratories: heart rate, oxygen concentration, respiratory rate, VO2Max, glucose and, indeed, cough frequency.
Even as regular people become fluent in sleep quality, exercise, diet and heart rate data, health systems themselves don’t yet know what to do with all this data. This is because the whole system is designed around a concept of health data as a strictly static sample, as opposed to a continuous, dynamic stream.
We have gotten really good at studying drops of water, but have no concept of the proverbial river.
Historically doctors were few and far between so evolutionary this makes sense. The doctor visit was a rare and important event and all data sampling was done there, at the time: the doctor would study the patient, listen to their lungs, take the pulse and ask a bunch of questions:
“How bad is your coughing?”
Based on these data samples – also known as clinical evaluation – a diagnostic was given and treatment provided.
How did they know the treatment worked? Well, the symptoms would go away or, in the case of chronic diseases, people would learn how to live with the symptoms. And in any case, people would make the effort to call if anything was wrong, right?
Meet your smart companion
Now that we have sensors everywhere and are surrounded by computational powerful devices it is pretty easy to collect continuous data streams about our health and wellbeing. Meanwhile, increasingly competent AI models can process all this data and find patterns that can help us understand the relationships between our lifestyle and our health better and they can help us change those variables that increase our health and quality of life.
RELATED: Two Ways in Which Healthcare is Changing Right Now
Data savvy medical professionals can now look at data of thousands of patients. They can screen them before the first visit and they can monitor their treatment continuously, without having to actually see them in person. For the first time ever, doctors can truly scale themselves. They can run scripts and event listeners on all this this data to alert them to any outliers, sudden changes or meaningful patterns and they can learn more about how treatment and symptoms are affected by environmental variables and patients’ daily habits.
Medical relevance of lifestyle choices
This is one of my favorite things about smart companions. It is also the main reasons why I am tracking and quantifying as much of my life as I can. There are so many variables that I can tweak to optimize the impact of my exercise and diet. My overall quality of life, as well as my treatment plans.
With a similar mindset, doctors – with the help of powerful AI models – can now increase the efficacy of treatment and reduce adverse events without even messing with the drug itself. Some of the variables that can be tweaked include:
-
- Time of the day/ distribution of treatment dosage around the day
-
- Symptom/ therapy correlations with things like diet or sleep cycles
-
- Correlation with environmental factors – temperature, humidity, air quality
-
- Correlations with behavioral components
-
- The role of exercise
Input/ output
Doctors used to adjust treatment based on reported changes in symptoms, every several weeks or months. Basically, they’d prescribe a treatment and wait for a few weeks or so to see how it works. Then they would adjust based on a follow up visit or call . This simply doesn’t make sense in this day and age.
Digital companions provide a feasible way to see input/ output in real time, or as close to real time as possible. This means that right now we can use input/ output evaluations to increase effectiveness & reduce adverse event rates. Maybe even reduce treatment time.
One powerful way to track input/ outputs is symptom frequency and specifically, longitudinal symptom tracking – i.e. tracking symptom over time. Combined with AI powered pattern recognition models, doctors and patients can now work together to develop custom treatment plans for patients that can be adjusted dynamically based on real world data. Below is an example of a personal respiratory risk matrix (curtesy of Hyfe.ai) that uses cough frequency as a variable and correlates that with environmental factors as well as aggregate community-level frequency data (to control for things like seasonal flu).
Being able to customize and personalize treatments, and learning more about treatment effectiveness in conjunction with behavioral variables would not only reduce the cost of care, but it will also unlock an additional wave of innovation.