Digital Health Stack

Medicine is on a profound journey, similar in scope to the transition made in the 1930’s and recounted in Lewis Thomas’ seminal work “The Youngest Science”.  At that time, physicians shed many of the horrific practices from the 19th century and embraced the scientific method.  Our understanding of physiology, began to match our knowledge of anatomy and modern medicine was born.  The discovery of antibiotics, notwithstanding the serendipitous nature of discovery  involving mold floating through an open lab window,  powered a sense that health, illness, life and death were now the realm of human understanding.  

At the turn of this century, the human genome was sequenced and a new journey began in earnest, albeit more slowly and involving more complexity than was imaginable.  This movement is progressing   to a complete molecular understanding of health, wellness, illness and disease.  The underlying mechanisms of human physiology are being illuminated and we are casting off the assumptions of the last century.  Drugs that we have thought of as “hypertension” drugs are simply drugs that work on particular biochemical pathways in the body; useful in hypertension but perhaps similarly so in other conditions or circumstances.  Diseases that in the 20th century were thought to be worlds apart, like coronary artery disease and pemphigus (a rare and severe dermatological condition) are found to be closely related at a genetic level.  Our division of specialities by anatomy, will look awfully quaint when we look back at the last hundred years.  For example, conditions treated by gastroenterologists concerning our gut, run that gamut from cancer, autoimmune, endocrinological, and now increasingly microbiological as we learn more and more about the effects and consequences of the 10 trillion bacterial residents of our microbiome.

What powers this transition in medicine, is a what I call the new digital health stack.  It starts with the digitization of everything:  

• Molecular:  genome, epigenome, transcriptome, proteome, metabolome. 
• Anatomical: Digital CT, MRI, Functional MRI, PET, Ultrasound
• Physiological: external biosensors capable of measuring every aspect of our physiology in real time at very low cost.  Developing internal biosensors capable of measuring both molecular and physiological.
• Lifestyle – sensors that track activity in all aspects – exercise, diet and even personal exposure to harmful substances
• Environment – sensors that monitor environmental changes in terms of pollution, air quality,  UV index, solar activity
• Clinical records – finally the linkage to traditional clinical outcomes from clinical records that have been digitized over the past 20-30 years

What follows this mass digitization will be data aggregation, putting together datasets that for the individual will provide a complete picture of what they experience and the interplay between all of these aspects of what determines our health and wellness.  This data aggregation will be powered by big data techniques and  will start with correlative discoveries that will ultimately need to be hardened by science.   The potential for this discovery, will lead to digital health platforms where patients seeking to fully understand their health will demand their data.  

Data aggregation will provide the substrate for artificial intelligence and the creation of predictive, diagnostic, therapeutic and prognostic algorithms.  The platform with the most users, with the most data will benefit from the most innovators building the most useful algorithms.  As in everything digital, this virtuous feedback loop will exert “winner take all” pressures to the market and accordingly there will be a very limited number of digital health platforms.

Algorithms will initially be supported by human expert workflow, initially as part of a virtual care encounter, but evolving from being a part of every encounter to more of a quality control role and eventually disappearing entirely.  These algorithms, once functioning autonomously will require compelling user interfaces, with profound understanding of human behavioral change built in to their design.

For many health and wellness needs, people will of course continue have the need to see doctors and health practitioners, but even these encounters will be strengthened and improved by the integration digital health stack.
​
Finally the stack is driven by the complete capture of outcome information, fed back into the platform and available to improve the algorithms.

​This emerging digital health stack foretells a future in health care where a few digital health platforms emerge to disrupt, demonetize, dematerialize and ultimately democratize health care globally.  How long will this take?  It probably depends on where you live and the strength of  the counteracting forces (regulation, health care culture, entrenched interests…etc.) in your location.  In parts of the world that have not developed modern health care systems,  the shift to digital health may come first, in ways similar to the developing world’s adoption of cellular technology without first building land lines.

Over the next few months, I will continue to develop and expand this thesis, identifying, unpacking and validating the forces that will combine to determine the trajectory of digital health.