This year marks 200 years since French physician René Laennec invented the stethoscope, one of the more recent pieces of technology to be incorporated into the standard by which we assess our health in 2016: the annual physical exam.
Many doctors agree the standard physical is suboptimal. During the last several decades, numerous studies have failed to find a connection between periodic health evaluations in healthy people and reduced mortality. This is in part because many of our deadliest diseases do not produce easily noticeable symptoms in their earliest and most treatable stages. The solution is not to do away with check-ups, but instead transform the primary care visit itself.
The founders of modern medicine didn’t have the tools to non-invasively explore our physiology in detail, but today we do, and our capabilities are rapidly advancing. We have the technology today to eradicate treatable disease as a cause of death. Unfortunately, our healthcare system hasn’t evolved to incentivize methods of care that work toward this end.
We don’t need to settle for a system that allows people to die in order to fit backward-looking actuarial calculations. It’s time to stop viewing the economics of healthcare as a zero-sum game and completely reinvent primary care as a prevention-based science: one that measures much more about our bodies, realigns cost incentives, and no longer depends on simple, single variable point-in-time measurements of symptomatic individuals for predicting disease.
It is possible to build a preventive and personalized healthcare system that gets cheaper and more effective over time and is accessible to everyone. In the US we already have a pretty good example showing us how to get started.
Dentistry: Today’s Model for Prevention-Based, Personalized Medicine
Dentists today are taught that they are the frontline of healthcare because they see patients so much more frequently than primary care physicians. The majority of Americans see their dentists about twice a year, and much more information is collected in single visit to your dentist than during an annual physical to assess the health of your entire body. There are a number of diseases related to our general health that dentists have been able to correlate to our oral health. This isn’t because our mouths contain the best predictors for these diseases, it’s because we have more longitudinal data on the health of our mouths than the rest of human physiology combined. Our dental care system boasts millions of longitudinal data sets of unbiased measurements tied to outcomes, creating a self-optimizing information feedback loop. It is not a coincidence then that dental care by many metrics has gotten cheaper or stayed flat in inflation adjusted dollars all while our general dental health is improving. Over the same period of time, the cost of our healthcare system has skyrocketed and threatens to bankrupt our country.
Aside from the frequency of measurement and volume of quantitative information collected, our dental care system also does a good job of de-conflating the concept of a “diagnosis”, into two separate concepts: immutable measurements and mutable analysis.
Dentists didn’t invent the concept of longitudinal tracking of dynamic systems in order to better understand them: this is a basic part of the scientific method and has been the cornerstone of most major scientific discoveries in human history, so why don’t we use it to better understand our bodies?
We need to improve our understanding of human biology as a system of dynamic and highly interconnected subsystems and the only way to do that is to measure more about ourselves, more frequently, not less.
Preventing Preventative Medicine
Our healthcare system has many components, none of which is likely to be intentionally malicious. Yet, the health care system as a whole incentivizes providers to avoid using technologies we know would save lives — all in an effort to save money in the short term.
Doctors are not to blame here. It’s hard to point the finger even at the insurance companies that want doctors to perform fewer tests. It is a fact that tests can lead to more tests, some maybe resulting in invasive procedures that produce worse outcomes and generate higher costs than if no initial test was performed at all. This is in part because many of our best clinical, non-invasive diagnostics are not sufficiently accurate to use in asymptomatic people as determined by the actuarial models used by insurance companies. There are other common reasons cited for why we don’t use our existing tools more frequently that raise fundamental questions about our rights as patients to have access to information about our bodies, which is worth a separate discussion entirely.
But there is a clear path towards continually improving our accuracy in predicting pathology at its earliest stages, and not just known pathologies, but also pathologies we aren’t even aware of yet. We need to improve our understanding of human biology as a system of dynamic and highly-interconnected subsystems and the only way to do that is to measure more about ourselves, more frequently, not less.
We live in a world where we use millions of variables to predict what ad you will click on, what movie you might watch, whether you are creditworthy, the price of commodities, and even what the weather will be like next week. Yet, we continue to conduct limited clinical studies where we try and reduce our understanding of human health and pathology to single variable differences in groups of people, when we have enormous evidence that the results of these studies are not necessarily relevant for each and every one of us.
We cannot hope to have personalized and preventative healthcare, then attempt to understand our health in terms of generalized population averages alone. The majority of preventative tools our doctors practice medicine with were developed based on assumptions that ignore two unchanging facts about human biology. First, it is a long tail distribution. The combination of genomic and environmental factors that we are exposed to during our lifetimes is unique for every human that has ever lived, even identical twins. This should force us to reconsider how representative a “representative population” could ever be in a clinical study.
Second, human health and pathology is based on non-stationary patterns. This simply means that as our population changes — due to our technology, our nutrition, and the environment around us — the relevance of a clinical study will likely decrease over time. There is scientific evidence that simply moving to a different place during your life can create significant changes in your body’s physiological processes which, in turn, affect your health. Evidence of the deficiencies in the current methodology is clear when you look at the false positive rates for the gold standard diagnostics for many of our most common fatal pathologies, as well as how difficult it is for most clinical studies to be reproduced on other “representative populations”.
Gaga for Genome Sequencing
Despite what popular trends in the media claim, whole genomic sequencing of our germline DNA is not the key to personalized or preventative medicine. In fact, there are relatively few known diseases caused exclusively by our germline DNA. The sequence of nucleotides that make up your DNA is relatively static; however, the environment within each cell that the five trillion copies of our DNA sit in is highly variable. This variation can dramatically affect how the same sequence of nucleotides may be interpreted and in turn affect your health.
By some estimates, your physiological state at any point in time contains roughly 10¹⁸ (that’s a million trillion) times more information than resides in your genetic code. This represents an enormous amount of information and complexity that we are continually accumulating over our lives and is encoded into our physiology, the vast majority of which lives outside of our genes. Having your genome sequenced will not tell you if your family has been exposed to toxic drinking water, nor how badly you injured yourself in a fall, or even how a recent surgery or change in medication affected your health. Genome sequencing cannot determine if you are healthier this year than you were last.
We believe our health cannot be determined by our genomics alone, and the things that it can be used for in isolation are relatively small. Our genome is our programming, but no one knows what a program will do unless they know the inputs to it, and even then it is unclear whether this problem is decidable. They key to personalized preventive medicine is about tracking what is changing in your physiology. It is understanding these changes, the inputs to our programming, that gives us the context in which our genome is most useful in medicine. Simply stated, right now we need to know the inputs to our programs more critically than we need to understand the programs themselves.
The genome itself was inferred to exist and discovered because we measured how visible population traits changed over time. If we were armed with enough longitudinal physiological information of detailed changes in our physiology, we would be able to infer the existence of certain genetic variants even if we didn’t know DNA existed. It is then reasonable to assume that the key to unlocking the secrets of the genome starts with longitudinal tracking of detailed changes in our physiology.
Your doctor will then use this as their primary tool to make personal health forecasts the same way meteorologists use weather simulations to tell you whether or not to pack an umbrella next week.
The Future of Primary Care and our Healthcare System
In the near future, the standard physical will consist of increasingly cheaper and high-resolution snapshots of your physiological state generating terabytes of information gathered within minutes. These snapshots will not only provide a form of backup that can be used to “debug” your health in the future, they will also dramatically increase the accuracy of our tools for predicting pathology well before a person becomes symptomatic and often most treatable. Eventually, these snapshots will be used to create personalized simulations of your physiology, which will replace the notion of a healthcare record entirely. The preventative checkup will be done at intervals dependent on your personal risk factors and current state of health.
In this paradigm the primary purpose of the preventive primary care checkup will be to refit your physiological simulation with your current physiological state. Your doctor will then use this as their primary tool to make personal health forecasts the same way meteorologists use weather simulations to tell you whether or not to pack an umbrella next week. These forecasts won’t just project the trajectory of your health based on your current health state. Your doctor will also be able to make forecasts about your health based on hypothetical changes in your diet, exercise or even medications in order to recommend changes that are optimized to meet your personal health and quality of life goals.
As a result, there will be a shift in how we view the macroeconomics of health care. We will no longer see a dollar spent on healthcare as a dollar lost, but as an investment in keeping a person a healthy and productive member of society, that, on average, has positive long-term economic returns. As we become an increasingly information-based economy, the scope of human knowledge increases at an accelerating rate, and so will the economic cost for society to raise and educate a person to the point that they are productive. Losing people to treatable diseases too early will be not just tragic because of human suffering, it will be considered bad for our economy.
When we are of age, we not only will be given the opportunity to be organ donors, we will be given the option to be data donors. Almost half of the adults in the US choose to be organ donors. For those of us who opt in to be data donors, when we pass, all the data that was ever measured about our bodies will be anonymously uploaded into a public database. If you are an organ donor you can save one life. If you are a data donor you can not only help save the lives of those you leave behind, you can help save the lives of every person who will ever be born thereafter. This will result in an accelerating positive feedback loop for humanity, where every generation is healthier than the last, and treatable diseases no longer take lives.
While all of this may seem fantastical, it is inevitable. Many clinically accurate technologies required to cheaply digitize our physiological state already exist and the biggest missing pieces will be available within five years. The majority of these technologies are already increasing in resolution and decreasing in cost at a rapid pace, and the next generation of technologies will be at least an order of magnitude better on a cost-performance basis. Simultaneously, we continue to have exponential increases in computational power per unit cost. This nexus will enable the opportunity to reinvent healthcare so that it gets better, cheaper, and more accessible over time.
Contrary to what many predict, we believe these advances will help restore the relationship between the patient and primary care physician to be the pillar of our healthcare system rather than further diminish it. We believe this relationship is important because we are emotional beings, and no matter how quantitative healthcare becomes, there will always be a human element required at the point of care — because no one wants to be told they are ill by an algorithm that will never get sick and will never die.