By Dr. Elia Stupka
Boehringer Ingelheim Pharma GmbH & Co. KG Research
Health and the “Atoms vs Bits” paradigm
Negroponte coined the famous “atoms vs bits” paradigm in his book “Being Digital”. Indeed, as predicted, large amounts of information and knowledge have shifted from atoms to bits: think of books, music, films, encyclopedias. As both small and large companies started digitizing life there seemed to be no limit to what could shift from atoms to bits, be it an audio file, a conversation, or the pace of a jog.
And then came health. Time and again somebody assumed that the key to crack health was “just” a matter of “standardizing and digitizing health records” (remember Google Health?). And that the main challenges were the heterogeneity of the data, standards and formats, but once that would be solved, it would lead to discoveries, therapies, solutions, in other words, real impact on real people’s lives.
In parallel came genomes. As long as we sequence hundreds, thousands, millions of genomes, share the information across the wider global scientific community and enable discoveries, we will crack a significant chunk of the health challenges facing humanity.
Reality Check 2016
Let’s fast forward to 2016. We have progressed steadily in improving the sharing and exchange and standardization of health records. We have created possibly the largest scientific communities ever witnessed in history around genome sequencing and interpretation (just think of some of the great work done by ExAC, Genomics England, the global alliance for genomics and health!). And yes, genomes have had tremendous, rapid impact on rare disease diagnosis and understanding and tailoring cancer therapies that we should all be tremendously proud of, and which we can only hope will keep accelerating at the current pace.
But we all have one, two, way too many friends or relatives who are affected by some of the most common diseases affecting humanity, be it those affecting wealthy western countries, or the public health challenges affecting developing countries. In some they have no treatments options available, in some others even the most recent and innovative therapies are likely pills, containing chemical compounds (or more recently antibodies) targeting mostly the same 500-1,000 genes encoding proteins, which were the result of long, expensive and time-consuming “good old” research projects encompassing a variety of in vitro testing, pre-clinical animal testing, toxicology testing, years and intense efforts in identifying and optimizing a compound, and even more expensive and intestive clinical development, through the three, required, and crucial clinical trial phases. And how many of those will have truly, concretely benefited from either vast electronic health record collections or genomes? A small fraction, and mostly as additional “supporting evidence”, with the exception of a few interesting targets that emerged from GWAS studies. But there are countless drug targets which came from reading a good paper, or in depth research and analysis of diseased tissue affected samples (think gene expression, proteomics, epigenetics, etc).
The HEALTH DISRUPTION mantra: how fare are we?
The theory goes that any large, old, industry that has not been disrupted by a clever tech startup and/or by a tech giant such as Google or Apple, will be disrupted very soon. It has happened to taxis and hotels, it is happening to insurance, mortgages (“fintech”), and it is now even feared in law firms and other human consulting services, as lawyers are feared to be substituted by artificial intelligence.
My LinkedIn feed is streaming every hour by hypothetical scenarios where the same will happen to health: a “magic” combination of big data, artificial intelligence, wearables, clever algorithms, massive pre-competitive data exchange and sharing will yield the upcoming, disruptive business model for health, where the pace of therapy development will be accelerated 10-fold (down to a few years from the current 10-15 years), the cost will decrease 10-100 fold (from billions to peanuts), and the real world, perceived, patient outcomes will significantly increase (from marginal, statistically significant, carefully designed trial results, to a large fraction of affected patients screaming “it works, I feel good and functional again!”).
It has not happened yet. In fact, if anything, some of the greatest disruptions that are coming to health are not necessarily in the area of electronic health records, but in fields you can touch and feel, and use, e.g. the potential that gene editing (CRISPR/CAS technology) brings to the table, to customize, test, develop very rapidly specific modifications, vectors, tools that might alter the biology of disease. Or the potential that wearables bring to measure disease relevant aspects of a person’s life continuously. In other words, the potential to bridge atoms and bits!
Ups…. health is complex!
And here comes a topic that isn’t particularly trendy to talk about when we talk about innovating in health: health is intrinsically extremely complex, from at least two angles:
- It bridges a vast number of disciplines, just to name a few: medicine (to understand disease), biology (to understand the underlying mechanisms), chemistry (to develop tools that could modify the disease), computer science (to bring all the data together and analyze it), statistics (to distinguish signal from noise), engineering (to develop devices), physics (to develop tools such as imaging reagents), law and ethics (to do what is agreed to be “good”, and regulate it how it is done, since we deal with human beings!), economists (to understand the financial value of a therapy), etc. etc. the list goes on.
- It connects different components of society, who share a passion for improving human health, but act in accordance to completely distinct functions, capacities, roles: patientsincreasingly informed and aware of the work surrounding their condition, increasingly frustrated at the slow pace of change, and increasingly interested in playing an active role in the innovation process, scientists, who want to learn, discover, produce and access as much data as possible to push the boundaries of what can be done, doctors and nurses, who see the patients on a regular basis, and attempt to make the best possible choices for them based on existing, concrete, validated and safe options available, payers who need to establish clear, quantifiable ways to decide what should be reimbursed and how much based on the health impact it has (e.g. QUALY, quality of life years gained), regulators, who must ensure a good compromise between pace of innovation, discovery, research and the safety and ethical compliance and real utility of this progress. And all of them acting within a fragmented, globally diverse universe of national and international government organizazions, companies, charities, universities…
It then becomes a little more obvious, perhaps, why “health tech” would not produce the same, rapid, disruption that is seen in other industries. Health cannot be improved simply by moving more data from atoms to bits, but by developing platforms, business models and new approaches to connect all the dots described above (disciplines and involved components of society). Thus connecting atoms, bits… and souls!
The unifying factor: patients
Search the text of most presentations pertaining to health and you are bound to find some reference to heterogeneity, fragmentation, diversity, be it relating to the way a disease is classified, or the biological causes of a disease or indeed how it is reimbursed in different countries.
So what is the one and only unifying agent in this crazy, complicated field? The patient. A patient is likely to be represented in a 100 different ways in different systems. They are likely to be “known” as a biopsy image by a pathologist, a prescription history by a pharmacy, a health insurance account by payers, a clinical record by a doctor, trial protocol number by a pharmaceutical company, And yet it is still the same patient, suffering and hoping that a cure is found for her/his disease. And, as Jamie Heywood from Patients Like Me often reminds me, a positive outcome is a unique, very personal measure. A pianist needs to use his hands, a computer geek needs to be able to program, a barman needs to stand many hours, so a positive outcome when trying a new therapeutic approach means different things to different people.
Probably one of the few people who really knows her/him is a dedicated, hardworking, nurse, who is among the few to know that she/he has ups and downs in reacting to treatment, that they have been biopsied 3 times by 3 different pathologists, that their DNA was sent for genome sequencing, that their data was used in a research study, and that yesterday was their birthday, and they fear it is the last. And they help in the oldest, way known to humankind, with an understanding smile, a warm gesture, the most powerful and beneficial placebo treatment. A soul to soul treatment, no bits involved, some atoms hopefully affected.
Now let us talk about “disruption”
Where does this leave us in terms of innovating, disrupting, and improving discovery and innovation resulting in real improved outcomes for patients? Does it mean it is an insurmountable challenge? Not at all. It just means we should not simplify disruption and innovation in health to bits. And even bits and atoms (e.g. big data plus a great biobank) will be sufficient. Without the souls, it will not take off. Who are the souls that need to be involved? All of us who care, but above all, the patients.
So how do we get there? We need to alter radically how we operate in the health space. It won’t work anymore for scientists to discuss their latest greatest research results without communicating with patients. It won’t work for individual companies to try to build up large datasets around specific diseases in isolation. It won’t work for doctors to exchange unusual case reports of “unusual patients” without investigating them with scientists. And it definitely won’t work to ignore patients, their families, and their caregivers and the wealth of knowledge they have about their condition, and their very personal, specific view of how they are living their condition.
Indeed one way to look at this is a “circle of care” around the patient. To ensure that all relevant knowledge, formal and informal, biological data and samples, clinical history, wearable data, caregiver comments, prescription medicines are provided and integrated and where a patient would be empowered to consent dynamically (see work on dynamic consent by Deborah Mascalzoni and other renowned ethicists) on how this wealth of information should be used or not used. The world is not divided anymore into “big bad corporations” and “good, loving, charities”. Good and evil can occur in any organization, and the greatest protection from evil is transparency, engagement, human relationships.
Imagine a patient who from early symptoms, through diagnosis, hospitalizations, prescriptions, and progression of disease was able at a tap of a finger to see all their information in their own “health dropbox”. Imagine that they would get a request from a scientist, nurse, doctor, biotech company, social scientist to mine their data, and to be updated on how their data was used. And imagine he was supported by knowledgeable staff, highly trained nurses who would know their history, could assist the patient in providing precious samples for further research, and support the understanding of more complex scientific or medical concepts, “disease coaches” who could ensure that the patient lives in an ethically appropriate balance of contributing to research, benefiting from research, and having the best possible treatment known at any particular time. In other words imagine a world where the universe of “health” (intended as a combination of data, samples, and interested parties) revolves around the very person we all want to help: the patient.
Then we might really start disrupting this crazy, amazing thing we call health, by bridging bits and atoms to touch souls.