The Xconomy San Diego Forum on the Human Impact of Innovation offered a window into how information technology and science, marching hand-in-hand, will shape healthcare. “IBM’s Watson is not the only breakthrough game in healthcare,” was the best summary of the afternoon.
Today’s unpredictable markets demand that as you maximize performance from your current business model, you must in parallel be designing and developing new business models to disrupt the current cash generators. Otherwise, start-ups will turn your company into a dinosaur. In doing this leadership work, consider four key trends highlighted in the conference. Each one alone portends major changes for a number of players in the industry and the combined effect of which will surely reinvent the entire sector.
1) Advances in “micro” medicine
Healthcare technology is increasingly leading to a better understanding of what’s happening at a genetic level. For example, researcher Larry Smarr, Director of the California Institute of Telecommunications and Information Technology, and others are investigating the relationships between the genetic profile of microbes in our gut and autoimmune diseases like Crohn’s and IBS. The insights suggest pathways to better treatments and cures.
Sequencing of the genome by companies like Illumina (a sponsor of the symposium) has moved from research like Smarr’s to the clinic level. Illumina’s VP/GM for Reproductive and Genetic Health, Jeff Hawkins, discussed how sequencing is leading to better diagnostic tests, such as a more accurate and less risky replacement for amniocentesis. Another new sequencing application is the rapid diagnosis of children’s rare diseases, cutting the cost of healthcare and advancing family wellbeing.
2) Advances in “macro” medicine
At the other end of the spectrum are advancements that allow investigations across groups or populations: large data sets, affordable storage, much faster data processing hardware, better availability of mobile sensors (e.g., EKG, heart rate), and machine learning technology. Collectively, these create previously unavailable insights into population health drivers.
Machine learning is a set of techniques that, by recognizing trends and categorizing data using neural network computing, forms predictions of health outputs from multiple inputs. Application of this technology holds the promise of identifying the emergence of a disease before symptoms appear based on comparing sensor streams from multiple individuals. Smarr argues humans are quite similar, making machine learning highly applicable to prediction, diagnosis, monitoring, and health coaching.
Cardiologist Steven Steinhubl of the Scripps Translational Science Institute reported on the largest National Institutes of Health multi-year study – one that will collect the whole genome, microbiome data, activity data, and digital health measures of a large population over time. (It’s an exponential expansion of the Framingham Study, initiated in 1948, to identify heart disease risk factors.) Machine learning will unearth rich insights.
3) Increasing Use of Artificial Intelligence
When known relationships are programmed into a computer to guide mechanical action, we call it automation. Think of the mechanical robot on an assembly line carrying out routine operations in place of human hands.
When you add methods, algorithms, and technologies like machine learning that make software not pre-programmed, but instead self-learning, you have artificial intelligence (AI). AI is also called smart software, so named as it “may seem human-like to an outside observer,” according to Lynne Parker, director of the Division of Information and Intelligence Systems for the NSF. We’ve seen a lot of static machine learning (e.g., the kind that teaches a computer to recognize human faces). The future brings dynamic machine learning, one where the software continues learning as the application is running.
Patryk Laurent, head of AI and Engineering for LeEco US, a company that connects content across all devices for seamless user experiences, shared his views of where we will see consumer-focused AI in healthcare. Childproofing a home, automated physical therapy assistants, and making a home easy for elderly to live in are applications that will incorporate local processing and sensors to create actions sensitive to the user’s behavior and goals.
Dan Goldin, the longest-serving NASA leader and founder of KnuEdge is creating a private cloud service for scalable machine learning and AI. It’s built on new kinds of hardware and uses structured and unstructured data. He argues AI will be a huge business, transforming all industries. (He cited a McKinsey & Company estimate of $25T by 2025.) Devices that today capture images or data today will capture patterns of life that predict health issues and security risks as they arise.
Intel’s Urs Koster shared developments in Intel’s hardware that will accelerate the adoption of AI. His case studies – automated driving, identifying fault lines in rocks for shale oil exploration, and fraud prediction in financial markets – while not healthcare examples, conveyed the power of new hardware technologies. The hardware will accelerate the use of surgical automation and AI for imaging.
4) Precision Medicine
The three preceding trends combine to enable precision medicine – focusing the diagnosis and treatment to the individual level versus “demographic group” level. Instead of going through multiple levels of treatment before you get the right high blood pressure or cancer treatment, you’ll start out with the right mix given your unique make-up.
A wellness company that is already acting on precision medicine, Arivale, identified pre-diabetic patterns in healthy individual clients, coached them for behavior changes, and eliminated their risk of developing diabetes, according to Arivale’s Clayton Lewis, CEO and Founder. Drug companies selling expensive therapies are already using precision medicine to identify the right candidates for their drugs.
Implications for business models
When trends are in opposition, they create sudden breakpoints in a market. Rising healthcare costs coupled with declining affordability resulted in a sudden change from cost-plus provider reimbursement to prospective payment in the late 1980s. We face similar challenges today. Value-based payment approaches and employers’ directing employees to Centers of Excellence for costly surgical procedures are two recent examples of breakpoint changes. Another breakpoint, given the trends above, will be a rapid shift of healthcare spending from sickness care (where our nation currently spends most of its healthcare dollars) to better management and prevention of disease.
My takeaway from the conference is that if healthcare sector companies are not concurrently considering how to adopt new technologies and link more closely to ongoing discoveries, their 3-year strategies will fall short. These technologies are vital to lowering costs and improving outcomes. And, when considering the longer-term boundaries of their business, they must discuss prediction and prevention to remain relevant over the longer term.