by Christopher Jankoski
Separating Digital Health Fact From Fiction
Posted by Chris Jankoski from Ogilvy Health North America on March 10, 2019
It’s hard not to get caught-up in this whirlwind of hype around the next-generation innovation on-display at SXSW. In the healthcare space, we are no stranger to this kind of fanfare. From the latest wearable promising to track your every step along the patient journey to the widespread urge to look at artificial intelligence as a silver bullet to hurdle barriers to quality care; HYPE IS EVERYWHERE.
A panel from yesterday’s agenda, Healthcare’s Digital Disruptors: Hope vs. Hype, looked to take specifics examples of today’s digital health landscape to explore how they are being used today and how we can optimize innovations to sort hope from hype.
Predictive analytics, in the healthcare space, is the use of current and historical health data to track and forecast health behaviors and outcomes. They hope lies in that by pulling (and making sense of) all the data points, we can drive efficiencies for healthcare professionals, intervene in people’s lives at critical points along the patient journey, and ultimately provide better outcomes with fewer resources than ever before.
The panelists were collectively hopeful for this technology to make real impacts for the patients of today and beyond. Because we are collecting so much health data from our wearables, electronic health records, and other devices which are embedded in our daily lives, the promise to use that information for clinical decision support is huge. According to the panelists, some ways that predictive analytics is making real impacts today are in reducing mortality rates, lowering the risk of readmissions, and illuminating the root causes of no-show appointments. All of which would improve many areas of our healthcare systems if driven to become more efficient. Although the health data is not perfect and there is still a long way to go as far as improving the algorithms to showcase relevant insights — the panelists agreed that this technology is here to stay and support healthcare experiences of today and tomorrow.
Beyond the in-office data collected, there is huge promise with incorporating other data points of our lives to paint a holistic representation of our health and wellbeing. With things outside of medical care like socioeconomic, environmental, and daily behaviors making up 80% of what contributes to healthy outcomes,* often called social determinants of health. The promise lies in pulling together those data points outside of the healthcare experience to level-up the use of predicative analytics and drastically improve the quality of care in a proactive manner.
Machine learning uses statistical models and algorithms to analyze the health data being collected to extract insights. A process which improves over time through learned patterns of information and behavior. The hope lies then in machine learning as a critical tool in making sense of the billions of health data points being collected each day.
The panelists were cautiously optimistic around this technology to make tangible benefits in healthcare experiences. Some areas which are already seeing machine learning’s promise are the medical imaging and diagnostic landscapes. Machine learning can be a key element to disrupting how and where diagnostic testing can happen. An example from a panelist was in pregnancy testing. All pregnancy testing used to require a trip to the doctor’s office, but now the at-home testing mechanisms have made it a more seamless and convenient process. Applications of machine learning to diagnostic tools could mobilize technology and bring some of these other tests into the home moving from a test in a medical environment to a test in a more consumer-centered environment.
Although the ability to effectively make sense of all the data around us is a critical element to the next generation of digital health, the panelists did proceed with caution with the idea that we have a long way to go in getting and processing quality health data points. One speaker, Irna Nash, explained, “The value of this is only as good as the data we feed the machine.” So, we aren’t there yet, but machine learning does hold promise to make sense of our data-driven world in the future.
Precision medicine is “an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person.”* The hope with this is that by using this information, we can provide truly personalized drugs and treatment methods that will be most effective to the patient.
Here is where the panelists seemed most conservative in their views. While the concept of providing more personalized care was a shared hope, the hype around using genomic data in today’s environment was met with resistance. There are some areas where precision medicine is making strides like limited oncology treatments, but this was the exception to the promise because the understanding of genomics is not at the point where it can make actionable insights. The speakers offered an alternative school-of-thought to make it a bit more tangible in that we can shift from precision medicine to precision health. The shift that the panelists discussed was to think of patients more holistically which can help intervene at the points that will make the biggest impact in their lives. This thinking honed-in on the latter part of the above definition which focuses on environment and lifestyle factors of health. The panelists offered examples like thinking about a patient’s availability to housing, access to healthy foods, ability to purchase and store medications, and more. By understanding these outside factors, we can tailor other areas of behavior change to ensure treatment methods make sense in the daily lives of patients. While the panelists ultimately put precision medicine into the “hype” category, the thinking around personalized healthcare engagements via factors outside of the doctor’s office offers both hope and promise.