2020 Trends in Health Analytics
10 January, 2020
Originally published in Analytics
Editor’s note: Predicting what might happen in the dynamic world of healthcare is risky business, particularly in a world where technological advances and efficient, cost-effective, data-driven care often bump up against institutional resistance, politics and patient privacy concerns.
Nevertheless, Jeff Terry and Ian Worden, C-level executives at their respective health organizations, offer the following health analytics trends to pay attention to in 2020.
1. Don’t Sleep on Retrospective Analytics!
The last decade saw new appreciation for actionable information. Everyone in healthcare administration has seen dozens of “analytics maturity curves,” which show a progression from retrospective through predictive, prescriptive and self-learning analytics (or some version of this). And while there is some truth to these frameworks, they also created the misconception that some analytic types are “better” than others.
Retrospective analytics have been so consistently presented as the bottom rung on these ladders that they are now often treated as unworthy, uncool, whatever. We think this misses the point.
The measure of an analytic is usefulness, not technical sophistication. Predictive and prescriptive analytics are useful for many things. They’re also interesting because they enable us to impact problems that we couldn’t with traditional analytics. But make no mistake, retrospective analytics are the foundation of understanding any problem. They are useful for an infinite and ever-growing set of problems.
We think a new equilibrium is emerging as leaders gain experience. These leaders value the right tool for the job and discourage heady talk of which analytic type is best. They need all analytic types at the right time for the right problem. These days that often means breathing a bit of energy back into good old business intelligence and statistics. Don’t sleep on retrospective analytics!
2. Coming in 2020: New FDA Guidance
Keep an eye out for new U.S. Federal Drug Administration (FDA) guidance regarding clinical decision support software later in 2020. It is focused on clarifying whether certain types of functionality in software are classified as a medical device. In September 2019, the FDA issued for comment a draft guidance (https://www.fda.gov/media/109618/download). The comment period closes in December. A final guidance is expected sometime in 2020.
3. Three Cheers for Integrated AI
Artificial intelligence (AI) is all the rage, but success stories at scale in healthcare are hard to find. This is probably because AI by its nature isn’t a solution, and it’s probably not even a tool. Instead, AI is something that makes other tools better. For example, speech recognition makes data entry faster. Machine learning makes algorithms better. Natural language processing creates more value from free text already entered somewhere.
It is for this reason that we find the trend by several major companies toward “integrated AI” to be positive. Integrated AI suggests that product managers and engineers at those companies are starting from the simple but powerful perspective that the AI must integrate into an existing workflow, process or tool. We hear this new clarity of focus from electronic medical records (EMR) and device companies. If it continues it could propel AI in healthcare to leap the chasm from interesting to important. Here’s hoping that it does. Three cheers for integrated AI!
4. The Necessity of Apps Outside the EMR
The last 20 years saw the rise of the holistic EMR. As far as we can tell, every “electronic medical record” in developed countries now refers to not only the patient record, but also a collection of workflow systems like lab, pharmacy, radiology, orders, bed management, transfers, patient transport, etc.
EMRs have largely consumed what was once an ecosystem of workflow systems – those workflows are now mostly just another module in the “EMR.” And while there is still much to be done to optimize EMRs (perhaps most notably to simplify them in order to liberate clinicians from “screen time”), they have generally been a positive force in healthcare.
The scale, sophistication and reliability of modern EMRs may in fact be among the most unsung technological success stories ever. EMRs are amazing.
We now also know that EMRs are not enough. Even the best and biggest EMRs have limited utility. Feedback from providers in at least five countries is that their EMR is most useful for managing each patient, but not very useful for managing all the patients. For example, EMRs struggle to spot needles in haystacks, facilitate planning meetings or prioritize resources. The reasons for this have to do with EMR architecture and the complexity of the underlying data.
It is for these and other reasons that “apps outside the EMR” have emerged; caregivers need key information quickly, and current tools aren’t enough. These new apps are providing at-a-glance decision support in areas like discharge planning, clinical surveillance, resource prioritization, etc.
In most cases the apps are Real Time and draw data from various EMR and other workflow systems. We see real impact and major potential with this concept and expect it to accelerate in 2020. Feel free to contact the authors for a list of the vendors we see emerging. Expect these new tools to be a permanent feature of the healthcare landscape.
Ian G. Worden is the chief operating officer of CHI Franciscan, a nonprofit healthcare network that includes 11 hospitals and other healthcare facilities in the Seattle/Tacoma area, where he is responsible for directing the operations of acute care services along with coordinating, integrating and growing these services in communities while ensuring the delivery of cost-effective care. Prior to joining CHI Franciscan, Worden led St. Vincent Health System, a 23-hospital system in Indiana, as its executive vice president/chief operations officer. As the COO, he drove the system from a financially struggling operation to the most profitable regional system in Ascension Health. He has master’s degrees in neuroscience and anatomy, business and finance, and health services administration.
Jeff Terry built and is the global CEO of the Clinical Command Centers group for GE Healthcare Partners. His innovative team delivers transformational change to help top-performing healthcare organizations achieve even better outcomes and efficiency by re-engineering clinical operations and imagining, designing, building, and activating healthcare command centers. Since joining GE in 2001, he has led a number of strategic initiatives and solutions in Six Sigma, clinical asset management, clinical operations, healthcare strategy and patient safety. Prior to GE, Terry served as a combat engineer lieutenant and captain in the U.S. Army’s 1st Armored Division. He is a Fellow of the American College of Healthcare Executives (FACHE) and holds an MBA.