Editor's Pick (1 - 4 of 8)
Leveraging Technologies To Better Position The Business
Reducing And Recovering Your IT Spend
Disrupting the CIO Comfort Zone to Innovate and Transform How...
IT Strategy in Healthcare
Data Science Helps UCHealth Improve Infusion Center and Hospital...
Steve Hess, CIO, UCHealth
The Coming Era Of High Performance Medicine
By Jason Burke, System VP, Enterprise Analytics & Data Sciences, Unc Health Care
Part of that opportunity can be framed as an over-reliance on business intelligence (BI). Whereas retrospective reporting capabilities like BI are very useful, they provide few insights into what is likely to happen with patients right now, what will be happening tomorrow, what are the contributing factors, and where are the opportunities to influence outcomes. If healthcare were a performance race car, its dashboard gauges today would usually display car data on at least a 24-hour delay–not very useful when the car is on the track moving 180 miles per hour.
Over the past eight years, the federal focus on electronic medical records adoption has opened huge doors to data-driven transformation. Historically, less than 20 percent of physicians used electronic health records; that number is now over 90 percent. Given these large and growing repositories of health-related data, if BI alone doesn’t deliver high performance, how can health organizations more effectively leverage this data to transform every patient’s outcomes and costs?
One starting point may be to consider the technological capabilities supporting other chaotic, high performance environments. If healthcare is a race against diseases, what would a health care race car look like? If the changing on-the-ground conditions in healthcare can be compared to evolving battlefield conditions that military teams face, is it possible to adopt similar situational awareness and responsiveness strategies? How can health IT provide the same sorts of intelligent, actionable insights that weather forecasters enjoy daily?
In considering these and other questions, capabilities for enterprise architecture begin to emerge.
Over the past eight years, the federal focus on electronic medical records adoption has opened huge doors to data-driven transformation
Weather sensor arrays, battlefield communications, car engine sensors, and other technologies ensure critical data is available for decision-making. Health care has started on this journey with the adoption of electronic records systems, though many of those systems were designed more as electronic filing cabinets. Incorporating other data—genomic data, business process measures, medical devices, consumer wearables, external data aggregators, and many others—provides opportunities to develop more sophisticated models of medical decision support. But it requires an enterprise architecture that can handle “big data” with greater agility than the systems of the past. And it requires organizations embrace disciplines such as data governance to improve the quality and semantic utility of their data assets.
Real time data
A seismograph that reports earthquake readings two weeks after the quake occurs does not lend itself to decision making and emergency responsiveness. The value of information is highly time sensitive. And whereas many health care organizations are well positioned to handle the batch loading of periodic data files, capabilities associated with real time interoperability across heterogeneous systems and environments are far less mature. But newer standards like FHIR are demonstrating the promise of greater interoperability both within and across IT architectures.
Recently, health care providers and insurers have been designing and implementing new models of care management and reimbursement that focus on empowering stronger health teams–physicians, specialists, care managers, nurses, and others–to better support patients and their care needs. In order to manage each patient’s health risks, team members need greater visibility into shared business processes. For example, if a physician refers a patient to a specialist for a consult, did that appointment get scheduled? Did the patient get there? Were any next steps identified? Were they followed? Just as racing teams need to know what car tunings produced high performing results, health care teams need to know what activities help patients along their journey to better health.
Predictive modeling and simulation— techniques routinely used in weather forecasting, for example—will become foundational for improving the quality and costs of care. Part of the opportunity is in gaining a better understanding of the factors influencing health outcomes; for example, why do some patients tend to come back to the emergency department within 30 days of discharge from the hospital? But part of the opportunity is also in developing more personalized views of improvements: what would best help this particular patient avoid complications that would result in a new hospital visit? Advanced analytics can change how we respond to medical problems, but it can also help to prevent those problems from ever occurring in the first place.
Many of these capabilities represent daunting challenges, but they are far from “pie in the sky”. Health industry conferences and publications are regularly featuring early examples of these changes in motion—detecting diseases by analyzing free text data, preventing hospital readmissions by modeling patient risks, uncovering unnecessary variations in cost-sensitive care processes, identifying patient subgroups that respond differently to treatments, developing “precision medicine” approaches to diseases, and many more. For technology-savvy leaders–CIOs, CTOs, and increasingly Chief Analytics Officers–there has never been a more exciting time to work in healthcare.
Data Science Helps UCHealth Improve Infusion Center and Hospital Efficiency
Steve Hess, CIO, UCHealth