What healthcare organizations can learn from other industries

December 9, 2014

Joining forces with ZirMed has given me even more opportunities than before to answer the question of why the future of healthcare is in predictive analytics. Sometimes I can answer in just five words.

  • Because executives will demand it.
  • Because their boards will demand it.
  • Because patients—as consumers—will demand it.
  • And because the new financial realities of healthcare will demand it.

OK, some take a few more than five. No matter though—the truths behind them are the same.

  • Healthcare is one of the largest sectors of the American economy.
  • Unlike other major sectors, it hasn’t yet been transformed by predictive analytics.
  • Healthcare organizations, in turn, are the only fixtures of the American economic landscape that haven’t yet reaped the benefits of exception-based workflows, true data mining, and cloud-based analytics that enables economies of scale through massive gains in efficiency.

Big surprise.

Or not. After all, healthcare isn’t like other parts of the American economy. The healthcare landscape itself can be reshaped by legislative or executive action, at pretty much any time and without much notice. Healthcare isn’t just a business—or, perhaps more accurately, it’s much more than just business.

It’s also still in the process of transforming from paper-based models and standalone systems to electronic records and cloud-based software. That gap—from where healthcare is today to where it will be in five years—is a huge leap. But it’s one that other industries such as banking, shipping and logistics, as well as retailers made more than a decade ago.

When I look at what it will truly take for big healthcare organizations to optimize revenue, manage population health, and capture new efficiencies—especially those currently beyond their grasp—interoperability is the common thread I see in every situation and every challenge. Interoperability is the key to empowering predictive analytics across the myriad systems installed at the enterprise level, and cloud-based solutions are the key to interoperability. Any platform that isn’t fully interoperable should be viewed as a non-starter.

It’s a whole new world. Wait…this world looks familiar.

Predictive analytics hasn’t just transformed how businesses find new efficiencies and hidden pockets of revenue—it’s changed how we shop, how we’re marketed to as consumers, how we travel, and even how governments function.

Alibaba recently went public at a valuation of $25 billion; every unit of its business is a part of an e-commerce platform rooted in an engine that runs on predictive analytics. Apparel retailer Zara increased sales 3%-4% using predictive modeling to optimally estimate per-store demand for fashions, maximizing the availability of clothing sizes for display merchandise. These per-store stocking decisions are made by crunching the numbers only hours before shipments leave the warehouses to more than 1500 stores worldwide.1 Amazon is so confident in its predictive analytics that it’s experimenting with shipping items to customers based on predictive modeling alone—items that those customers haven’t ordered yet. And from the prices you pay on travel sites to the security systems that seek out meaningful anomalies among passengers, itineraries, and baggage, predictive analytics has completely reshaped air and rail travel.

And yes, it’s true that the only way governments can use big data in a substantive way is through predictive analytics. Some people focus on the national security and intelligence applications—but just as real and just as meaningful to daily life are the predictive analytics that underlie modeling of traffic, demand for utilities, and exactly how much of any publicly provided resource or assistance might be needed in the case of natural disaster.

All these things add value—and because of the shift in healthcare toward value-based models, the insights that predictive analytics delivers will increasingly align with value as well. Improved clinical outcomes and optimized clinical quality—value rather than volume.

Clearly it’s a good thing for healthcare organizations to be able to optimize their performance. Today too many provider organizations of all sizes are hitting a technological ceiling as they work to optimize financial workflows and outcomes—in part because healthcare data is growing at an exponential rate, and all but the most sophisticated cloud-based solutions can do little if anything to make managing this big data feasible.

Predictive analytics, by comparison, has the horsepower to ensure that organizations accurately identify the claims that have the greatest financial impact—so that staff can focus their efforts on those claims rather than issues that have comparatively negligible impact on reimbursement and quality reporting. On the patient-collections front, predictive analytics can accurately forecast propensity to pay for self-pay patients as well as those covered by high-deductible health plans (HDHPs)—all based on geographic, demographic, and historical payment data, which time and again prove to be better predictors of propensity to pay than credit score alone. Predictive analytics will show providers which patients are truly at risk for developing pneumonia following a bout of bronchitis, and which patients within a given inpatient population are most likely to be readmitted within 30 days.

Predictive analytics translates to more coordinated care, more targeted and effective care outreach, and evidence-based clinical best practices. That’s the potential I believe in—and it’s why I believe healthcare is the most critical segment of American life where this potential must be realized.

Jeff Kaplan is ZirMed’s Chief Strategy Officer

What healthcare organizations can learn from other industries

Stay Informed

Stay Informed