The Limits on Healthcare Learning From the Business World

The Limits on Healthcare Learning From the Business World

A favorite criticism of EHRs is that they are glorified billing platforms, rather than clinical tools.

Despite being sold to — and subsequently, by — the federal government as being healthcare’s ticket into the modern age, and to leveraging big data, EHRs haven’t so far facilitated the kinds of analytics initially touted as the new standard in medicine.

Part of the shortfall in realizing the potential of EHRs may be fairly ascribed to overselling; new technology and new applications of existing technology tantalize imaginations. In an industry with as many challenges and problems to solve as healthcare, it is understandable that advocates got carried away with silver bullet thinking about EHRs and let development and implementation — not to mention security — fall behind.

But part of the problem is also an overextension of the analogy that what works for the business world, ought to also work for healthcare. Even outside of EHRs and questions of technology, the assertion that business leaders, models, systems, and tools have pedagogical value for healthcare leaders, and practical value for clinics, has become so popular that for many it sounds indistinguishable from conventional wisdom

That may have its merits, but the reality (all too familiar for those actually working in healthcare) is that business lessons very often don’t apply to the health sector.

 

Are EHRs Failing to Deliver Analytics?

When big data doesn’t work, it isn’t necessarily a failure of information, but a user error; a failure to properly organize information, or to ask the right questions of that information. In other words, big data doesn’t just happen by virtue of keeping digital records or even hiring data scientists to get things in order. There has to be a compelling use case, a specific goal associated with the data to turn raw information into something actionable. This is where business and healthcare diverge most dramatically.

The business use case for big data is, first and foremost, about competition. A forensic look at marketing initiatives, supply chains, tax planning, even compensation, all serves to make businesses more lean, more efficient, more profitable, and ultimately, more resilient in the face of stiff competition. McDonald’s managed it before “big data” was a buzzword, by simplifying its menu and streamlining its kitchen. Today, it is synonymous with “fast food” not because it is holistically the best, but because it led the pack in turning analysis into a competitive advantage. And it continues to use technology, analytics, and big data to further hone everything from sandwich assembly to locating new franchises around the world.

It should hardly require saying so, but healthcare has no McDonald’s model to follow.

Healthcare data analytics — carried on the backs of EHRs — are not necessarily intended to support competitive improvements or advantages. By and large, major clinics and hospitals have a virtual monopoly, if not geographically, then often in terms of insurance networks, or both. So the idea that competition drives innovation, optimization, or introspection is a non-starter.

 

The Profit Motive

Businesses are looking for improvement opportunities not just to aid the bottom line, but to boost profitability. The majority of clinics in the United States are, at least on paper, nonprofits (or government-operated). So in these hospitals, that “bottom line” under scrutiny by CEOs and data scientists often has more to do with volume, sustainability of operations, and especially coordinating with insurers in order to remain solvent.

Big data in business enables corporations to minimize the costs of their own operations, and to pass on some measure of savings to customers. That boosts profitability not just by making the cost of business lower, but by incenting consumers to buy more, or at least, to elect to buy from the optimized company. Everyone along the supply chain is looking for the best, for the least.

Healthcare is never so straightforward. Prices are hopelessly opaque in healthcare, and the relationship between the many stakeholders along the supply chain — from universities to providers, clinics to insurers, consumers to pharmaceutical companies — is all but impossible to optimize because there are so many different motives, inputs, and contradictions involved. People are looking for the best, but seldom have any way to judge quality, or have no access to competitive alternatives, or to balance quality with cost, or to hold anyone along the way accountable for quality or, for that matter, setting prices.

What this all amounts to is a limitation on the ability of healthcare organizations to make use of their data in the same way their business sector counterparts have been doing with any hope for success, insight, or actionable conclusions. That the finances of free market corporations and health systems are different is itself not an especially novel observation, but the fundamental difference of motivation extends further than price-setting and value-shopping.

 

Optimizing for Engagement

Although broadly similar, and often looked for in the data, the effect of “engagement” in a normal business setting is critically different from the sort of engagement providers and health systems are trying to achieve with patients.

Engagement in marketing is a matter of driving conversions; the more consumers hang out on your site or are exposed to your brand, the more likely they are to convert to buyers. This kind of engagement takes shape as funnel: get the widest possible audience to begin engaging, then optimize every node, webpage, or conversation to drive them all toward one destination: purchase.

In the business world, you see this driven by big data in the form of things like A/B testing to maximize webpage performance. Optimizing ecommerce or brand websites, targeting marketing messages, streamlining design for user experience and ease of navigation — it all funnels down to that old bottom line. When a given consumer’s experience seamlessly and pleasantly flows from landing on a website to buying a product or service, the engagement effort has worked. Engagement for business, in other words, is discrete.

Engagement in healthcare has a very different connotation, with extremely different end goals: engagement is about adherence, first and foremost. Getting patients engaged with their care is a function not of encouraging brand loyalty or making a sale, but of trying to optimize the value of the care they have already received. In other words, engagement after the sale is more important than leading up to the sale, because what happens after a visit to the hospital can be more critical to patient health than the limited encounter they have with providers.

In medicine, engagement is continuous, and more a matter of perpetual relationship-building, of exchanging feedback, than of driving everyone to one universal outcome. Individual patient health goals are unique; sales goals are easily generalized. A/B testing a patient portal may help improve general user experience, but the substance of a patient’s chart, or conversation with a provider, can’t be optimized the same way a product page can. While a specific retailer or brand can optimize experiences to their specific consumer demographic, healthcare organizations have the impossible challenge of optimizing all patient engagement pathways to anyone and everyone who needs medical attention.

 

Redesigning Health Data

The other example of A/B testing in healthcare, of course, is the control study for medications, new procedures, or determining best practices. This is where the real value, the maximum return on investment, from adopting EHRs should be sought. The big data EHRs deliver can only do so much to highlight wasted revenue, inefficiencies, or optimized patient experiences in the sense that the business world so often makes use of. But outside of the profit motive, or of engaging consumers to make a sale, big data in healthcare can begin to reveal population trends, problems with current standards, pathways of disease, and where health resources are needed most.

The best use case for data in healthcare is not a matter of competition as it is in the business world. It is a matter of learning, of monitoring populations not to take advantage of trends, but to anticipate and prevent disaster or outbreaks. The best use of the data is not presenting it to leadership or business-minded members of the C-suite, but making it accessible to the academic community, to researchers and scientists who can turn it into a competitive advantage against death and disease, rather than the marketplace.

Achieving this takes standardization, interoperability, and some amount of relief for providers feeling taxed by the need to play data scientist and doctor at the same time. All easier said than done; what’s worse, interoperability among legitimate and authorized users is lagging behind security failures and vulnerabilities across the healthcare industry. But progress can start with recognizing that EHRs don’t need more help from the business world to fulfill their promise. EHRs, like scalpels or stethoscopes, don’t belong in the boardroom, and their use and design is best left not to administrative types, but to medical professionals.

Leave a Reply

Your email address will not be published. Required fields are marked *