Wednesday, 25 January 2012 00:00 |

Last Updated on Friday, 27 January 2012 19:52 |

January/February 2012

Analytics & the future of healthcare

Market-driven reform already shifting emphasis from volume to value as industry leaders weigh in on what’s next.

By Peter Horner and Atanu Basu

Healthcare will be a hot topic during the 2012 U.S. presidential campaign as the Patient Protection and Affordable Care Act signed into law by President Barack Obama nearly two years ago is attacked or defended by the respective candidates and their surrogates. However, no matter who wins the White House this year, the U.S. healthcare system will be reformed, and more likely transformed, in the near future, and analytics is certain to play a leading role in that transformation. In fact, reform is already well underway, driven by increased competition within the healthcare industry, the trend toward “accountable care” and the realization that spiraling costs make the current system unsustainable.

According to the Centers for Medicare & Medicaid Services, U.S. national health expenditure totaled $2.5 trillion in 2009, or $8,086 per person, and accounted for 17.6 percent of gross domestic product [1]. The United States spends more money per person per year on healthcare than any other nation in the world [2], yet the World Health Organization ranked the U.S. healthcare system 37th in overall performance [3] (just behind Costa Rica and just ahead of Slovenia) in 2000, the last year the rankings were compiled.

Why is healthcare so expensive in the United States and why doesn’t all that money produce better outcomes across the populace? Some of the more notorious contributors to the problem include misaligned incentives among the various stakeholders, bloated administration costs (someone has to shuffle all that paperwork), fraud and abuse, overtreatment and defensive treatment (from fear of malpractice suits), system failures and a lack of coordinated care, almost all of which are target-rich environments for analytical intervention.

To be sure, the United States offers arguably the best healthcare in the world, but at what price? According to the American Journal of Medicine, medical bills capsized 62 percent of the people who went bankrupt in 2007 [4]. Clearly, healthcare in the United States can benefit from a strong dose of analytics to help improve the performance of a massive, complex, fragmented, hugely expensive system struggling to sustain itself.

Volume vs. Value

The U.S. healthcare system has historically operated on a fee-for-service model. The more patients a doctor sees, the more operations a surgeon performs, the more beds a hospital fills, the more money the care provider in question makes. While patient outcomes and experiences are obviously a concern for all involved, they don’t impact the fee schedule. In short, the fee-for-service model emphasizes volume over value. That is about to change.

The mandate requiring individuals to purchase health insurance has turned into a popular talking point for politicians, but the provision of the Accountable Care Act that has caught the attention of the healthcare industry is the one that imposes financial penalties for providers who don’t meet certain standards of care for Medicare and Medicaid patients. The most prominent yardstick is hospital readmissions – patients who come in with certain ailments and then have to be readmitted to the hospital within 30 days after they are discharged. If a hospital’s number of such readmits exceeds a national standard, the hospital will suffer financially in terms of Medicare and Medicaid reimbursements. That means patient outcomes are now part of the healthcare fee structure, which makes it a whole new ball game.

“If I, as a healthcare provider, am now financially at risk if you as a patient have to be readmitted to my hospital within 30 days, it changes the relationship I have with you,” explains Steve Conti, senior director of clinical innovation and population management at Seton Healthcare Family and head of the analytics committee at the Integrated Care Collaboration (ICC), a nonprofit alliance of healthcare providers in Central Texas. “In a fee-for-service environment, the system is not financially affected by how many times you get admitted. It may call into question the quality of the care you receive, but from a purely financial perspective, it is advantageous to have you readmitted. In a value-based system, it’s just the opposite.”

Conti predicts that within five years, the U.S. healthcare industry will move from a largely fee-for-service, volume-based system to a value-based system. “And the way you get to that new type of structure is through analytics,” he says.

ICC, one of the seven highest-rated Health Information Exchanges (HIEs) in the nation, has embraced and employed analytics since the alliance was founded in 1997. In support of its mission to provide high-quality healthcare in a cost-effective manner, particularly for patients who can least afford it, the ICC operates a regional health information exchange called ICare that contains data on more than a million patients and more than 8 million encounters (provider visits) at 70 locations throughout the Central Texas region.

According to Conti, analytics coupled with the wealth of patient data available in ICare enables ICC provider organizations to identify and reduce duplications in services, thus cutting costs and driving value. ICC also uses analytics and a team of epidemiologists and database analysts to measure and assess everything from readmission rates to clinical ventures. The team uses statistical models to compare how its member providers are managing their diabetic care clinics, for example, to see which ones are doing well and where there’s opportunity for improvement.

“Healthcare is too expensive,” Conti concludes. “When we look at the national expenditure for healthcare it becomes pretty evident that it’s unsustainable. As you back out from that, it causes large health organizations to begin to ask the tough questions. How are we contributing to that cost, and what can we do to become a change leader in the process of making healthcare more affordable, more effective, more efficient and more accessible? And the only way we can understand and improve the process is by having strong analytic capabilities.”

Realigning Incentives

Until fairly recently, the provider side of the healthcare industry had been reluctant to embrace analytics. Humans are naturally resistant to change, and doctors are notoriously wary of ceding the control they’ve historically wielded regarding their patients’ diagnoses and treatment to others, let alone a “mathematical model.” After all, who could possibly know a patient’s medical history and issues better than the patient’s personal doctor?

For-profit hospital organizations had reason to resist employing analytics because “optimizing” their systems could theoretically hurt profits. Imagine a major hospital group that used analytics and electronic health records pre-Affordable Care Act to eliminate overtreatment and unnecessary lab tests and imaging, while simultaneously cutting patient queue times, improving patient outcomes and reducing readmissions. At the end of the fiscal year, everyone would be happy except the company CEO, whose bonus is tied to profits and who has to explain a multi-million dollar drop in revenue to shareholders.

Can for-profit organizations prosper in a value-driven healthcare environment? Of course they can. After all, competing on value and service demands efficiency and effectiveness, assumed private sector strong suits in a competitive marketplace. The only question is, are for-profit healthcare organizations willing and able to adapt and change the way they do business and realign their incentives and pay structure to fit the new industry model? With their survival at stake, the likely answer is yes, and big data and analytics are ready to help.

But isn’t the healthcare industry different from, say, manufacturing? You’re talking about healing human beings, not making widgets, and you can’t just apply systems and global optimization techniques in the traditional, industrial engineering sense to the healthcare industry, can you?

“Health is something that is very difficult to quantify,” says Ryan Leslie, Ph.D., vice president of analytics at Seton Family of Hospitals, a network of 31 healthcare facilities throughout Central Texas. Seton is one of the 32 Pioneer Accountable Care Organizations announced by the Centers for Medicare & Medicaid Services in December 2011.

“As emotional human beings, it’s hard for us to be objective when talking about health,” Leslie continues. “By its nature, healthcare has complexities that you don’t necessarily see in other industries. The outcome of a healthcare work process isn’t a widget. You’re doing something with the intent of helping a patient, but the patient is looking subjectively at the whole experience. You may properly treat an illness, but the patient may not like your bedside manner. You may be incapable of treating an illness, but the patient may still be satisfied with the way that you have delivered care. It’s the peculiar nature of the products and services that come out of healthcare.”

Leslie points out that the economics of healthcare also differentiate it from other industries. For example, healthcare predominantly uses a third-party payer system, and supply-and-demand and market and competitive forces do not necessarily work the same way in healthcare as they do in other industries.

“If you look at the high-tech industry, you see innovation right and left because if you’re not constantly transforming and innovating, you’re out of a job tomorrow. With healthcare it’s a little different because there are so many legal barriers to competition. The decision to consume healthcare products and services and the payment for those products and services are so disconnected that you just don’t have the same economic forces that you do in other industries.”

Leslie’s colleague, Conti, offers a different perspective. “What has led us to where we are in healthcare today – rising and unsustainable costs and the resistance to change – is this philosophy that healthcare is different, that it’s unique,” Conti says. “Obviously, caring for people is very different, but if you look at just the process that is involved in moving individuals in, through and out of a hospital or a clinic, it’s very similar to the hospitality industry – a hotel or a restaurant, for example. … The logistics involved in coordinating all the moving parts required for delivering care to an individual are complex and not dissimilar to air traffic control moving airplanes around airports and around the world.

“At ICC,” Conti continues, “we’re trying to create a system that provides whatever care is needed wherever that care is needed across the entire spectrum of delivery venues and throughout the patient’s lifespan. That’s a fundamentally different way of thinking than what would be required in a hospital-only organization.”

Payers vs. Providers

To some extent, insurers/payers are caught in the middle between patients and providers in the healthcare economic tug-of-war. Providers (and economic factors beyond their control) tend to drive healthcare costs up, while patients and employers obviously want to keep their insurance premiums down. Kaiser Permanente is both a provider and a payer that offers healthcare and health insurance under one big roof, which gives Dr. Yan Chow, director of KP’s Innovation and Advanced Technology Group, a unique 360-degree view of the “pay-for-performance” trend in healthcare. Chow also holds an MBA as well as an MD, further cementing his feet in both the business and medical sides of the industry.

So what’s Chow’s take on the changing healthcare landscape and how will analytics shape KP and the industry going forward?

“Over time, payers will push more and more for the collection of data,” he says. “We need data to know what’s going on. We need shared data to integrate a very fractured healthcare system. We need data to better inform physicians and empower consumers.”

Much of that data will come from electronic health records (EHR). In 2003, Kaiser Permanente became one of the first healthcare organizations to install an enterprise-wide EHR system; today the system holds information on nearly nine million patients and continues to gather information at an astonishing rate.

“We have the largest clinical data repository in the world,” Chow says. “It contains more information than the Library of Congress.” Now the trick is to analyze all of that information and turn it into something useful, Chow says, starting with a couple of applications in its integrated care delivery system.

“Even though we are an integrated system, revenue capture remains a big issue for us and for most healthcare providers,” he says. “We want to make sure when we bill Medicare, for example, we receive the reimbursement that is due based on the data collected during a patient’s visit. Beyond that, from both the payer and the provider side, we’re very concerned about the quality of healthcare, so we look very carefully at data for care quality outcomes and plan to use predictive analytics to find out what kind of treatments or courses of treatment work best for different individuals. It’s a highly personalized analysis and it’s all driven by analytics.

“From the payer’s side, we’re looking very closely at efficiency, and that’s all driven by data,” Chow continues. “We collect data on the time it takes to do things, the time it takes to get resources to get where they are needed, the implications of certain kinds of decisions. We use analytical tools to see how we can become more efficient while providing the same or better quality of care. As a very large organization, we look at the data challenge from many points of view, and while different internal groups may have different agendas, we are all on the same page as far as quality, service and affordability goals are concerned.”

Chow expects other healthcare organizations to follow KP’s lead, and suspects the ones who don’t will be left behind.

“With all the information that’s out there, patients are going to be looking at their healthcare situation very carefully and comparing their options,” he says. “In the future, there will be much more visibility and accountability. If another care provider down the street is getting better results than you for the same cost, you’re going to lose patients.”

Big Pharma Perspective

Andy Palmer, co-founder and former president and CEO of Vertica Systems and co-founder of a half-dozen other successful startups over the past 20 years, made his mark and his money as an entrepreneur in analytics, big data and data management software. Today, in addition to his entrepreneurial activities, he spends a considerable amount of his creative, analytical talent on the life sciences, working with companies such as pharmaceutical giant Novartis (as an executive director) in the area of drug discovery. Palmer has a particular passion for cancer research. For example, he served as an advisor to the Lance Armstrong Foundation on the development of an iPad app for cancer patients.

“Drug discovery and research is a fundamentally interdisciplinary activity that requires scientists on a daily basis to integrate information from across many different organizations and many different scientific disciplines,” Palmer says. “I think of it as the hardest analytical and data integration environment on the planet, but also one that offers the biggest benefit: finding important new medicines for patients with unmet medical needs.”

The bottom line for any pharma company is developing new and more effective drugs and getting them to the marketplace as soon as possible, given all the research and clinical trials that are typically involved. How does analytics help make that happen?

“Pharmaceutical companies make decisions every day on where they are going to invest research dollars,” Palmer says. “It’s a huge distraction and waste of time, energy and effort to start a new clinical trial only to find that there are not enough patients out there in the general population to make the trial successful. It’s important to make sure that we are investing our dollars and our energy as efficiently as possible, and that we’re making the best possible decisions so that we can effectively discover new drugs for patients.

“Every decision should be analytically and data-driven: which trials to run, how those trials are going to be run, where they are going to be run, which patients to recruit for which trials. All of these things should be reviewed in an integrated way. It’s a very challenging analytics and big data problem.”

Chronic Diseases & Telehealth

According to the Centers for Disease Control and Prevention (CDC), chronic disease accounts for about 75 percent of U.S. aggregate healthcare spending [5], which makes chronic disease an attractive target for any entity interested in reducing national healthcare costs. Since unhealthy behavior and non-adherence to prescribed meds often results in complications involving chronic diseases such as diabetes, the concept of telehealth is one way to improve chronic disease management, yet it, too, has run into resistance in certain segments of the medical community.

Unlike traditional medical intervention that basically stops when the doctor writes a prescription (at least until complications commence), telehealth is a means for healthcare providers to continuously reach out to patients through telecommunication technology. Telehealth can be as simple as sending an e-mail reminder to chronic care patients to take their meds, or it can be an intensive monitoring and information exchange system using sophisticated software systems and online questionnaires to ascertain a patient’s health status and to discover early on if the patient is developing a problem – all without the patient ever visiting the doctor’s office.

Adherence to meds and early detection of potential problems translate into fewer complications and fewer emergency room visits, according to a 2011 study published in the policy journal Health Affairs. In the study, Stanford University researchers, reporting on a telehealth demonstration project involving 1,700 Medicare patients with various chronic diseases, found that a telehealth system trimmed spending by 7.7 percent to 13.3 percent, or $312 to $542 per person, each quarter [6].

Dr. Jasper zu Putlitz is president of Palo Alto, Calif.-based Robert Bosch Healthcare, developers of the telehealth system used in the demonstration project cited in Health Affairs. “I think that telehealth will be an integral part of healthcare in the future, not only in developed countries such as the United States, but in emerging countries like India where access to healthcare is really an issue,” he says. The prevalence of smart phones around the world, even in many emerging countries, enhances the capability of telehealth as a versatile, viable delivery system.

Dr. zu Putlitz, a medical doctor by training who practiced internal medicine in Germany and Switzerland for seven years, envisions that telehealth systems will be integrated with electronic medical records and health information exchanges, creating a wealth of integrated data and opening up a world of opportunities for analytics. “You want to be really smart about how you analyze the data,” zu Putlitz cautions. “Whatever we do in analytics should be about enabling the physician to make the right decision. The goal is to take all of the data and create a prediction around what a particular patient will require next and make it a very personalized intervention.”

To date, relatively few healthcare organizations have adopted telehealth. Ironically, one of the reasons often cited is the perceived “non-personal” nature of the technology, but that perception is beginning to change, says zu Putlitz. “Doctors, especially those affiliated with larger healthcare systems, are starting to appreciate all the benefits that come from additional information and all the things that technology can do for them,” he says. “Clearly, there’s still reluctance from doctors who feel that this technology might take away from the patient-doctor relationship, but once they see how teleheatlh actually enables them to take care of their patients in a better, more efficient way, they change their minds.”

No one has to convince the Veterans Health Administration, which runs the largest and one of the most cost-effective healthcare systems in the United States [7], about the merits of telehealth. The VA has been a Bosch customer for 11 years and just extended its contract for another five years. “The VA is absolutely a trailblazer in the use of analytics and telehealth,” zu Putlitz says. “What they are doing with our technology is clearly impressive. They published a study linking telehealth and 17,000 VA patients with chronic disease that showed a tremendous impact – nearly a 20 percent reduction in hospital admissions.” [8]. (For more on telehealth in this issue, click here.)

Bottom Line on Healthcare Reform

The transformation of U.S. healthcare from volume-based system to a value-based system – with the goal of improving efficiency, access and outcomes while reducing costs – demands the realignment of stakeholder incentives and the development of a new payment structure that rewards those collective goals for all concerned.

“The changes that are taking place at the national level are fundamentally changing the way the healthcare system is organized and aligned,” says ICC’s Conti. “Whether it’s a healthcare system, an insurance company or a pharmaceutical company, all three of these players can no longer function in silos. They are going to be aligning themselves in a way that lowers the cost for the entire system and drives value for the individual patient.”

“It’s in the interest of both providers and payers to change their business model to fit the new environment,” says Kaiser Permanente’s Chow. “Costs concerns, resource restraints, increasing demand, better technology – all of these factors will cause change. The people who survive will be the people who can figure out how to adapt. Many consumers are ready to accept a new model of care. Eventually it will have to happen. We don’t have enough doctors and nurses. We don’t have the resources to keep charging for things that we can’t pay for.”

“When healthcare becomes more analytics-based and maybe a little less subjective on a population level, there will be savings and there will be efficiency gains,” adds Bosch’s zu Putlitz. “The payer (insurer) is almost naturally incentivized to do it because the savings would show up where the risk is sitting. The provider side – doctors and hospitals – will see better outcomes. Improving the care of patients is a very strong argument, and it’s a way for providers to differentiate themselves in a crowded, competitive market. Providers are talking to us because they are being held accountable and need to demonstrate quality and value in this new environment.”

And what about the patient, the presumed focus of healthcare reform in the first place?

“If we do it right and healthcare organizations really use advanced analytics to drive better access, quality and outcomes, you as a patient are going to derive more value from the healthcare system,” Conti concludes. “Along with better care and better outcomes, you will get more value for every dollar you spend on healthcare, whether it’s on a doctor, a pill or an insurance plan. The ultimate metric is ‘health per dollar.’ Over time, that will become the gold standard for healthcare.”

Along those same lines, entrepreneur/technologist Palmer gets the last word: “Five years from now we’re going to look back and realize that the companies that adopted predictive and prescriptive analytics are the ones that are going to have a significant advantage in the market. And I think there is no more important application of these high-end analytical technologies and capabilities than developing new, more effective therapies and improving healthcare for patients.”

Peter Horner ( This e-mail address is being protected from spambots. You need JavaScript enabled to view it ) is the editor of Analytics and OR/MS Today, the magazine of membership of the Institute for Operations Research and the Management Sciences. Atanu Basu ( This e-mail address is being protected from spambots. You need JavaScript enabled to view it ) is the founder and of CEO of Ayata, a prescriptive analytics® software company whose customers include Dell, Cisco and Microsoft.