Healthcare Delivery in the US

  • Mark L. Braunstein
  • Mark L. Braunstein
Chapter
Part of the SpringerBriefs in Computer Science book series (BRIEFSCOMPUTER)

Abstract

The United States has a uniquely complex and expensive healthcare system. We are alone among the industrialized countries in not having a “single payer” or at least a single entity responsible for making the rules. As a result, each individual health provider may have to deal with dozens of different health plans, each tailored by the patient’s employer to try to manage rising health costs. This complexity adds significantly to administrative costs which are estimated at 25-30% of spending. One study suggests that US administrative costs at 31% are proportionately nearly twice those in Canada. [1] Many studies show that we spend around twice as much on healthcare as compared to our peer nations. Yet we get relatively poor results, particularly for routine public health issues and for managing chronic diseases, the problems that affect most people and drive most healthcare costs. It is beyond the scope of this book to examine the merits of the various proposed solutions to these problems but the belief that it can help with them is the core rationale for federal funding of the deployment of health informatics.

The United States has a uniquely complex and expensive healthcare system. We are alone among the industrialized countries in not having a “single payer” or at least a single entity responsible for making the rules. As a result, each individual health provider  1 may have to deal with dozens of different health plans, each tailored by the patient’s employer to try to manage rising health costs. This complexity adds significantly to administrative costs which are estimated at 25-30% of spending. One study suggests that US healthcare administrative costs at 31% are proportionately nearly twice those in Canada. [1] Many studies show that we spend around twice as much on healthcare as compared to our peer nations. Yet we get relatively poor results, particularly for routine public health issues and for managing chronic diseases, the problems that affect most people and drive most healthcare costs. It is beyond the scope of this book to examine the merits of the various proposed solutions to these problems but the belief that it can help with them is the core rationale for federal funding of the deployment of health informatics.

The Chronic Disease Problem

Most Americans think of healthcare in terms of dramatic, high-technology, life-saving interventions. It’s what they see on television; it’s what they hear about when a friend, neighbor, relative or celebrity develops cancer or some other serious condition. Arguably, the US has the best system for providing that sort of care, as shown in Fig. 1.1. Patients here needing what Intermountain Health’s Dr. Brent James, a national thought leader in health quality improvement, calls “rescue care” have a better chance of survival than in other advanced industrialized nations.
Fig. 1.1

Mortality in Life Threatening Situations (lower is better)adapted from and courtesy of Dr. Brent James of Intermountain Health

However, well over 90% of Medicare spending is for patients with chronic diseases such as diabetes, hypertension and coronary artery disease. [2] I’m aware of unpublished data that suggests that around 70% of spending for younger families whose care is paid for by an employer is due to chronic disease. Key economic differences between the acute high technology care we excel at here in the US and the care of chronic disease are that chronic disease care takes place mostly outside hospitals, uses little advanced medical technology and is, therefore, relatively inexpensive and less profitable for hospitals and providers.

While inexpensive to treat directly, if not managed well, chronic diseases cause complications that are expensive to treat. For example, poorly controlled hypertension can lead to stroke, heart attack or kidney failure. Moreover, chronic diseases can cause other chronic diseases. Diabetes, for example, is a major risk factor for heart disease. This compounds the cost problem because those patients with multiple conditions account for a much greater proportionate share of spending. [2]

The US healthcare system is poorly designed to manage chronic disease, in large part for historic reasons (Fig. 1.2).2 For millennia morbidity was mostly due to infectious diseases. In the 20thCentury life expectancy increased substantially as a result of improved public health and sanitation, vaccinations and the antibiotics that became widely available after World War II. In the past few decades lifestyles here in the US became more sedentary and highly caloric, fast and processed food became a larger part of the American diet leading to a striking growth in obesity, a major cause of chronic disease. Today, only a few decades from when most people died of infections, the “single greatest cause of rising healthcare spending in the U.S. is the growing prevalence [emphasis mine] of chronic disease.” [3] That point is worth re-emphasizing. Here in the US we have a much higher rate of chronic disease than in other countries. It is the health problem driving the largest share of increasing healthcare costs.
Fig. 1.2

The Evolution of US Healthcare Over the Past Few Decades

However, despite these substantial changes in disease patterns, the healthcare system remains relatively unchanged. A core reason is the nature of current economic incentives. For example, “Medicare continues to operate under a fee-for-service model, which complicates the adoption of chronic care treatment models.” [2] In most communities healthcare is dominated by the local hospital, a place where little of the care for chronic disease can or should be done. Physicians offering high technology procedures and treatments make far more money than those offering primary care, the mainstay of chronic disease management.

Any medical student already knows this. Primary care physicians such as family doctors, general internists, pediatricians and gynecologists can expect to make far less than they would earn as a surgeon or specialist. [4] As a result, we have a shortage of primary care physicians, the front line warriors in the battle against chronic disease. [5] To leverage those scarce resources and achieve better results requires a different approach than the traditional physician-centric, episodic office visit with little or no contact with the patient managing their disease at home between these visits.

Why a different approach? Consider any acute, life-threatening problem. It is likely to be diagnosed and treated within a relatively short period of time. The care will often take place in one highly specialized venue where care coordination, though hardly guaranteed, is substantially simpler and the patient is essentially a passive participant.

Contrast that to a diagnosis of chronic disease which, by definition, is not curable. Treatment will take place over years or even decades in multiple venues that may be widely geographically dispersed. The complexity of coordinating care is compounded with multiple chronic diseases, particularly in our highly fragmented and specialized system of care. So “while the average Medicare beneficiary sees between six and seven different physicians, beneficiaries with five or more chronic conditions see almost 14 different physicians in a year [emphasis mine] and average 37 physician visits annually”. People with five or more chronic conditions fill almost 50 prescriptions in a year. [6] In fact “virtually all of the [Medicare] spending growth since 1987 can be traced to patients treated for five or more conditions.” [2]

Success in treating chronic disease depends heavily on the behavior of patients when they are outside of what is traditionally thought of as the healthcare system. Not only are these diseases often caused by issues such as poor diet and lack of exercise but proper diet, exercise and medication compliance is almost always critical to successfully managing them. New technologies such as the Internet and wireless and mobile devices are of great interest because of their potential to engage these patients at home and change their behavior both to prevent and more successfully manage chronic disease.

Health Data Logistics

“Our healthcare system needs all the help it can get. And health information technology is some of the best medicine we have.”

Tommy G. Thompson [7]

Among the key problems for which information technology is clearly good medicine is insufficient data sharing among the many care providers treating patients with chronic diseases, particularly those with multiple diseases. According to the IOM, the “fact that more than 40 percent of people with chronic conditions have more than one such condition argues strongly for more sophisticated mechanisms to communicate and coordinate care. Yet physician groups, hospitals, and other healthcare organizations operate as silos, often providing care without the benefit of complete information about the patient’s condition, medical history, services provided in other settings, or medications prescribed by other clinicians.” [8]

The practice of medicine is largely comprised of collecting data, analyzing that data, making decisions based on it, conveying those decisions to other members of a care team and the patient, following what happens, and adjusting according. Any effective coordination of care depends on making a patient’s data available when and where it is needed. In essence, we have a “health data logistics problem”. That is what the national effort to deploy health informatics is primarily aimed at solving.

This is very analogous to problems other industries have identified and solved using contemporary information technology. For example, Wal-Mart famously scans every item sold in its stores and transmits the information, as sales occur, to databases at its Arkansas headquarters where it is virtually instantaneously available for analysis and action. If snow shovels are fast sellers in Minnesota, Wal-Mart makes sure more are on the way. If swim suits are languishing in Florida, it slows their resupply. Wal-Mart can only do this because it has a standard electronic record of every sales transaction and a network in place to seamlessly and quickly move that digital information from where it is generated to where it is needed.

Countless manufacturers have similar systems in place to manage a global supply chain and logistics network feeding components to their plants. This is the sort of technology that makes “just in time” possible. Today most automobile manufacturers rely on suppliers to deliver entire and often even customized subassemblies as they are needed for installation in a particular car. Their computer systems are integrated with those of the companies that supply them so the entire process can be orchestrated efficiently without error or delay.

Paper records are hard to share. The original can physically exist in only one place. They can be copied and faxed but this is slow, time consuming and relatively error prone. The results may not be very legible. The information contained in them is rarely standardized except where required for billing.

Yet, in many respects, managing chronic disease is a “data logistics problem” comparable to those we just described. More effective management of these increasingly common problems will depend to a large degree on creating the same data liquidity that drives all current global logistics and supply chains. Of great, or perhaps even equal, importance is the need to substantially improve care processes based on this wider access to more timely and accurate data.

Consider a seemingly simple issue: when should a chronic disease patient next be seen by their physician? The traditional approach is somewhat arbitrary but is hopefully guided by experience and intuition. The next appointment is scheduled in three months, six months or a year. If Wal-Mart used this approach of stocking their inventory on a fixed schedule unneeded swim suits might arrive at Florida stores even though the shelves are overflowing. Potential sales of snow shovels in Minnesota might be lost as customers go elsewhere because the inventory is exhausted.

Perhaps Mr. Smith, a diabetic patient who just saw his physician, will do fine. He will follow the prescribed diet and take his medications in the proper amount and at the proper time. If so, he won’t need to be seen in six months, as scheduled, but, in most clinics, the physician will see him anyway, congratulate him on his success and schedule the next visit after another six months.

At the same time Mrs. Jones, the hypertensive patient who came in right after Mr. Smith, decides she is feeling fine and does not need her medications which, after all, are quite expensive on her limited budget. Two months before her scheduled return visit she suffers a stroke and ends up in the hospital and, after discharge, requires months of expensive, painful care and rehabilitation.

Healthcare deals poorly with these patients in large part because traditional practice models only deal with them during visits. This provider-centric, office-based view of care is partly historic and largely the result of our reimbursement system. Physicians have normally only been paid when they physically see patients. Moreover, unlike Wal-Mart and the auto manufacturers, care providers typically don’t have a “real time” view of how their patients are doing.

A “data logistics” approach would capture data at the source – the home – and transmit it to where it is needed so Mr. Smith is not brought back for an un-needed visit and Mrs. Jones is brought back when her hypertension is clearly starting to get out of control, possibly preventing her stroke.

Changing Processes

“The hospital is altogether the most complex human organization ever devised.”

Peter Drucker [9]

The IOM goes on to say that “If we want safer, higher-quality care, we will need to have redesigned systems of care” which “must be designed to serve the needs of patients, and to ensure that they are fully informed, retain control and participate in care delivery whenever possible, and receive care that is respectful of their values and preferences.” Finally, it recognized the critical importance of improved management of chronic disease which it said “needs to be a collaborative, multidisciplinary process.” [8]

The IOM identified six challenges healthcare organizations would face in developing these new care systems. The first was “to redesign care processes to serve more effectively the needs of the chronically ill for coordinated, seamless care across settings and clinicians and over time.” [8]

We’ve seen that information technology has been utilized by many other industries to transform business processes. Consider another example from an organization that is not typically associated with high technology – the US Postal Service (USPS®). We all dread going to the post office to send packages, particularly during busy holiday periods. The predictable waiting line is a major factor that drives many individual consumers to use a UPS® or FEDEX® store, even if they are more expensive. So, in response, the USPS has introduced “Click-N-Ship®”. Go to their website, answer a few questions and print the bar coded shipping label. Once it is attached to the package you can go to the head of the line and drop your package off at the counter and leave without any interaction with the clerk.

This is a transformed business process only possible through an innovative application of technology. Make things convenient and people are more likely to use them. Hopefully, you can see that it has many similarities to the data logistics approach to chronic disease described at the end of the previous section. Among these are the use of the Internet to bridge time and space by taking advantage of the availability of technology in the home. Most importantly, both depend on the willingness to “think outside the box” with respect to how we do business. If the USPS can do it, why can’t the healthcare system?

Perverse Incentives

“A problem is something you have hopes of changing. Anything else is a fact of life.”

C R Smith[10]

The answer has a lot to do with financial incentives. If the USPS can transfer the entire package preparation and drop off function to the consumer, they reap any resulting financial benefits. It costs less to process the package because their customer is doing much of the work. Their customer is more likely to use their service because it is more convenient.

Though there are clear financial benefits to avoiding unnecessary care, the beneficiary is usually the employer that insures their own health benefit, an insurance company or government. In most circumstances, if a physician uses technology to avoid unnecessary care, visits to the emergency department or hospitalizations, the physician’s income is certainly not increased and may even be diminished. Moreover, any time spent by staff using some new technology-mediated process may represent lost income. To make matters even worse, up until very recently, if a practice wanted to utilize technology to facilitate an innovative care model, it would bear the entire investment.

This conundrum, in which there is a mismatch between who has to invest and who might benefit, is characteristic of what my colleague, Bill Rouse, and others refer to as a “complex adaptive system”. [11] In such a system there are multiple “independent agents” each of whom are intelligent, adapt to changing conditions and act to optimize their own self-interest. Moreover, no entity is in control.

So, despite the availability of enabling technology to transform healthcare, adoption would continue to lag without incentives to bridge the self-interests of the independent agents. Solutions have been tried for decades. For example, in a health maintenance organization (HMO) employers typically contract for the care of their employees on a fixed annual basis. If the care costs more, the HMO has to make up the difference and loses money. If it costs less, the HMO benefits financially. This is a clear example of trying to change incentives. This contrasts sharply with traditional fee-for-service model where providers typically make more money the more services they provide. There is room to argue about whether the fee-for-service approach leads to the delivery of unneeded services but few could argue that it provides a disincentive to invest in technologies to replace traditional approaches to services that can be reimbursed with new technology-based alternatives that are not.

A landmark 2008 survey of 2,758 physicians [12] strongly supports this conclusion. Despite the long time availability of electronic health record systems with the capability to improve clinical decision making and expedite the management of prescriptions and other orders, in 2008 only 4% of physicians had a “fully functional” system that could do those things. At the top of the reasons given for not having one were the cost and the uncertainty about achieving a return on the investment. Interestingly, physicians who had implemented such a system had a high degree of satisfaction and saw clinical benefits to their patients, a part of the survey not as often cited as the low adoption numbers. Some might be tempted to be critical of physicians based on this, but they are acting rationally and exactly as the theory of a complex adaptive system predicts – in a manner to maximize their own perceived self-interest.

Fixing this requires new incentives. In our highly fragmented health system usually only Medicare which, according to MEDPAC (the advisor to Congress on Medicare issues), paid for some 23% of personal healthcare spending in 2010 [13], has the size and market presence to introduce transformative change. This study and other data led to the decision by the federal government to essentially pay the cost of electronic health record technology in provider offices and hospitals if those systems are used in a manner called “Meaningful Use” that could transform care, at least modestly. In parallel, the Affordable Care Act of 2010 introduced Medicare Accountable Care Organizations (ACO), to create financial incentives, similar to those in an HMO, but utilizing a contractual arrangement with existing community-based care providers. Many major private health insurance companies are now also adapting a similar model. The needed new incentives may now be in place. We turn to them next.

Footnotes

  1. 1.

    I have generally used the more inclusive term “provider” in preference to “physician”. Provider includes physicians and other professionals such as dentists, nurses and nurse practitioners and, increasingly, care coordinators. The major exception is cases where I feel a system is quite specifically designed for use by physicians.

  2. 2.

    For these, among other reasons, we compare poorly to the other industrialized countries in many measures of health and public health in particular. For the official statistics on this visit http://stats.oecd.org/ and click on Health Status.

References

  1. 1.
    Woolhandler S, Campbell T and Himmelstein DU (2003) Costs of Healthcare Administration in the United States and Canada. N Engl J Med; 349:768–775CrossRefGoogle Scholar
  2. 2.
    Thorpe KE and Howard DH (2006) The Rise In Spending Among Medicare Beneficiaries: The Role Of Chronic Disease Prevalence and Changes in Treatment Intensity. Health Affairs 25 (2006): w378–w388CrossRefGoogle Scholar
  3. 3.
    Kumar S and Nigmatullin A (2010) Exploring the impact of management of chronic illnesses through prevention on the U.S. healthcare delivery system – A closed loop system’s modeling study. Information Knowledge Systems Management 9:127–152CrossRefGoogle Scholar
  4. 4.
    Medscape Physician Compensation Report: 2011 http://www.medscape.com/features/slideshow/compensation/2011/ Accessed 19 July 2012
  5. 5.
  6. 6.
    Anderson G and Horvath J (2004) The Growing Burden of Chronic Disease in America. Public Health Reports 119:May–June 2004Google Scholar
  7. 7.
  8. 8.
    Committee on Quality of Healthcare in America (2001) Crossing the Quality Chasm: A New Health System for the 21st Century. The National Academies PressGoogle Scholar
  9. 9.
    Drucker P (2002) They’re Not Employees, They’re People. Harvard Business ReviewGoogle Scholar
  10. 10.
    Smith CR (1969) Publishers Weekly, September 8, 1969Google Scholar
  11. 11.
    Rouse WB (2008) Healthcare as a Complex Adaptive System. The Bridge, Vol. 38, Issue 1, pp. 17–25Google Scholar
  12. 12.
    DesRoches CM et al (2008) Electronic Health Records in Ambulatory Care— A National Survey of Physicians. N Engl J Med 359:50-60CrossRefGoogle Scholar
  13. 13.
    http://www.medpac.gov/documents/Jun12DataBookEntireReport.pdf Accessed 15 September2012Google Scholar

Copyright information

© The Author(s) 2013

Authors and Affiliations

  • Mark L. Braunstein
    • 1
  • Mark L. Braunstein
    • 2
  1. 1.School of Interactive Computing College of ComputingAtlantaUSA
  2. 2.Health Systems Institute for People and Technology Georgia Institute of TechnologyAtlantaUSA

Personalised recommendations