How to incentivize hospitals?

Let's say you are the head of the Centers for Medicare and Medicaid Services (CMS) in the United States and are trying to answer a simple question: how should CMS incentivize U.S. hospitals? In an ideal world, hospitals that provide superior clinical care while minimizing costs should be optimally rewarded. The problem is, of course, that those two goals - good care for cheap - are often at odds. So we're back to our original question:

How should a government that pays for healthcare (for at least a sub-set of its citizens) incentivize hospitals?

From a cost perspective, it's a fairly simple question. You provide a set payment to a hospital for a set service and allow the free market to work its magic. Hospitals that keep their own costs lower than the government's payment make a profit and hospitals that do not control costs lose money. I'm going to call this the cost reduction handle. By constantly squeezing payment rates for various healthcare services, the government incentivizes hospitals on cost controls.

But what happens if a hospital swings entirely from the cost reduction handle at the expense of clinical quality? Ethical concerns aside, it is not hard to imagine that hospitals who provide inferior care may end up having lower costs and therefore higher profits than hospitals who provide superior care. As a result, the government needs to establish some type of quality improvement handle to incentivize hospitals on clinical quality. If a government can establish a mechanism for continuously monitoring clinical care within hospitals and punishing under-performers, it can incentivize clinical quality.

Here-in lies the rub. Measuring clinical quality is notoriously difficult. What metrics should we use to compare clinical quality in the care of ischemic stroke? Once we have decided which metrics to use, how do we develop the infrastructure to compare outcomes across U.S. hospitals? The systems simply don't yet exist.

In an ideal world, hospitals would be stretched between the cost reduction handle and quality improvement handle so that at no point could they sacrifice clinical care for cost controls or cost controls for clinical care.

Sometimes It's Good To Steal

Given my interests in health system performance, I was drawn to this New York Times article highlighting how the Swiss and Dutch Health Systems - as hybrid private-public models for universal heatlhcare insurance - are in vogue among U.S. politicians. Other than to briefly note that citizens in these two European countries are forced to purchase their own health insurance at the risk of penalty, the article is light on the details.

It's still unclear to me whether these health systems are in the press because they are effective (from a quality and cost perspective) or because they are politically feasible in a United States political environment that is increasingly looking towards universal health insurance but that is also wary of single-payer models of universal coverage. Nonetheless an interesting article and something I'm going to read up on before my next post.

A Nursing Shortage

In 2005 the health division of the OECD - a best practices research group for the world's wealthy countries - published a working paper entitled Tackling Nurse Shortages in OECD Countries. The authors - Steven Simoens, Mike Villeneuve and Jeremy Hurst - take a diagnostic look at nursing shortages in multiple nations and take a stab at identifying best practices employed by nations to either increase the number of nurses entering the workforce or reduce the number of nurses exiting the workforce.

It's an illuminating read. From a macro-economic policy perspective, it will open your eyes to how little we know about the factors that affect nursing recruitment and retention. And this is despite the fact that nurses represent the single largest employee group in the hospital environment. I am always fascinated by the things we learn from our peers and this paper highlights several putative best practices employed in socialized healthcare systems.

Among the highlights. In Ireland, the government set up the Nursing Careers Center in 1998 to promote and market nursing as a career. In Canada, the national health system has introduced various flexible scheduling and family care initiatives to improve job satisfaction among nurses and aid in retention. Yet, it is surprising how little hard data we have on these measures. What initiatives are most effective at recruiting and retaining nurses? What initiative are most cost effective?

Technology in Public Health

With our first world country bias, we tend to view advancements in medical technologies through the lens of capital technology. We measure clinical advancement in 64-slice CT scanners, intraoperative imaging suites and innovative radiation therapy delivery platforms.

However, we tend to forget that for the vast majority of human beings - people without access to clean water and nutrition - medical innovation can come in more basic forms. Scientists at Proctor & Gamble, in collaboration with the U.S. Center for Disease Control (CDC) have made amazing advancements in water purification technology. Though it emerged more than a year ago, the PUR water filtration technology is worth highlighting for its ingenuity and potential to improve access to clean water in developing countries. The size of a sugar packet, a PUR packet costs less than seven cents and can purify up to 10 liters of pathogen-laced water. Given that pathogen-acquired diarrhea is a leading worldwide killer of children under the age of five and that contaminated drinking water is a prime cause, any improvement in access to purified drinking water will dramatically reduce childhood mortality in the developing world.

It's easy to forget that, for the vast majority of this planet's inhabitants, access to basic needs is a necessary pre-requisite to improving public health. First, the clean water and reliable food source. Then, the 64-slice CT scanners.

More Transparency, This Time By the MHA

The Massachusetts Hospital Association recently posted data on the number of patient falls per 1,000 inpatient days and various other measures of clinical outcomes for heart attack, heart failure and pneumonia care. It's an admirable move and more evidence that full clinical transparency, if not here, is at least gathering steam.

The more transparency there is in hospital outcomes, the easier it will be for hospitals to identify negative and positive deviants. At the very least, such information will help hospital administrators identify areas for improvement. And, if you believe that the age of consumerism in healthcare is already here, perhaps such information will allow consumers to apply strong market pressure to under-performing institutions.

Choosing a Specialty

Yesterday I was speaking to a friend of mine who will be entering medical school in the fall. When I asked him what specialties interested him, our conversation shifted towards a discussion of the relative advantages and disadvantages of the different medical specialties.

Why become a neurologist? Why become a dermatologist?

Presumably medical students - as any other consumer of goods or services - respond to market incentives. Certain specialties involve more patient contact, others are better compensated and still others have better qualities of life. In making their decision, medical students must weigh personal interests against debt loads, training times, expected financial rewards and quality of life.

Neurosurgeons, though well compensated, have lengthy residencies. Although they train for less time, neurologists are relatively poorly compensated and often participate in grueling call schedules. Simply on a financial basis, it is not surprising that medical students graduating with over $200,000 in debt gravitate towards higher-compensating specialties and avoid residencies in lower-compensating fields like family medicine and pediatrics.

As an outsider, it's easy to forget that both monetary and non-monetary incentives are highly likely to influence the medical specialty, region and city that a medical student selects.

Psychological or Physical

Perhaps I shouldn't be, but I am constantly surprised by the links that are uncovered between psychological and physiological health. Today the Associated Press reported on a recently published study in which individuals in bad personal relationships had a significantly greater risk of heart disease than individuals in good personal relationships. While this study certainly doesn't prove causation (merely correlation), it is indicative of both the basic challenges of medical research in humans and the greater challenge of disseminating research findings to the lay public.

Commenting on the results of the study, one expert noted "it is still not clear what to recommend. Do we tell people who have negative relationships to get therapy? They may have other reasons to do so, but I see no basis for them doing so only to avoid a heart attack." Ending a bad marriage is not necessarily the answer either, he said, given evidence that being unmarried also could be a risk.

When even the experts appear at a loss, how is the public supposed to internalize the health ramifications of a clinical study?

Networking Hospitals to Improve Care

For better or for worse, we Americans live in a country that is generally suspicious of anything that smells even remotely of government interference in free markets. With that in mind, I came across an interesting article in the Wall Street Journal discussing how state governments are passing laws mandating a multi-tiered system for hospitals providing trauma care.

What is the problem that these laws attempt to address? To be blunt, you wouldn't want to be a trauma patient at a rural community hospital that lacks the staff, technology or care protocols to optimally treat your injuries. However, in many states there are few regulations mandating transfer protocols for severe cases.

In brief, these state laws attempt to address this problem by differentiating hospitals' trauma capabilities and ensuring that the most serious trauma cases get routed to the most advanced trauma center as rapidly as possible. If a patient arrives at a rural hospital with serious injuries, they are immediately transferred by ambulance or helicopter to the nearest level I trauma center.

There is an interesting ethical question here. If patients in states with trauma networks have substantially better outcomes than trauma patients in states without trauma networks, isn't there an ethical mandate to create these trauma networks in other states as well? And if the networks improve outcomes for trauma, might networks for other medical needs also improve outcomes? There is already evidence to suggest that stroke networks founded on the same clinical principles as trauma networks - rapid patient transfer to the best equipped hospitals - improve patient outcomes.

It will be interesting to watch if these trauma networks continue to proliferate among the states and serve as models for other clinical pathologies. There are certainly many hurdles left to clear - a distrust of government interference, resistance from hospitals and regulatory challenges in enforcement, to say the least.

How Does Innovation Spread?

In a word, slowly. Our colleagues in the field of psychology tell us that human beings are incredibly resistant to change. But what about in the healthcare industry where innovation is expected and even demanded of physicians and hospitals? Thousands of clinical studies are published every year and based just on the sheer volume of clinical literature, one might assume that medicine is universally dynamic, infused with an ethical mandate to roll-out innovations as rapidly as possible. But even if we assume that a true innovation in clinical care has been identified, historical experiences suggest that the healthcare industry is as resistant to change - if not more so - than other industries.

Caused by blood clots in the brain's blood vessels, ischemic stroke is a leading cause of death and disability in the United States. Prior to the mid-1990s, there were no effective treatments for ischemic stroke. Although ischemic stroke patients arriving at the hospital would receive supportive medical care, there were no known pharmaceutical or surgical interventions proven to lyse or remove the blood clot. However, in 1996, on the strength of a study published in the New England Journal of Medicine, the FDA approved the intravenous delivery of the clot-busting tissue plasminogen activator (tPA) as the first treatment for acute ischemic stroke.

The first true therapy for ischemic stroke had been approved, the optimists sat back and waited for a dramatic spike in the use of tPA and a concomitant improvement in clinical outcomes.

However, despite tPA's FDA approval and a rash of subsequent studies confirming the drug's clinical efficacy for treating ischemic stroke, the use of tPA remained surprisingly low. Even today, only 1-5 percent of ischemic stroke patients nationally receive tPA. The reasons are many. First, to be eligible for tPA, patients must present to the hospital within three hours of stroke onset. Second, there are a host of physiological contraindications to tPA that disqualify many patients. Yet, even among patients who are eligible for tPA, the national treatment rates range from 15-20 percent.

Why aren't more eligible ischemic stroke patients receiving tPA? Innovation spreads slowly.

Transparency

NetDoc recently combined data on hospital quality outcomes - as reported by hospitals to the Department of Health & Human Services - with Google Maps to provide a diagnostic snapshot of hospital quality. In grading each individual hospital, the tool combines data on heart attacks, heart failure, pneumonia, surgical infection prevention to create an overall "clinical quality metric" for each hospital.

This website is just one more indicator of the slow shift towards transparency in the U.S. healthcare system. Information is the mediator of free market competition and an increase in transparency in hospital performance will play an important role in improving overall quality of care. Patients will be able to make more informed decisions about providers and hospitals will emulate positive deviants in the market to improve outcomes and stay competitive.

Measuring Things

I recently read Better: A Surgeon's Notes on Performance, a wonderful collection of essays penned by Atul Gawande, a general surgeon and frequent contributor to The New Yorker. I was struck by one passage in particular:

"Regardless of what one ultimately does in medicine - or outside of medicine for that matter - one should be a scientist in this world . In the simplest terms, this means one should count something. The clinician might count the number of patients who develop a particular complication from treatment - or just how many are actually seen on time and how many are made to wait."

How can a hospital improve clinical outcomes without first understanding how its outcomes compare to outcomes at peer institutions? The larger argument is that in any given endeavor - medicine, baseball, farming - there are positive deviants that, by definition, outperform the average. Just as some baseball players consistently bat above the league average, some physicians and some hospitals consistently create superior clinical outcomes. In the case of hospitals, how much could they improve the quality of care delivered if they could identify and emulate positive clinical deviants among other hospitals? Why is the Mayo Clinic more successful at treating pituitary tumors than 95 percent of other U.S. hospitals? What are they doing right?

At the Organization for Economic Cooperation and Development (OECD), a sort of best-practices organization for the world's developed nations, researchers are developing a comprehensive set of healthcare quality indicators that will eventually allow policy makers to compare clinical performance across healthcare systems. The development of metrics for the assessment of clinical quality - breast cancer survival rates, stroke 30-day mortality rates and colorectal cancer screening rates - will help answer important questions. What does the French healthcare system do better than the Spanish healthcare system? Why does it do it better?

I think they're great questions and will hold important lessons for individual institutions and systems.