PHYSICIANS and HEALTH CARE REFORM Commentaries and Controversies

Commenting on the President’s budget, an editorial in the Times on Feb 2^nd juxtaposed three of our nation’s dilemmas: the deficit, jobs and health care.

“President Obama got his priorities mostly right. The deficit, compared with what it could have been, is $120B. That’s a lot of money. But it’s not too much at a time of economic weakness, when deficit spending is needed to put Americans back to work.”

“Medicare and Medicaid will cost $788B; that should be another reminder of why the country needs health care reform.”

The fundamental question about health care spending is, therefore, what does it mean for jobs?  Approximately 15 million people work in health care, and that doesn’t count jobs at the 140 companies that specialize in constructing health care facilities or 56,000 pharmacies or the dreaded health insurance companies, nor does it include all of the jobs of people supplying goods and services to the 18 million folks engaged in health care in these various ways.  But a more important question is, where is the job growth? The answer is health care. Over the past decade, the growth in health care jobs has equaled the total growth of jobs. Many are high-skilled, but many are entry-level jobs that help to move people out of poverty. So we had better be careful in measuring the impact of health care. Quite apart from its beneficial effects on well being, it just may be the engine of the economy.

A friend – a highly respected health economist – expressed concern that my criticism of the Dartmouth group’s studies “is a justification to do nothing.” I told him that I, too, have been concerned that my critiques might be taken that way, but they shouldn’t. What I have tried to do is to prevent the wrong things from being done. Regulations based on geographic variation will do nothing to improve the quality curve. Quality is not geographic. But poverty is, and despite the Dartmouth folk’s denial, what the studies of geographic variation demonstrate is that poverty is a major contributor to health care costs. That’s why Mayo shuns poor Medicaid and Medicare patients. However, I have had an even bigger concern - that the Dartmouth group’s twisted logic would eventually be figured out, and its close association with the Orzag-Obama plan would undermine health care reform. It’s difficult to conclude that this was the problem in Massachusetts, but “lack clarity,” “arrogance of ideas” and similar expressions of distrust with health care reform have been cited as reasons for the election’s outcome. Twisted truths are not a basis for sound policy. We’ve got to get this right if health care reform is to proceed on a rational basis.

The favorite sound bite of Dartmouth disciples is to compare some high cost locale with a low cost locale. First it was Miami vs. Mayo, then Birmingham vs. Grand Junction, then Los Angeles vs. Green Bay and now it’s Los Angeles vs. Portland. This time, Tom Brokaw delivered the message on Meet the Press: “At UCLA Medical Center, they spend $92,000 on the last two years of a life, but in Portland, Oregon, just north of there (it’s actually 825 miles north of there), they spend $52,000 because they’ve got better controls on Medicare.  So until you begin to pay for value and pay for performance, health care reform is not going to work.”

What do Miami, Birmingham and Los Angeles have in common, and what do Rochester MN (home of Mayo), Grand Junction CO, Green Bay WI and Portland OR have in common. One thing is poverty. The maps below show the density of poverty in each (light green shows census tracks with 20-40% poverty and red shows tracks with >40% poverty).Miami, Los Angeles and Birmingham all have poverty ghettos. There certainly is poverty in the other cities, but not in ghettos. The dense poverty ghetto in central Los Angeles includes more than 2 million poor people.

So is it “better controls” or less poverty that results in lower spending. The first has a certain appeal, and we do seek better ways of doing things, but the latter is what’s really driving spending higher, and nowhere more vividly that in Los Angeles. Poverty is expensive, and ghetto poverty is really expensive. To paraphrase Mr. Brokaw, until you begin to solve the problem of poverty, health care reform is not going to work.

Does David Leonhardt, the New York Times publicist for the Dartmouth group, really care if reports about geographic differences in health care are the whole truth? You be the judge when you read his latest article. It’s true, as Leonhardt says, that the supply of hospital beds in Richmond VA decreased between 1996 and the present, and it’s likely that health care didn’t suffer, as he reports. But the whole truth is that Richmond is no exception. Bed supply decreased everywhere, as health care shifted from inpatient to outpatient. Richmond didn’t decrease its beds while everyone else kept extra beds functioning at increased costs. Operating under certificate-of-need, Richmond’s bed supply decreased 15% between 1996 and 2008. But data from the Kaiser Family Foundation show that bed supply in the US as a whole decreased twice as much – 29% between 1990 and 2007, and 35 of the other states also had certificate-of-need. So even with its decreases, Richmond still has more beds than the national average and more than other regions in Virginia, and that doesn’t even count the Richmond VA, which has been growing all the while, principally with patients in the Medicare age group. Of course, what Richmond also has is low population, low population density, lower-than-average poverty and, like similar towns across America, lower health care costs overall. Forcing bed closures will not save money and certainly not lives. In fact, it’s likely to cost more of both. Come-on, tell the Times readers the whole truth – stop peddling Dartmouth malarkey. It might allow us to design valid improvements for our health care system.

Provisions in the Senate and House health care reform bills propose to reallocate resources based on geographic differences in Medicare spending. While well intended, they will penalize providers who care for the poor and impair access for these vulnerable patients.

A reallocation of resources to lower-cost states has been endorsed by members of Congress from states with lower Medicare spending who believe that, by receiving less from Medicare, their states are currently being penalized for being “efficient.” However, it is not efficiency that accounts for their lower spending. It is less poverty and better health status.

The map shows (in blue) the ten states represented by members of Congress who have publicly endorsed reallocation plans based on geographic differences. Six neighboring states with similar characteristics are lightly shaded. On average, the Medicare spending in these 16 states is 20% less than in the rest of the country.

The “low-cost” states cover almost 40% of the land mass of the US but encompass only 14% of the population and only 3% of the African-American population. While they include many prominent cities, there are no major urban centers with the dense zones of poverty, as are found in Chicago, Los Angeles and New York. Nor are there broad bands of poverty, as are found in Louisiana, Mississippi, Alabama and southern Texas. Yet it is in “poverty ghettos” and broad “poverty regions” that health status is poorest and health care spending is greatest.

 We must not confuse the added costs of caring for the poor with inefficiency in health care. The greatest “inefficiency” is poverty. The US will never slow the growth of health care spending unless it addresses the special needs of its most disadvantaged citizens. Health care reform should assist the hospitals and physicians who care for them. Unfortunately, a number of sections of the current bills do just the opposite.

Hospital readmissions. Section 3025 of the Senate bill (Hospital Readmissions Reduction Program) and its companion Section 1151 of the House bill (Reducing Potentially Preventable Hospital Readmissions) would penalize hospitals that have higher rates of readmissions. While increased rates may reflect substandard care in some hospitals, the more common reason for higher rates is more patients who have complex diseases processes and little social support, most of whom are poor. Indeed, when fully adjusted for severity of disease, most inter-hospital differences in readmission rates disappear.

Value and efficiency. Because providers in counties where poverty is prevalent have higher per-beneficiary spending, they would be classified as “inefficient” under Section 1123 of the House bill (Payments for Efficient Areas), while providers in the 20% of counties that have the lowest Medicare expenditures would receive a 5% bonus. In like manner, Section 3001 of the Senate bill (Hospital Value-Based Purchasing Program) would reward hospitals that have lower per-beneficiary costs for certain defined conditions (acute MI, congestive health failure, pneumonia, etc.), although it is known that expenditures for such conditions are much greater among low-income patients. I am hopeful that the Senate bill will not include the Finance Committee’s call for penalties for physicians whose resource use is in the highest decile, which would mainly affect those whose practices include poor patients with multiple comorbidities.

IOM study of geographic variation. The same logic pattern that has been applied to readmission policy and to “value-based purchasing” exists in Sections 1159 and 1160 of the House bill, which instructs the Institute of Medicine to develop payment policies based on geographic differences in health care, with the assumption that differences in the Dartmouth Atlas are relevant to cost containment. Yet MedPAC has shown that, even without adjusting for income, much of the variation disappears with adjustment for health status. But even though the bill instructs the IOM to consider income and other social determinants, there are no standards that can be applied nationally to adjust adequately for these factors. On the other hand, there are partial remedies for their effects, such as the wider use of interpreters and transitional care coordinators, as Section 1151(8) of the House bill proposes to support. 

Efficiency and value are important goals, but variation in their geographic distribution is principally a reflection of variation in poverty and the prevalence of disease. That should not be a surprise. We all know that poverty varies geographically, and we know that the care of the poor is expensive. If health care costs are to be constrained, it must be through addressing the needs of low-income patients; not through penalizing the providers who care for them.

Respectfully,

When the New York switches sides, it’s probably true, and now, after years of extolling the virtues of Dartmouth’s geographic malarkey, the Times has published an article about the work conducted by Tom Rosenthal and his colleagues at UCLA and elsewhere in the U Cal system demonstrating that more care is not wasteful, as the Dartmouth-Orszag team professes. It saves lives, which is what health care is all about. Senators should take note before they vote on their historic bill, which will penalize hospitals and physicians who provide added care for patients who are sicker and often poorer. It’s not too late to do it right.

Be sure to read the four-part series in the Oakland Tribune by Suzanne Bohan and Sandy Kleffman describing life and death in the Bay Area’s poorest ZIP Codes. 

“Closing the gap in life expectancy between those perched on different rungs of the socioeconomic ladder isn’t just unfinished work in the Civil Rights Movement, health advocates say. It’s critical to controlling runaway health care costs.”

Part 1  

Part 2    

Part 3    

Part 4

On the heels of the American Hospital Association’s recent demonstration of gross discrepancies in the Dartmouth group’s data, MedPAC released its December 2009 report to Congress showing the same. Confirming data for 2000 (reported in their 2003 report), MedPAC demonstrated much less variation among states and metropolitan statistical areas (MSAs) than described by Dartmouth for states or hospital referral regions (HRRs). Closer scrutiny of MedPAC’s data reveals even more.

Adjustments. First, the Dartmouth group has claimed that they adjusted their measures of Medicare spending for age, sex, race, mortality, disease incidence and prices. But Dartmouth’s adjusted data are indistinguishable from MedPAC’s  unadjusted data, both among states (in 2000) and HRRs (in 2006), as shown to the left. This confirms suspicions that Dartmouth’s “adjustments” are all shadows and mirrors, or just malarkey.

Adjusted variation. Second, as reported by MedPAC and consistent with the above, MedPAC found much less variation in Medicare spending among MSAs after adjusting for prices, health status and special payments than Dartmouth found among HRRs after supposedly adjusting for prices and a host of other factors. The two figures below demonstrate that. One need only look at the broadly dispersed bars in the illustration of Dartmouth variation and the more tightly packed ones in MedPAC’s.

Sociodemographic realities. But, despite finding much less geographic variation, MedPAC pointed out that there still was plenty. The greatest likelihood is that most of this residual variation is related to differences in income, which MedPAC does not account for (except as it correlates with health status). The final graphic supports that view. It distinguishes a cluster of eight southern states, which house 89 MSAs, from the other forty states (Alaska and Hawaii were excluded), which house 312. Compared to the other forty, the “southern eight” has a poverty rate that is 55% higher (89% higher for blacks); its rate of uninsurance is 31% higher (89% higher for blacks); and its mortality rate is 16% higher (30% higher for blacks). Excluding the “southern eight,” Medicare spending is within 10% of the mean in 87% of the MSAs in the other forty states. Only five MSAs out of 312 fall beyond +15%.

While no one would deny that the practice of medicine varies among practitioners for a host of reasons, both good and bad, such variation is not responsible for geographic variation, or at least not usually. Rather, geographic variation in health care spending reflects geographic variation in prices and variation in two patient-related factors: income and burden of disease. As I said before, the Dartmouth Atlas is the “Wrong Map for Health Care Reform.” Getting to the core of geographic variation can help us get health care reform right.

According to MedPAC, 18% of hospitalizations among Medicare beneficiaries resulted in readmission within 30 days, accounting for $15 billion in spending. Since treatable chronic illnesses are responsible for many such hospitalizations, it is assumed that they represent failures of the health care system. MedPAC claims that 84% of readmissions are potentially preventable. However, as will become evident, most readmissions reflect differences in co-morbidities, poverty and other social determinants, all of which deserve attention, including better transition care, but few of which are under the control of hospitals. Nonetheless, health care reform assumes that regulators can accurately adjust for such risks and estimate the “excess.”

Both the House and Senate bills include reductions in payments to hospitals with “excess” readmissions. Payment would be reduced 20% for “excess” readmissions within seven days and 10% within fifteen days. Hospitals with 30-day risk-adjusted readmission rates above the 75th percentile would incur penalties of 10-20%, scaled to the time to readmission.

The legislation gives ultimate authority to the Secretary of Health to define national and hospital-specific benchmarks according to methodologies that would be determined by the Secretary and that would be free of administrative or judicial review.

So, just how many hospitals are over the threshold? The three illustrations below present the dilemma for one category of disease: congestive health failure. Each presents risk-adjusted data for more than 4,000 hospitals. MedPAC risk-adjusted using 3M’s “all patient refined diagnosis related groups” (APR-DRGs), a proprietary package that defines the severity of illness for 314 indications. CMS used their own statistical model that estimates the independent effects of age, gender, past medical history and approximately 40 comorbidities. The research group at Yale headed by Harlan Krumholz, the nation’s most prominent cardiac epidemiologist, used a hierarchical logistic regression model for patients with heart failure, drawing on age, gender, 26 comorbidity variables and 9 cardiovascular variables.

All of the hospitals in the upper 25% as evaluated by MedPAC had readmission rates that were more than 2% above the mean, but that was true for only half of the hospitals as evaluated by CMS and for fewer than 10% of such hospitals as evaluated by the Krumholz group. Indeed, when CMS put its assessment to the statistical test, it found that, despite the wide variation depicted, only 5% of hospitals had readmission rates that were statistically “worse than the US average,” far short of the 25% that would be dunned.

This poor policy is a tragic manifestation of Dartmouth malarkey, which has created the false belief that reducing geographic variation could save enough to make health care reform possible – the “30% solution.”  The result, instead, would be serious damage to hospitals that care for the most vulnerable patients, a large proportion of which are safety-net hospitals and academic medical centers, and that’s no way to reform health care.