Unnecessary Double Chest CT Scans

Today’s Managing Health Care Costs Indicator is 75,000

Saturday’s New York Times had a great example of using variation to identify waste in the health care system

It’s almost never necessary to do two chest CT scans in a single day – one without contrast, the other with contrast.   The ordering physician should know in advance whether she is looking for disease that requires imaging of the vasculature system. 

It’s bad to do sequential chest CT scans of patients for at least three reasons.  Each chest CT scan is the equivalent of 350 chest x-rays – and we should avoid the extra radiation exposure, which does cause some cases of breast and lung cancer. The  cost of CT scans is high – CMS reports that these duplicate CT scans cost Medicare alone $25 million.  Doing extra tests poses the danger of finding “incidentalomas,” findings that are not relevant to health, but that require additional tests which pose new health risks and additional expenses.

Yet there are some hospitals that do double chest CT scans on almost nine of every ten patients who get a single chest CT.   Many hospitals are under 1% - yet the national average is 5.4%.  75,000 Americans had double chest CT scans in 2008.

I encourage you to look at the interactive geographic map showing excess utilization – it shows pockets of overutilization including Texas, Oklahoma, southern California, and the midsection of the country from Illinois to Mississippi. 

Fee for service payment is one of the culprits here – hospitals with high rates of repeat chest CT scans make more revenue – and for a high fixed cost item like CT scans, make even more margin on this service.   However, there is a straightforward fee for service fix.  We should simply bundle together any two chest CT scans done on the same person at the same facility within 48 hours of each other. 

By the way, CMS also announced on Friday that it will use predictive modeling to proactively identify fraud in health care bills.  CMS until now has paid all submitted bills, and chased any fraudsters retrospectively identified.  Many of those billing CMS fraudulently have disappeared long before Medicare could recoup money -- so this could help lower Medicare costs.

These are two good examples of studying variation to improve health care cost-effectiveness.

Variation Redux: The IOM Weighs In

Today’s Managing Health Care Costs Indicator is 45%

Click to enlarge image 

The Dartmouth Atlas, as highlighted by the 2009 Atul Gawande article reviewing the high utilization of high margin services in McAllen, Texas, has focused attention on the high level of variation in Medicare spending in different hospital referral regions across the country

The Affordable Care Act offers higher Medicare fee updates to physicians practicing in areas of low utilization, and diminished fee increases for those practicing in high utilization areas.

A number of academics have long been suspicious that the Dartmouth methodology overstated differences, and have complained that the Dartmouth researchers

•     Used only Medicare data
•     Did not do substantial risk adjustment
•     Did not adjust for Medicare geographic payment differentials 

The Institute of Medicine just released a long-awaited study of variation of Medicare costs with this additional adjustment completed. Costs of 45 percent of all hospital referral regions changed by at least 10%.

Honolulu remains the lowest cost HRR, and Miami remains the highest.  But the cost in Miami drops from 200% of average to 135% of average.  As you see in the graphic above, the Bronx (NY) drops from 55% higher than average cost to 10% lower than average cost after reasonable adjustment for payment differences, illness and demographics.

This highlights the difficulties of using apparent variation to lower overall health care costs.


Addendum: Thanks to reader MarilynMann - I have fixed the last link.

Variation: Study Suggests Cost Based on Illness More than Delivery System

Today’s Managing Health Care Costs Quote:

“Insurance is expensive because it's paying for medical care, and it won't be affordable unless the medical care it's paying for becomes affordable.”

Ezra Klein  , January 28, 2011

There’s another volley in the ongoing academic food fight about how much provider variation affects variation in health care costs, and how much of variation is about how sick patients are.

The Center for the Study of Health System Change  has just circularized an article published last week in Health Services Research  which looked at a year of claims data from 1.6 million Medicare beneficiaries in communities where the Community Tracking Survey collected independent information about providers.  High cost beneficiaries cost $48,000, while low cost beneficiaries (top and bottom quintile) cost $7000. 

The high cost beneficiaries were older, more likely to be eligible for both Medicare and Medicaid, and had more diseases.    They were more likely to get health care in multiple census districts, and more likely to have a specialist as their usual source of care.  However, they were not more likely to live in areas with higher hospital, physician or specialist supply.

This is a complicated study of Medicare beneficiaries only, and includes scads of statistical manipulation and adjustments.  It’s not proof that supply has no impact on resource cost. However, I believe this is another indication that we can’t expect to solve the health care cost crisis by profiling providers and controlling capital expense alone. 

We’ll need to make real and difficult change in the actual delivery of care to genuinely sick people to get the cost savings we need.    

Smoked Pig

Today’s Managing Health Care Costs Indicator is $2159

Today’s New York Times features another tale from the land of variation.

  

Cardiologist Mark Midei inserted 30 cardiac stents in patients in a single day in 2008. The stent manufacturer (Abbott) purchased a $2159 barbeque dinner including a slow-smoked whole pig for the cardiologists’ home two days later.

The pig is a great headline, and a savory metaphor for the problem of overuse of high-margin procedures, some of which are invasive and have clear associated dangers.  The hospital subsequently notified 585 patients that they might have received medically unnecessary stents.  585!  This is mind-boggling.  

This story, initially reported in the Baltimore Sun, is reminiscent of the Tenet hospital in Redding, California.  That hospital built an empire providing cardiac surgery with very low complication rates – but it turned out that many of the cardiac surgery patients had normal coronary arteries. See Shannon Brownlee's excellent 2007 book Overtreated for the details.  

It’s especially interesting how this physician’s behavior was revealed.  His cardiology group had agreed to be purchased by the local hospital, but the hospital turned around and made him a separate deal and decided not to purchase the entire group. At that point, his spurned group attacked him, and a former chief executive vowed to “spend the rest of my life trying to destroy him personally and professionally.”  The hospital has paid $22 million to settle federal charges of kickbacks to Dr. Midei's former group. 

Most variation in health care is not as extreme as the allegations at St Joseph’s in Baltimore.   Most physicians who practice at the 90^th percentile of resource use don’t realize that they are using more resources than their colleagues, and few believe that they are performing unnecessary care.   We need better reporting on resource utilization and continuing effective peer review.  It’s a pity when peers are only effective at reining in rampant overutilization when they no longer inure benefit from that utilization.

Variation in Practice: Rosiglitazone Case Study

Rosiglitazone, a drug used for treating diabetes, has been shown in recent studies to be highly associated with increased risk of cardiovascular complications. While it's effective lowering blood sugar and helping patients attain a good hemoglobin AIC, which is how we judge diabetes control, it appears to increase the chances of a bad outcome.

The initial study showing that this medicine appeared to be 'trouble' was a metaanalysis published in 2007, and accompanied by an FDA "black box" warning.  The good news is that this new information and warning was associated with an impressive ~45% drop-off in drug utilization.  We often worry about the slow pace of incorporation into practice of new information.  In this instance the communication around the dangers of rosiglitazone looks like a big success.

click to enlarge 

These images come from an article in the New England Journal of Medicine on November 25. 

Here's a worry though.  Look at the geographic variation of use of this drug from 2005 to 2010.

Click image to enlarge 

The dropoff in rosiglitazone use is pretty uniform across the country- about 75% decrease in use for each quartile of previous utilization.  (The authors don't give state-specific data).  However, the northern plains and New England states had a low rate of use of this medicine in the first place -and maintained this over time.  Use of this (expensive) medicine represented an exceptionally large portion of diabetes medication costs in states including Idaho, Utah, Wyoming, Oklahoma, and Kentucky in 2005. Although these states used the highest amount of an exceptionally expensive diabetes medicine - they are not states known for differentially higher quality diabetes care.

This is another illustration of the Dartmouth Atlas contention.   Variations in overall cost usually don't purchase better quality - and sometimes purchase higher risk.

Observations on Managing Health Care Costs (Part 1 of 2)

Happy New Year.

I’ve just finished my first full year of blogging about managing health care costs, and just completed my sixth year of teaching a course “Managing Health Care Costs” at the Harvard School of Public Health.  I wanted to share some observations from our final class of 2009.  Class slides are at this URL.  

This posting will be in two parts – I’ll post the second part tomorrow.

Observation One: Sick people are expensive to care for.

 The top 1% of nonelderly patients represent 30% of all medical costs.  We need programs to better manage those with serious illness, and a regulatory framework to discourage risk shifting and patient dumping.

Observation Two: The problem in the US is much more unit cost than utilization

In the US, we have fewer doctor’s visits, fewer prescriptions, fewer (and shorter) hospitalizations compared to all other developed countries.  But our average hospitalization costs over $12,000, compared to under $10,000 (France, Canada) and under $4000 (UK and Netherlands).

In Japan, MRIs cost under $100, compared to $1500 in the US. 

Observation Three: Our lifestyles cause large medical costs

The good news is that we smoke less than we did. The bad news is that we’re getting more and more overweight. 

Observation Four: We don’t like to make tradeoffs

Everyone agrees we should perform more care that increases quality while lowering cost.  This means we should give more vaccinations –but there aren’t many money-saving medical interventions. We also agree that we should do fewer things that raise cost while lowering quality.  So let’s not give middle aged men Vioxx, which works as well as ibuprofen but increases the risk of heart attack.  The challenge is that we are not willing to give up tiny quality increases at enormous costs. See, for instance, a new cancer drug that for $36,000 a month decreases tumor size in under 1/3 of patients.   We’re also reluctant to design systems that  are 'decrementally cost effective.'   

Observation Five: There is huge variation

Atul Gawande’s “The Cost Conundrum” in the June 1 New Yorker   brought well-deserved attention to the work of Jack Wennberg, Elliot Fisher, and others at Dartmouth who have been showing us the vast variations in utilization of health care.  In expensive areas, we have too many hospital beds and too many doctors – and we use them.  Good example of how decreasing hospital beds does not decrease quality in David Leonhardt’s column in the New York Times this past week. 

Observation Six: Fee for service is toxic

Imagine if we paid auto manufacturers by the bolt rather than for a completed car. We would have cars chock full of bolts –each one an extra cost, extra weight, and an extra ‘point of failure.’  That’s what we’ve got in a fee for service health care system 

Observation Seven: There is a cultural clash between those seeking to preserve the “art” of medicine, and those looking to create more reliability and cost effectiveness through industrial redesign

Jerome Groopman worries that electronic medical records and standardization will take the personal relationship out of medicine.  I worry that lack of accountability and standardization is responsible for many medical errors –and we just can’t rely on the extraordinary effort of individual physicians to insure quality and cost-effectiveness.

Observation Eight: We pay a heavy economic and noneconomic price in our effort to banish uncertainty

It’s our intuition that every additional piece of data increases our knowledge.  This is simply not true. We often gather data in our quest to banish uncertainty, and that data doesn’t much change the chance of real, serious, treatable disease, but does increase cost and increase the risk that we’ll do further, more dangerous tests or interventions. I’ll be blogging more on the vain quest for certainty in the coming days. 

I'll post Part II of these thoughts tomorrow.
[Addendum: thanks to Wellescent Health Blog for note, and I have finished the sentence in observation four]

Part Two of this post is here.

Failure of Massachusetts health care reform?

Last week, Physicians for a National Health Plan and Public Citizen, two advocacy groups, released “Massachusetts’ Plan: A Failed Model for Health Care Reform.”
The authors state that Massachusetts’ reforms
• Have not reduced uninsurance as much as as advertised
• Have not addressed the soaring cost of health care
• Expose those of modest means to large copayments and deductibles
• Reduce the subsidies for “safety net” providers, which threatens access to the most vulnerable
• Increase waste in the system, by using commercial (though nonprofit) health plans and by imposing a fee to support the work of the Connector, the agency that administers the insurance exchange at the heart of health care reform here.
Jon Kingsdale, the executive director of the Commonwealth Health Insurance Connector Authority Board, disputes the statistics on cost increases and the number of uninsured.

This report correctly reports many of the challenges faced by health care reform here in Massachusetts. But let’s be sure not to blame the health care reform for problems beyond its mission, and let’s not forget about the real accomplishments.

Some big accomplishments:
• The Connector merged the individual and small group markets, allowing for better pooling and much more affordable policies for many Massachusetts residents.
• No one argues that the number of uninsured has gone down, and the PNHP document points out that the amount of free care delivered went down by over a third (although the authors complain that this decline should have been 75%)

Some problems beyond the scope of Massachusetts health care reform
• The number of employees in the private workplace has declined due to the economic maelstrom around us. In Massachusetts, we’ve learned before that economic malaise can torpedo efforts to expand health insurance.
• Reductions to safety net providers were mandated by the Bush administration, who threatened to cut off $350 million in federal funds. This can and should be revisited as part of the federal stimulus plans going forward.
• The health care reform authors explicitly avoided addressing the cost issue in the initial legislation. They recognized health care inflation would need to be addressed, and felt this was substantially more difficult than expanding coverage, and should be deferred until we provided coverage to more residents.

We all know that health care reform will ultimately be unsustainable if we don’t control costs. The costs we have to control are in the delivery system – and sharply limiting administrative costs alone will not fund large expansions in access. However, if 1/3 of all health care costs are unnecessary (Fisher, Annals, 2003), it will take a long time to unwind these costs. Perhaps Boston will need one or two fewer teaching hospitals, with dozens of thousands of job losses. It might not be wise to try to eliminate all those extra costs right this moment, when so many other portions of the economy are shrinking so violently.