PWC Estimates 9% Health Care Cost Increase in 2011

Price Waterhouse Cooper  published its projections for health care inflation in 2011 last week.  PWC projects that medical costs will rise by 9.0% - better than the 9.5% from 2010 – but certainly not a game changer. Further, since inflation is likely to remain low or nonexistent and the GDP is likely to be close to flat, the portion of GDP dedicated to health care will continue to increase.

PWC suggests that the factors that are pushing health care costs down include:
-          Move to generic medications
o        Lipitor goes generic in 2011. Note that the huge cost savings are generally 18 months after the first generic comes out; but there are some savings even in year one
-          Increasing patient cost-share
o        This increases price sensitivity, and decreases demand. Above is a summary of the PWC survey on increased cost-shifting to members/patients

-          End to COBRA subsidies.
o        Members who have recently lost their jobs tend to have very high claims costs, and the much higher unemployment coulpled with the federal subsidy of 2/3 of the cost of COBRA has swollen the number of people on this program.

The factors pushing health care costs up include:
-          Medicare cost-shift.
o        Medicare underpays hospitals and other providers less than commercial, employer-sponsored plans (although not as badly as it would if the 21% physician fee schedule decrease had not just been reversed for another six months by Congress.)  The PPACA (health care reform) decreases rate of increase for hospital fees – so the relative underpayment will increase.  While Medicare rates are still substantially higher than payment rates in other countries, they are almost always below
-          Provider consolidation.  
o        Hospitals are merging, and they are acquiring physician practices.   There’s been a dramatic shift of cardiologists from private practice to hospital employment over the past year with reimbursement changes, for example.  Larger provider groups mean less competition and higher prices
-          Health Care Information Technology
o        PWC notes that the largest expenses for implementing Healthcare IT are front-loaded in 2011 and 2012 – and suggests that the savings will accrue later.
o        The New York Times http://www.nytimes.com/2010/06/27/business/27digi.html?hpw had an article today suggesting that EMRs can allow faster physician bill transmission, which can lead to quicker adjudication (payment), and lower administrative costs.  
o        It’s also possible that EMRs will lead to better charge capture, which can increase health care costs as long as we stay in a system of fee for service reimbursement.

All told, the PWC estimate of health care cost increase is especially distressing in light of the sour economy.  

Medicare Chemotherapy Rates: No Good Policy Goes Unpunished

Click to enlarge graphic 
Many private practice oncologists have made a substantial portion of their income from the “markup” on chemotherapy they administer in their offices.  The GAO calculated that oncologists were paid six times the acquisition cost for the chemotherapy agent Paclitaxel in 2004.


The oncologists have long argued, convincingly, that the ‘profit’ from chemotherapy makes up in part for the poor payment they receive for the cognitive work they do, including careful deliberation about the most serious of diagnoses, and nonbillable patient and family counseling.   The financial incentive to give more chemotherapy, and to prefer brand-name (high margin) chemotherapy, has always seemed like a bad idea. 

Most of us don’t ever want to need an oncologist. But if we do, we don’t want the oncologist to make much more money if she treats us with more expensive chemotherapy!

Medicare made this system dramatically more rational six years ago.   CMS initially moved the price from 95% of Average Wholesale Price to 85% of AWP, and then moved to 6% above average sales price.   AWP is a fiction – and almost all medications can be purchased at dramatic reductions from AWP.  Estimates suggested that the markup for chemotherapy agents from Medicare decreased on average from 22% to 6%. 

Health Affairs  just published an evaluation of the results of Medicare’s change in payment methodology on the web.  The article was briefly mentioned in yesterday’s New York Times.  The researchers reviewed Medicare claims data for over 200,000 patients with the diagnosis of lung cancer. 

The results are stunning. In the context of these payment cuts, physicians
è Prescribed chemotherapy to more patients (18.9% got treatment within a month of diagnosis, compared to 16.5% before the payment changes
è Dramatically increased the use of one medicine, docetaxel, where price decreased by just 8% and where a high price means that the 6% margin represented a larger profit for the oncologist.
è Paclitaxel use fell, and it represented a much lower portion of all chemotherapy administered.  (The authors present a graph showing a huge decline, but when I multiply out the percent of those given chemotherapy with the likelihood they got Paclitaxel, it appears that the utilization is flat while the utilization of many other drugs increased.   I’ve put these calculations in this appendix) http://isites.harvard.edu/fs/docs/icb.topic659062.files/chemotherapy%20paclitaxel.xls
è Another chemotherapy drug which did not have a dramatic price cut nonetheless showed a drop in utilization

The authors note that economic theory suggests that when physicians face a fee cut that will have a large impact on their income, they increase utilization.  On the other hand, when there is a fee decrease that impacts merely a small share of physician income, the affected services have lower utilization.

This is an important insight as we review how to control medical inflation.  Many support a large recalibration of health care reimbursements to move income from procedural specialties to cognitive specialties  -- especially primary care.  However, this study suggests that very large or across-the-board fee schedule decreases would be likely to lead to much higher utilization in affected specialties.



Economist Vs. Economist: Will Health Care Reform Lower Costs?

This month’s Health Affairs has sequential articles by David Cutler  and Douglas Holtz-Eakin  debating whether the health care reform bill will REALLY save money.  Cutler has been an advisor to the Obama campaign and the White House.  Holtz-Eakin was the director of the Congressional Budget Office, and was an advisory to John McCain’s Presidential campaign.

Cutler’s points:
1)       We can achieve a trend 1.5% lower than the current trend; we did it in the “managed care” era.
2)       The projected savings from health care reform increase in the second decade
3)       Administrative waste, clinical variation and medical errors are ripe for cost savings

Cutler argues that Medicare reductions will demand improved value from the health care delivery system.  He notes that bundled payments, accountable care organizations, pay for performance and better coordination can all improve care and lower costs. 

Holtz-Eakin’s points:
1)       We have a fiscal train wreck, with government spending out of control
2)       Tennessee had to reverse its broad coverage approach when TennCare proved vastly more expensive than expected; Massachusetts has maintained its high rate of insurance, but the cost has been far higher than expected.
3)       There are some unfunded provisions, including new dollars for the IRS to collect penalties, and $20 billion a year to avoid those 21% fee decreases for physicians under the Sustainable Growth Revenue targets, which have been reversed every year.
4)       We collect premiums for long-term community care in the early years, but that should be segregated to pay for benefits. (He notes the same concern about increased Social Security revenue projected by the CBO).

Holtz-Eakin and coauthor Michael Ramlet suggest that the federal government will likely back down on some Medicare fee decreases, and further suggests that the Feds will not collect the excise (“Cadillac”) tax at the projected levels.

Neither of these represents dispassionate analysis. Each represents a political point of view. Cutler might well be overoptimistic in some of his assumptions.  Holtz-Eakin and Ramlet appear to me to be overly pessimistic.  In fact, the physician fee decreases are the only Medicare cuts that Congress passed but that were never enacted. Here’s a link to a Commonwealth Fund  report showing that the CBO has historically underestimated cost savings from health care reform efforts.

I’m sure the actual financial results will fall somewhere between these two poles. The important thing to keep in mind is that the default of doing nothing was not viable.  Without action, we were likely to end up with premiums that were more and more unaffordable, and higher and higher portions of the GDP going to health care despite fewer people having benefits.   With health care reform, we have an opportunity to rethink health care delivery.   The bill as passed seems to me to be the “least bad” of many alternatives – including the alternative of not passing anything during this Congress.


“As Good as it Gets”: RAND's Evaluation of Health Care Reform Bill


RAND researchers Elizabeth McGlynn et al have used a microsimulation model to conclude that the health care reform bill signed into law did about as good a job of expanding coverage without increasing the bill (much) as we could reasonably expect within the confines of the real political world.


The researchers did a sensitivity analysis with multiple variables, including:
1)       Varying individual or employer penalties for not obtaining or providing health insurance.  Researchers found that lower penalties increased the cost of expanded coverage.  Increased penalties lowered the cost of increased coverage – but were not likely politically palatable.  
2)       Varying the threshold for Medicaid eligibility. If this is lower than the federal poverty level, the rate of uninsurance remains high.   If it is set above 133%, there is more ”crowd out” with members leaving employer-sponsored plans, which increases the cost to government.
3)       Varying the restrictions on increased costs for older enrollees.,

All modeling was done as if there was a single national exchange, and the researchers did not consider penalties collected (essentially discounted these at 100%).

The researchers also evaluated which scenarios led to the highest value for consumers.  Invariably, there was a proportionate relationship between government spending and value to consumers – so to reliably give more benefit to consumers, government spending would have to increase. 

In the graphic above, the origin (red square) is the health care reform bill as passed.  Area 1 represents less government spending AND more people insured.  It’s the smallest area – meaning the fewest of the simulations were here.  All of these were judged by the authors to be political non-starters,.  Area 2 is unequivocally worse outcomes – higher government cost with fewer new enrollees covered.  Area 3 represents more coverage and higher spending, while Area 4 represents less coverage and lower spending.  (3B and 3B represent better ‘value’ – in that there is less government spending for each newly insured person).

We all know what’s wrong with the Patient Protection and Affordable Care Act (PPACA).  We wish that its cost saving was more iron-clad, and we wish some of the benefits came more quickly. We are worried that some of the cost savings might be overstated.  Having said that, the sausage-factory that is Congress ultimately passed a bill that does an admirable job of increasing coverage and being prudent with taxpayer dollars.  

Unanticipated Consequences: Shorter Hospitalizations and More Readmissions


A JAMA article last week showed that shorter lengths of stays for congestive heart failure over the last decade and a half have been associated with higher levels of readmissions.  

The study is impressive.  In each time period there are close to a million fee for service Medicare members admitted, and data is presented both with risk adjustment and in raw form. (The numbers are so large that the risk adjustment doesn't matter so much).  The study was based on claims data only, and the press accounts I've read have emphasized that this data represents a failure of the health care delivery system

We can do better.  Studies have shown that a simple intervention like a post-discharge phone call can prevent  many readmissions  The good news is there is a real spotlight on hospital readmissions, since Medicare will start paying hospitals with high readmission less in 2014.

There is good news in this study that  has been largely overlooked. Mortality rates have tumbled dramatically.

What does this mean for readmissions?  Those surviving initial hospitalization are dramatically sicker.  (Indeed, 53% more are being discharged to nursing facilities). These patients with more serious underlying disease might in fact be more likely to be readmitted to the hospital regardless of length of stay.  It's also possible that the increase in discharge to nursing homes with vigilant professionals picks up more patient decompensation - so some who might have died at home are transferred for yet another hospitalization.

Society is getting a higher survival rate because health care for those with late-stage congestive heart failure has improved dramatically since the 1990s.  The additional quality adjusted life years we are gaining are not coming without cost.


Pharmaceutical Detailing

The Wall Street Journal just published an overview of pharmaceutical company detailing - or provision of samples to physician offices. The report the WSJ reviewed is required by the new health care reform bill.

The main participants:
Pfizer: 101 million prescriptions worth $2.7 billion
Merck: 39 million prescriptions worth $356 million
Wyeth 52 million prescriptions worth $64 million
(Wyeth was acquired by Pfizer)
Abbot: 16 million prescriptions worth $32 million

Of note, different companies accounted differently for prescription value (retail vs. acquisition cost) or prescription unit (per pill or per package).

What do the drug companies get for their samples?

They offer samples for high margin medicines.  In most cases, physicians could prescribe lower cost (and therefore higher value) alternatives.   Once a patient starts on the high-cost brand name, it's unlikely he or she will switch to a generic.

The two brand name drugs are mentioned in the WSJ article as samples that are often provided by Pfizer and Lilly cost around $5 per day.  Generic medications cost as little as 50 cents per day (Source: Drugstore.com.)  There is no evidence of increased efficacy of the (expensive) brand name medications.

Many physician offices have eliminated pharmaceutical detailing altogether.   Those offices prescribe more generics, and leave their patients with lower overall drug bills.  This disclosure, along with disclosure of pharmaceutical and medical device company payments to physicians for consulting and other services, can help drive public policy and ultimately decrease medical trend. 

Gaining Cost Savings from Decreasing Variation: More Difficult Than It Looks!

 The New York Times has an excellent front page story today examining the work of researchers at Dartmouth who have shown huge variations of Medicare costs across the country.  The Dartmouth data has often shown that the relationship between cost and quality appears inverse – and many have inferred that we can have higher quality and lower cost at the same time.

The article reviews some of the critiques of the Dartmouth Atlas research, including the fact that the Dartmouth researchers focused on cost only, with no assessment of quality.

I’ve pointed out that the cost savings estimates are overly optimistic before. Here’s a link to a post from January 2009 on this topic. This was around the time that Professor Richard Cooper of Wharton published some provocative data showing that Medicare payment rates appeared inversely proportional to payments by private health plans.  Therefore, the areas like Louisiana that had VERY high Medicare payments did not have more specialists (even pre-Katrina). Rather, they had less cost-shift from Medicare to private payers.

The Dartmouth researchers have done a great service – and utilization differences are real and important across the country.  Their approach isn’t perfect – and too many people have extrapolated too optimistically about how easy it will be to save money based on decreasing utilization.  We have to be realistic.   People who live around the corner from 4 MRI machines will always have more MRIs than people who have to drive four hours in each direction to get to an MRI machine!

Another important point.   Compared to other countries, our main issue is cost per unit, not utilization.  Let’s be sure to decrease utilization – and in doing so we’ll probably improve quality, and at least not make it worse.  But first and foremost we have to address the problem of very high cost per unit for each service we deliver in the US.