Thursday, July 2, 2015

Medicare Part D Doesn’t Decrease Hospitalizations or Costs


Today’s Managing Health Care Costs Number is $76 billion



The mid-June issue of Annals of Internal Medicine had an article: “Did Medicare Part D Affect National Trends in Health Outcomes or Hospitalizations?”

This is a question that really matters.  The Medicare Part D program fixed a huge gap in the Medicare plan when it was passed in 2003.  There’s no reasonable health insurance plan that would include no coverage for ambulatory medications – drugs represent some of the most important advances in medical care. Drug coverage is especially are especially important for the elderly, who have suffer from more chronic diseases.

The Medicare Part D program costs $76 billion in public subsidies a year (about 1/7 the total cost of Medicare), serves 37 million Americans, and has been widely heralded as a success.   It was sponsored by the George W. Bush administration, and barely passed the House of Representatives after voting was extended beyond the time allowed by House rules.  Signups were initially plagued by confusion and chaos, but costs ended up lower than Congressional Budget Office projections, and currently Medicare beneficiary satisfaction with the program is high.

There are reasons to believe that increasing access to drug therapy could lower costs.   Treat hypertension to prevent future heart attacks and strokes.   Give asthmatics preventive drugs and they won’t go to the emergency room for treatment.

This study didn’t find that.  There were no differences in measured health outcomes or hospitalizations over the three year study period, despite dramatic increases in access to pharmaceutical therapies. 

 Five years after Part D implementation, no clinically or statistically significant reductions in the prevalence of fair or poor health status or limitations in ADLs or instrumental ADLs, relative to historical trends, were detected. Compared with trends before Part D, no changes in emergency department visits, hospital admissions or days, inpatient costs, or mortality after Part D were seen. Confirmatory analyses were consistent.  (Note ADL is “activities of daily living”)

An accompanying editorial suggests that this lack of effect might have been caused by

-       Higher population illness rates over this time period
-       Lag between decreasing risk and actual improvement in health or costs, and measurement of some metrics unlikely to be helped by drug therapy
-       Many beneficiaries (72%)  had previous drug coverage, which could lead to an underestimate of the impact of Medicare Part D.

I suspect there are other reasons why this study did not show cost savings associated with the beginning of the Medicare drug benefit.

1.     Drugs got more expensive during this time period. This is true of both brand names, especially specialty medications, and generic medications.  This unit price increase is partially due to innovation and  is also due to the decreased price sensitivity due to insurance, and because Medicare Part D has a severely fragmented purchaser community (more than 1000 different Medicare Part D plans), and the legislation prohibited Federal price negotiations

2.     Drugs have been responsible for absolutely HUGE improvements in quality (and quantity) of life.  However, pharmaceutical companies are good at pricing so that any cost saving benefit is captured by themselves.  For instance, hepatitis C drugs will certainly prevent deaths, and will certainly prevent liver transplants. But the cost of those drugs is greater than the potential transplant savings. That’s true over any time horizon.

3. If drugs prolong lives, they might lead to more hospitalization and higher medical costs - but they are still providing real value.

I’m more surprised that Medicare Part D was not associated with a decrease in hospitalizations, although that could come at a later point.  I don’t believe that during the study period (2010 compared to 2000) there were changes in practice that should have led to increases in hospital utilization that would have offset decreases due to Medicare Part D.

There are few drugs that are cost saving outside of tiny subpopulations.  The obvious class of drugs that are cost saving are generic oral contraceptives; these are not relevant in the Medicare population.   Most drugs for chronic conditions are ‘cost effective,’ meaning that we get quality adjusted life years from them for a “reasonable price.”  That’s not the same as cost saving.   So perhaps it was unrealistic to think that offering a drug benefit would lower health care costs.

I support the Medicare drug benefit, and it’s not because it saves money. I support Medicare Part D because you can’t give a patient excellent clinical care in 2015 if they don’t have access to a drug benefit.  Medicare Part D could be improved, and the program could negotiate lower unit costs. But don’t interpret this study to suggest we don’t need a Medicare drug benefit.

Note that bullet point 3 above and the note on increased generic prices are courtesy of one of my colleagues, and added mid-day 7/2.

Tuesday, June 30, 2015

EPA’s mercury regulations would lower health care costs


Today’s Managing Health Care Costs Number is either $6 million or $37 billion


Yes -this is Beijing. We don't want the US to look like this again!

The Supremes made more health care news yesterday – they sent the Environmental Protection Agency (EPA) back to the drawing board to do more cost benefit analysis for its airborne mercury regulations promulgated under the Clean Air Act.  The news articles note that the annual cost of this regulation, known not-so-fondly as “coal killing regulations” by Mitch McConnell and others, will be about $9 billion.   Coal-fired power plants are the largest source of mercury pollution, and mercury is well known to be toxic to nerves.  Opponents claim that the economic benefit will be a mere $6 million – an absolutely terrible deal.

The EPA notes on its web site  
·         The value of the air quality improvements for people's health alone totals $37 billion to $90 billion each year. That means that for every dollar spent to reduce this pollution, Americans get $3-9 in health benefits.
·         The benefits are widely distributed and are especially important to minority and low income populations who are disproportionately impacted by asthma and other debilitating health conditions.
·         Up to 540,000 missed work or "sick" days will be avoided each year, enhancing productivity and lowering health care costs for American families.


Environmental protection bills have striking impact on public health, and public health investments tend to have much larger payoffs than interventions in the medical space.  The reduction in smog in Los Angeles didn’t just mean people could see the mountains – it saved lives.  Some researchers estimate that 6% of all mortality is due to air pollution!


There’s substantial evidence that decrease in lead exposure, much due to regulations under the Clean Air Act, actually decreased violence – and even teen pregnancy. I’ve previously written about the evidence of public health benefits of efforts to decrease carbon emissions.


I don’t how much the mercury regulations will really save in terms of health care costs – it’s probably a lot more than $6 million,but probably lower than $90 billion a year.  The mercury rules apparently remain in effect while the case is reviewed by a lower court.

Monday, June 29, 2015

Marriage Equality Increases Those With Insurance



JAMA Internal Medicine published a review of the impact of New York state’s Marriage Equality Act on insurance coverage for same sex partners on Friday – just in time for the momentous Supreme Court decision confirming that states can no longer discriminate against gays who wish to get married.  The researcher used data from the US Census “American Community Survey,” which has a 97% completion rate because the Census Department doesn’t take “no answer” for an answer.

The results are clear.  Marriage equality has led to far more coverage of same sex partners, even while coverage of opposite sex partners did not rise. (This is before the Affordable Care Act took effect in 2014).

Not only that, the portion of male same sex partners who were insured by Medicaid fell from 3% to 1.4%, a 53% drop. (The absolute drop is larger for women in a same sex relationship, although the percentage drop for this group is "only" 26%.)  Marriage equality had the dual advantages of increasing access to employer sponsored health insurance and decreasing state Medicaid rolls.

Note that y axis is truncated here!

Thursday, June 25, 2015

Supremes Back Obamacare


Today’s Managing Health Care Costs number is Six


That’s six of the Supremes.    John Roberts wrote the opinion, and his fellow conservative Anthony Kennedy joined liberals Bader-Ginsburg, Breyer, Sotomayor and Kagan to uphold federal subsidies even in states that did not establish their own exchanges.  That means that the Affordable Care Act has dodged another legal bullet. 

I was in a train when the word broke –and there were “Six to three” conversations going on absolutely all through my car.

Here are some other numbers that COULD have been the featured number today

·         1332. That’s the section of the ACA that allows the President to waive elements of the ACA in states that want to “go their own way.”  A Democrat could have used this to allow Vermont to go single payer (although that’s now off the table).  A Republican could use that to undermine the individual mandate which could destabilize insurance markets.   Wonkblog
·         6.4 million: That’s how many people would have seen their premiums quadruple if the subsidies were eliminated.  They are concentrated in Texas, Florida and North Carolina. WaPost
·         $1.7 billion:  Monthly cost of the subsidy in those states without their own exchanges  WaPost
·         10%:   That’s how much hospital stocks surged at the news of the Supreme Court decision Vox.com
·        Three: That’s how many other major lawsuits the Wall Street Journal  posits the ACA could face in coming years. (House R’s suit alleging executive branch overreach, Ohio suit alleging no authority to tax, and suits against the Payment Advisory Board after it becomes operational

I breathed a sigh of relief that the insurance market was not thrown into chaos by a ruling against the Affordable Care Act. Many Republicans did, too, as they had little chance to pass meaningful "bridge" legislation to avoid destabilizing insurance markets across the country.  But there could always be another chapter.

Wednesday, June 24, 2015

One More Reason not to have an Anesthesiologist for your Colonoscopy


Today’s Managing Health Care Cost Number is $1.1. billion



We’ve known for a long time that healthy people having screening colonoscopies don’t need 
full anesthesia (or the bills of an anesthesiologist).   The NY Timesreported in 2012 that there are $1.1 billion in unnecessary anesthesiology bills for colonoscopies, and researchers in JAMA reported  that the portion of colonoscopies that had an anesthesiologist bill doubled between 2003 and 2009 to almost a third.

Now, we know another reason not to have an anesthesiologist present at your colonoscopy.  A northern Virginia anesthesiologist just lost a $500,000 malpractice suit from a patient who had inadvertently recorded the procedure and found that she spent much of the procedure mocking him.   The excerpts are frightening.

Conscious sedation (light anesthesia) administered by the gastroenterologist is enough.  There’s no reason to pay an extra $1000 or more.  I don’t feel differently about this just because the anesthesiologist’s bills might be covered with no cost sharing since this is a preventive procedure.   Waste is waste, whoever is paying for it!

Get your colonoscopy (every 10 years from age 50 on unless you’re at higher risk or have had a polyp or cancer or have a worrisome family history).  But pass on the anesthesiologist.

Surprise: Cancer Drugs are Not Value Priced


Today’s Managing Health Care Costs Number is $168,366



Source (Note that the line the authors drew here is NOT statistically significant. It’s best to view this graphic as if there is no line in place and you’ll see that the pattern appears random.

JAMA Oncology, a new journal, published a research letter earlier this month trying to establish a relationship between efficacy of new cancer drugs (approved 2009-2013) and their prices.   Simple answer – there was no relationship.  There was also no relationship between price and whether the new drug was a novel compound or “next in class,” and only a small difference in price among drugs which were approved by the FDA based on response rate, survival improvement, or time to progression of disease.

New cancer drugs are expensive – often topping $100,000 a year or a course of treatment.  Some new cancer drugs, like Nivolomab and Pembrolizumab  for  melanoma and selected lung cancers, are miracles – they sometimes convert a death sentence into a long-term chronic disease.  (This class of drugs also has a great name – “Programmed Death Receptor blocking antibodies.”  Sounds so science fiction!)   Many new oncology drugs, though, offer at best a month or two of increased survival, sometimes with bad side effects.    

It’s nice to know that drugs are not priced based on value, but what should we do about that?

A new project by the American Society of Clinical Oncology (ASCO) to publicly report on the cost and relative effectiveness of cancer drugs is important, but flawed.  Patients and their treating doctors should certainly know both the cost and the relative efficacy of medications used in treatment, and that’s especially true for drugs that can clear out a family’ life savings.   Anything to bring more transparency here is a good idea.

The challenge is that increasingly these oncology drugs are aimed at genetic targets -- and so will be enormously effective in a small portion of the population, and close to useless in others.  The best answer wouldn’t necessarily be knowing that a compound derived from thalidomide costs $150,000 and leads to an extra 7 months of life.   We’d really want to know how much extra life each drug would lead to depending on the genetic characteristics of the cancer and the host.  We need to know how valuable the drug would be in a population that is very similar to the relevant patient, not an unselected population. 

But that’s not the way medical literature is reported in traditional peer reviewed journals, which themselves are subject to substantial publication bias.  (This means that negative trials are rarely reported, leading to a rosy view of the impact of new medications).

A combination of publication bias and nonrepresentative patient populations (healthier, younger and more likely to respond) included in published studies makes us more optimistic of the impact of new drugs than we should be when we think about the impact on an individual patient.

This over-optimism will help propel us to be willing to pay for even irrationally priced oncology drugs.

Note that $168,366 is the annual cost of Omacetaxine, which treats chronic myelocytic leukemia.  14.3% of users have a major response, and the mean duration of response is 12.5 months.

Monday, June 22, 2015

Is Health Plan Consolidation Bad for Consumers?


Today’s Managing Health Care Costs Number is $47.5 billion

Health Plan Share Prices, Previous 12 months
It’s been a busy weekend of health plan mating rituals.  Anthem made public its $47.5 billion bid to take over its smaller rival CIGNA on Saturday, and CIGNA promptly said no, although the bid was 30% higher than CIGNA’s stock price had been just before rumors of a takeover offer hit the internet.    Aetna made its own offer for Humana on Saturday as well.   Humana hasn’t yet responded; there could be other offers, and it seems highly unlikely that Humana will remain an independent entity at the end of the dance.  United Health Group was rumored to be preparing an offer for Aetna, and Aetna/CIGNA combinations have also been suggested.

Each of the big five health plans has gained over 50% in value over the last year. The Dow Jones Industrial Average has been up 6.5% over the last year, and the Standard and Poor’s index has been up 7.4%.

Health plans have good reasons to consolidate:
·         Health plan infrastructure is expensive, and combined carriers don’t need two sets of C-suite executives and don’t need two sets of IT systems, although consolidating systems can take years.
·         Bigger health plans can push providers for better rates and demand participation in narrow networks.  
·         Less health plan competition can allow for higher administrative services only (ASO) fees and less pressure for aggressive performance guarantees– which can lead to higher margins, and higher costs for employers purchasing their services.

Some commentators have suggested that this health plan consolidation could lead to decreased competition – which could lead to higher cost of medical care.  The evidence suggests otherwise. 

RAND evaluated this in 2011, and showed that increased insurer market concentration was associated with lower hospital prices.  In fact, hospital prices in the most concentrated insurer markets were 12% lower than hospital prices in more competitive health plan markets.   Researchers from Harvard relooked at this in 2014, and found that

a market with two equal size insurers is associated with 3.9 percent lower medical care spending per capita (p = .002) and 5.0 percent lower prices for health care services relative to one with three equal size insurers (p < .001).

This suggests that the loss of one or more of the four (or five) largest health plans could actually help lead to lower hospital costs.  Larger health plans purchasing provider services can push for lower rates and better terms, which could help counteract our exceptionally high unit costs.   I expect that much of the opposition to health plan consolidation will come from the provider community, which has been consolidating rapidly and increasing unit costs over the last decade.