CMS Concurrent Hospice Demo

CMS has announced the hospice concurrent care demonstration that I blogged about in March 2014. The big idea in concurrent hospice care is that people can receive services from a hospice provider without having to “unelect” curative treatments, which has been required to receive hospice since 1983. Concurrent palliative care was one of three “non covered” benefits that around half of Medicare beneficiaries said they would fund via reductions of other care in some of my past work. This is not exactly that, however, because the demo announced yesterday requires that a patient be hospice-eligible, which means a physician certifies they are likely to die within 6 months.

A few thoughts on this.

  • Cudos to Medicare for innovating. In hospice, Medicare has always led the way, and will do so again. Further, over 8 in 10 people who die annually in the U.S. are Medicare beneficiaries, so getting end of life care straight in Medicare is key.
  • The biggest limitation to the demo is the fact that patients must be hospice eligible (a physician must say they are likely to die within 6 months). I would prefer to see a concurrent demo that attempted to push further up the disease course, say to the last 12-15 months of life.
  • Because a patient is hospice eligible, the $400 payment that hospice providers receive per month, is less than what they would receive in roughly 3 days of providing full fledged hospice care (after unelection of curative). Since the patient is able to continue receiving curative care, the notion is the services provided by hospice providers will be less intensive. However, my read is that the hospice is on the hook for delivering all of the hospice benefit. One thing to watch in the evaluation: how many patients start this and later stop the hospice concurrent care demo because they want more care than perhaps the hospice planned to provide (I think most hospice providers will do phone based monitoring, but we will see what happens).
  • In one sense it is surprising so many hospice providers applied. There was much grousing about the low payment and fact that providers have to be prepared to deliver the full hospice benefit. However, I think many thought they needed to be involved in such a demo, and the general concept of concurrent care is the way most who look closely think this end of life care should go.
  • Another key metric for the demo will be conversion rate of patients who start this demo into full-fledged hospice. In one sense, why would they? However, hospice providers will have lots of incentive to get them to do so, both due to payment and their normative belief that what they do is best for dying patients.
  • The evaluation design presents lots of interesting opportunities. Demo hospice providers, hospices who wanted to participate and weren’t selected, and other providers. Non experimental inference is key for policy research and this will be an interesting one from a methodological perspective.
  • We have a CMMI demonstration of early palliative care (further up the disease course) with Four Seasons Hospice and we are attempting to evaluate the impact on quality and cost to Medicare of this model (CMMI billboard May 16). The announcement that Medicare will begin paying for advanced care planning in January introduced an intervention into our control group, and maybe even the demo. There there is another new intervention to be accounted for.

New MEDPAC study says some past work (mine) overstates hospice savings

A MEDPAC contractor report considered the question of “does hospice reduce Medicare costs?” from three perspectives, and concluded that a matched-control approach that only compares costs from the time of hospice initiation to death overstates savings. This is the approach that we used in 2007, and that was replicated (and improved upon in terms of covariates) by Amy Kelley and colleagues in 2013.

The essence of the approach we used is to take hospice decedents, match them to otherwise statistically similar decedents, and then compare costs for the period of hospice use (if used for 10 days, compare only last 10 days of life).

I will blog more about this next week, but a couple of quick thoughts:

  • The MEDPAC report notes that the methods we used to pick a control group and that Amy Kelley adapted are not the most appropriate way to address the question “does hospice reduce Medicare spending?” I think the method we used is better than that alternatives, but I want to leave that aside for a second and say that all of the action in a non-experimental study of this type is how you pick the control group. As compared to what? is always the most important policy question.
  • In our ongoing work on palliative care, we are planning to use multiple methods of picking a control group, and further plan to think deeply about why the answer to the seemingly simple cost saving question can differ so much by method. There is information in this uncertainty, and non experimental inference is just about the most important thing for policy evaluation that will never be subjected to a randomized trial.
  • The answer to the question could differ by market area, as well as across time as an area becomes “fully exposed” to hospice and palliative care concepts. This part of the report (that I have not read carefully yet) looks to be very important. It is hard to think that the early hospice cost assessments that looked at the last year of life–for example–when the median stay was 15 days is the best way to evaluate this question. A market approach might be better.
  • I agree with MEDPAC that cost savings is not the most important question to address. Here is a list of many others.
  • Non experimental policy evaluation is hard. On January 1, 2016, about halfway through our CMMI project, Medicare will start to finance advance care planning. That means any control group in the last 18 months of our study had ready access to a key palliative care intervention that was not present from the beginning. I am not sure exactly what the answer to all these questions are, but I am confident that more data is better than less.

Medicare will pay docs to discuss dying

Medicare announced today that it will pay physicians to discuss dying, and the preferences, choices and options that patients have when they face the inevitable (it is only a matter of when, and from what). Six Augusts ago, the summer congressional recess exploded into “death panels” as a way to argue against the ACA. The offending provision would have simply paid physicians to have this type of planning discussion. Five Augusts ago, a RCT of early palliative care showed that patients with stage-4 lung cancer who essentially had such “goals of care” discussions–but who could subsequently choose whatever care options they wanted–actually lived longer, had better quality of life and cost a little less as compared to those who did not have such a discussion.

The idea that Medicare wouldn’t pay for a discussion of preferences and options for patients is absurd since over 8 in 10 persons who died last year in the U.S. were insured by the program. Medicare is inherently in the dying business, and each of us will do it once. The policy announced today will allow physicians to get paid to have goals of care discussions with patients and family members (see p 246), which is important to help patients to make the most informed choice possible. This policy represents a small step toward sanity in this area.

We at Duke have a Center for Medicare and Medicaid Innovation Award on palliative care with Four Seasons, which is seeking more to develop more comprehensive payment changes to how the Medicare program pays for end of life care, but we are still early days in the project (CMMI billboard May 16).

King v Burwell is a total win for the ACA

Chief Justice Roberts saved the ACA again, this time with a 6-3 ruling (Kennedy also joined the liberals) that says that tax credits can flow to persons in states that did not set up their own exchange, but who buy private coverage via

The case does not depend upon the so-called Chevron doctrine (where the executive branch of the govt is granted discretion when there is ambiguity in how to read a law).  Instead, the majority says that because tax credits (and therefore the ability of people to afford health insurance) is such a fundamental part of the law, that the plainest reading of a narrow section of the law couldn’t be read to mean that tax credits are not not available in all states, regardless of the type of exchange through which they get coverage.

This is a complete victory for the Obama Adminstration. A reporter just asked me “what does the North Carolina General Assembly need to do now?” Answer: they can keep doing what they have been doing on health care for quite a while: nothing.

I hope that they will not continue to do this, and that our state will work out a way to put together a reform of our Medicaid program (that I am in favor of), as well as a means of expanding insurance coverage (that I am also in favor of). Here is a concrete idea for how to do this that I put forth in January, 2014.

There have been some statements that they were holding back due to the uncertainty caused by the looming King v Burwell case. There is no longer any uncertainty on that front.

What will SCOTUS do? [Callie Gable guest post].

As the date of the Supreme Court ruling in King v Burwell looms, I knew that any self respecting health policy blogger should put forth some prediction. Alas, I could not muster the energy of a thoughtful one in the midst of an R01 deadline. Luckily, Callie Gable, a rising senior at Duke this fall is working with me, and so I outsourced the musings. Callie wrote this last Summer when she interned with Reihan Salam at the National Review, and she is one of the best health policy students I have ever taught, so listen up when you hear her name. You can follow her on twitter @CallieGable


The days to conjure up a futile prediction about the King v Burwell decision are running out. It appears that the Supreme Court will finally end months of speculation on Thursday or Friday, one of the two new decision days that were just added to the calendar. Despite the best efforts of wonks everywhere, the outcome of this case is still basically unpredictable – but my money is on a ruling that keeps the subsidies flowing.

The plaintiffs win the argument but lose the case. The actual text is plain, straightforward, and unambiguous. As Justice Scalia said, “is it not the case that if the only reasonable interpretation of a particular provision produces disastrous consequences in the rest of the statute, it nonetheless means what it says.” But context trumps text in this case. The court doesn’t have tunnel vision; a few words that don’t jive with hundreds of thousands of others are unlikely to dismantle the whole law.

The plaintiff’s make a good case, citing – among other solid arguments – the legislative history of the law and the fact that U.S. territories like Guam have Obamacare, sans subsidies. But at the end of the day, these points pale in comparison to the power of congressional intent, and they likely mean less in the eyes of the Court than some may think. Regardless of whether or not the law was written to become unworkable should states decide to not participate, the intent of the law was, and still is, clear.

Ruling in favor of King would conflict directly with the intent of the law, potentially seriously disrupting states’ insurance markets and causing lots of people to lose insurance coverage (although the outcome may not be as disastrous as some make it out to be). And as Justice Kennedy pointed out, such a ruling could open another argument regarding the possibility that linking the exchanges to federal subsidies is coercive – though it seems to me that such an argument may have little merit. Finally, the only thing we can be fairly certain about is that Congress can’t be counted on to clean up the mess if the insurance market does go bust, making handing the law back to it another con for a pro King ruling.

A more interesting question, I think, is how exactly a pro Burwell decision would be executed. The court could hand down a straight forward ruling or some sort of compromise – the possibilities are endless. But another possibility is Chevron deference, meaning the original agency would be left to interpret the law, read: Obama interprets the law as he intended for the next year…but we don’t know who is going to be in the White House in 2016, and how he or she will interpret the statute, which would certainly make the 2016 race even more interesting.

We do know that the Court already bent over backwards to keep the ACA on the books in 2012, when Justice Roberts made an unlikely alliance with the Court’s liberal core. And a ruling in favor of Burwell would require far less a legal stretch than did the 2012 ruling.

Could we shift low value spending to high value spending in Medicare?

This is really the big question behind the notion that value (defined as benefits of care greater than costs for patient) could be increased in the Medicare program while overall costs stay the same or decrease. This may sound too good to be true, and might be, but I think the best chance to pull this off is for Medicare to expand their covered benefits to include long term care. I outline the idea in this Health Affairs post (study referred to in the post).

The comments I have gotten (and that are left on the post) are roughly of two types:

  • great idea, LTC is such an important issue! Most of these folks have personal experience with our LTC system
  • you nut, Medicare is insolvent now and you want to add a new entitlement + unicorns don’t exist

The you are a nut response always has a chance of being correct, and this does sound too good to be true. And I am saying we can shift spending from low value, to high value care; I am not saying lets simply add more benefits on top of everything else. Why I think it might be possible in three simple bullets.

  • we spend so much on medical care that doesn’t help patients (so is low value)
  • LTC is such a tremendous need, and funding more of this will be high value (benefits > costs for patients)
  • If patients have broader options, including flexible LTC, many will choose a path that has less low value care (therefore shifting low value spending to higher value spending)

Pessimism is always warranted. But, this is my story and I’m sticking to it. The best way to increase the value of Medicare spending, is to expand coverage of LTC.

North Carolina ACA Plan/Insurer Update for 2016

The News and Observer has a story on proposed rates for 2016 ACA plan offerings. Some quick thoughts:

  • A fourth insurer is entering the N.C. ACA landscape; Humana will offer plans in the Winston-Salem and Charlotte areas. Blue Cross/Blue Shield is the only one to sell in all 100 counties. Consumers can choose from 4 insurance companies in some counties, and while BCBS NC remains the dominant insurer, the ACA has increased the number of companies selling indy plans in N.C. In some counties there are 4 choices. This wouldn’t have happened without the ACA.
  • The top line headline has been proposed rate increases, with premium changes across the 4 carriers ranging from a 4.3% decrease to a 40.2% increase. For BCBS NC, increases range from 5% to 32%. Premiums rise when actual costs are higher than expected among the insured book of business, and they vary based on age, where you live, whether you smoke. You can only by insurance based on your own characteristics in the county where you live, so the sensible thing to do is to check out how much your premium would be when open enrollment comes.
  • Premiums quoted in the story are actual premiums, and not what consumers pay. The ACA provides income based subsidies to defray the cost of insurance for persons with incomes between 100% and 400% of poverty. And around 90% of North Carolinians who are covered this year got subsidies, so most will be cushioned against these increases (magnitudes are big; avg NC premium ~$400/month, with subsidy ~$315/month).
  • A key issue will be increasing enrollments of younger, healthier customers, which is key for the long term viability of any insurance market. It is likely that expanding Medicaid would help with the risk pool as well, at least by reducing the uninsured (persons in the so called coverage gap–below 100% and above 0% of poverty if a childless adult–are exempt from the mandate penalty).
  • The following statement from the BCBS chief actuary is important (the bold is mine):

The company’s chief actuary, Patrick Getzen, said the ACA continues attracting people who had trouble getting insurance in the past: sicker, older customers who tend to run up medical costs. Getzen said 94.2 percent of Blue Cross’s ACA customers qualified for financial subsidies and nearly a fifth of them discontinued coverage after several months.

“Most of these customers purchased a plan, paid their initial premium, used costly health care services, then dropped their coverage,” Getzen said. “This is an unintended consequence of the way the law is written.”

He says “nearly a fifth” of their enrollees signed up, and later discontinued paying premiums. Then he says “most of these” (presumably those who disenrolled) used “costly health care services” before they disenrolled. Several points here.

  • Disenrollment could occur because a person got a job and therefore insurance, transitioned into Medicare eligibility, or bad reasons, like signing up, getting care and then disenrolling.
  • At least there is some magnitude of the number here, but this points out our need for better data with which to evaluate the functioning of the exchange in North Carolina (Nearly a fifth times “most of these” might be 0.18 x .66 ~.12 which when applied against the BCBS NC sign ups of 400,000 would be 48,000 people).
  • A national estimate of disenrollment is that 1.5 Million on a base of 11.7 Million signed up as of February, 2015 which is 13%.

It would be lots simpler if we just had good information about how many people disenrolled. in what counties? What were their reasons and/or to what insurance status did they disenroll to (uninsured, Medicare, private, etc.)? And how much care did they use prior to disenrollment?

Many important questions. I wish we were collecting better data.


Get every new post delivered to your Inbox.

Join 170 other followers