What discharges alive from hospice tell us

(cross posted at freeforall)

Paula Span recently wrote about hospice as a shield against unwanted medical treatments that provided the space for an elderly Medicare beneficiary to get better. The irony is that the family she writes about had to choose the option associated with imminent death to be able to take their mother home and allow her to receive competent care that her adult children could trust; eventually, she was discharged alive from hospice. Work that I did with colleagues at Duke shows that around 15% of Medicare beneficiaries who begin hospice are discharged alive (85% of Medicare beneficiaries choosing hospice used such care continuously until their death, an average of 50 days later; median 15 days, Group 1 below).

Groups 2 (~10% of patients) and 3 (~5% of patients) were persons that entered hospice and were subsequently discharged alive from hospice; those in group 2 later died while not receiving hospice care, while those in group 3 re-entered and died while receiving hospice care (the study only looked at decedents).

The subset of patients who began receiving hospice but decided to stop doing so while they were still alive likely represent a variety of motivations: prognosis improved; had a bad experience with a hospice provider; wanted to try a therapy that a hospice felt they could not provide because it was not palliative in nature (motivation could run from financial to following the rules to ethical); no longer having a family member that was able to help them receive such care while at home; or wanting a home-based benefit that provided more care than could be obtained via the normal Medicare home health benefits. Some of these motivations could be viewed as strategic in the sense of a patient (and family) seeking care that didn’t fit neatly into the conception of what end of life care should look like as defined by the Medicare hospice benefit for the past 30 years. Our study could not delve deeply into the motivations and outcomes of the different patterns (especially quality of life and burden on family members). More information is needed to fully inform policy. However, this blog post and data such as that shown above are signals that the hospice benefit needs to be reformed and changed with patient needs and preferences at the fore.

What is most demoralizing, is that the hardest part of doing so are the insane politics surrounding death, hospice and palliative care that is surely the worst byproduct of the “debate” about the Affordable Care Act. One of the intangible reasons that we need some sort of deal on health reform to move ahead, regardless of what the SCOTUS says, is so that we can move health policy a bit off the back burner politically (stop laughing!) and allow some actual policy making to take place, focused on questions of outcome and cost, while realizing that we will all one day die–it is only a matter of when, and from what. And what care is available to us all before then.

No one wants to kill your grandmother….but she will die, and so will you and I.

Author: Don Taylor

Don Taylor is an Associate Professor of Public Policy at Duke University, where his teaching and research focuses on health policy, with a focus on Medicare generally, and on hospice and palliative care, specifically. He increasingly works at the intersection of health policy and the federal budget. Past research topics have included health workforce and the economics of smoking. He began blogging in June 2009 and wrote columns on health reform for the Raleigh, (N.C.) News and Observer. He blogged at The Incidental Economist from March 2011 to March 2012. He is the author of a book, Balancing the Budget is a Progressive Priority that will be published by Springer in May 2012.

11 thoughts on “What discharges alive from hospice tell us”

  1. From your blog linked above: “…there is a good deal of care that is non-productive, which I would define as care that does not improve quality of life or extend life. We should start there.”

    Here’s a dilemma, Don. Suppose that, among the many options in late-life care, you identify one that is “non-productive” in 90% of cases. Furthermore, suppose the small minority that respond favorably are not characterized by any particular attribute that makes them identifiable in advance. In certain types of cancer, for example, they are generally non-responsive to any known treatment, but a small percentage of cases seem to be effectively slowed by certain types of chemo, apparently (for lack of Godlike knowledge) at random. My wife’s final months were made less pleasant by her chemo treatments, which did no good. Should I (or her doctor) have suggested that she not opt for the treatment, foregoing the 10% chance it might work?

    The dilemma is, to paraphrase Tip O’Neill, “all medicine is local.” The data are aggregated and analyzed, and the large sample sizes give good estimates of probabilities with +/- confidence limits. But every case is a sample of one, with a probabilistic/unknowable outcome, and every patient wants to be in the success column, even if the a priori chance is small.

      1. Unfortunately, I’m smart enough to frame the question, but not NEAR smart enough to answer it. If I were, you’d have heard of my Nobel Prize.

        1. My brother had a rule that he stated when our grandfather was nearing death. “Will you learn anything with this test? Will what you learn cause you to change your actions? Will the change in actions improve the outcome?” Unless you can get three “yes” answers, don’t do the test. And at the time, my brother worked for a company that manufactured PET scanners, so he actually had a (distant) financial interest in maximizing the amount of expensive tests run.

    1. True dilemma. Very hard and we will make mistakes….or we will spend ourselves to oblivion. In the aggregate we spend too much (this is a value judgement, or an issue of allocative efficiency). Underneath spending too much, we both under spend and over spend (issues of productive efficiency). I think the only practical way to do this is to start disease by disease. Say start with Cong heart failure and say under some circumstances, medicare will not pay for LVAD (just an example). Then define what the criteria for that judgement will be and constantly monitor and tweak. In other cases, you say we will do reference pricing, meaning only pay as much as the clearly effective therapy for the more expensive unproven one.For other things, you say the default is patients get it if they want and doc agrees and so on. We would always be tinkering with this….

      I also don’t mean to minimize the personal nature of what I am blogging about; it is not a theoretical issue for you based on your answer, and I am sorry for that. My initial interest in hospice/end of life came from serving as health care power of attorney for someone who was dying and I was surprised by the answers I got to questions I posed to her doctors: (1) what is the chance this will work? (2) how much does that cost? (3) what will her insurance cover (she was only 45 when she died) and so on.

      Perhaps we (as a nation) don’t have the stomach for this explicit rationing, which to me is preferable from the implicit way we do things now. If so, and and we will decide we really value the last ditch chance….then we need a higher Medicare payroll tax, etc. and/or less spending on something else

    2. Unnecessary or unhelpful treatments sometimes kill or injure patients, you need to consider that as well.

      My mother went through hell when they punctured a lung installing a pacemaker that she didn’t need. A friend of mine, the same age as my mother had both lungs punctured when they tried to install a pacemaker. This was in the same hospital. They never did put in the pacemaker and she’s doing fine without it, two years later.

  2. Your study mentioned earlier work in which some predictors of hospice discharge were identified: being male, unmarried, younger, having no “do not resuscitate [DNR]” order on the chart, or having a noncancer diagnosis.

    Quick question: how do you get into hospice without a DNR order? It seems reasonable to make DNR a requirement of hospice admission; doing CPR seems to violate the whole idea of hospice care. Having it as a possibility falls outside the hospice model, it would seem.

    Not-so-quick question: are there any data on hospice admissions with later discharge and survival for a prolonged period of time (a year or more)? Having only decedents in your study population means that the data would have to come from elsewhere. Probably a rare event, but may happen from time to time.

    When my mom was getting hospice care at home for congestive heart failure, she developed a forearm cellulites one day with pain and swelling. She was discharged from hospice; we took her to the ER, the condition was treated, and she was readmitted to hospice the same day. This was nothing unusual, I was told; the ER visit was for a problem which could be cured, but hospice patients were not eligible for ER admission. Is this a common event?

    So that’s three questions at once. Sorry ’bout that, chief.

    1. From my own experience, a lot of the issues about treatment and DNR and so forth are very local, sometimes to the hospital, sometimes to the doctor. (For example, when my father was dying, the hospital refused to issue a DNR order unless we agreed to forgo antibiotics to treat infection as well.) So it’s possible that the baggage associated with a DNR in some institutions would make it unacceptable to some patients, even though they were otherwise eligible for hospice care.

      1. That is very interesting! It has been many years since I was housestaff, but the chief of medicine emphasized repeatedly that “no cor” (which was ordered on charts when a “cor zero” was not to be ordered in the event of cardiac arrest) did not mean “no care.” The order simply meant that you did everything else (like treating infections, acid-base disturbances, arrhythmias, etc) but did not perform CPR when they died. My mom had a DNR order but got a panoply of interventions for her CHF, which enabled her to leave the hospital and come home and function at least a little bit.

        What do others know about the current meaning of DNR orders? “No antibiotics” sounds like a large departure from what the order meant in the past.

  3. How does this relate to the statistics demonstrating that long hospital stays tend to end with you dying of something other than what you entered with? Is it possible that hospices are somewhat better than hospitals at preventing cross-infection? If only because acute infections aren’t likely to result in you ending up in a hospice?

    It’s not just unwanted treatments that are a danger in hospitals.

    1. To reinforce that point, Brett, consider that the terms “entered hospice” and “discharged from hospice” often relate, not to a physical place called hospice, but to hospice care in one’s own home. My wife’s final month was spent as a “hospice patient” in our home, with frequent visits by hospice nurses, but with day-to-day care by family members. This, in my opinion, was WAY preferable to any alternative I could think of then or now.

Comments are closed.