I have been thinking a lot about the FY 2016  Hospice Wage Index proposed rule (NPRM).  The more I have thought about it, however, the more complex it seems. I have needed to harken back to my salami methodology (described here) and try to break it down into manageable chunks. The only chunks I am focusing on are the ones having to do with payment reform (tiered reimbursement and the Service Intensity Add-On) and the “clarification” to include all diagnoses on claim forms (and the related – excuse the pun – virtually all / related and not related / prognosis vs diagnosis conundrums).

Another, perhaps more apt metaphor comes to mind with these salami chunks – seeing the forest through the trees.  In this case, the forest is the Medicare Hospice Benefit and the trees are all the details in the NPRM – the first 60 days more money / after 60 days less money, the count of days, the Service Intensity Add-On (SIA) with only skilled visits, no SNF patients, all diagnoses on the claim forms including those related to mental health, etc., etc.  Yes, with all these “trees” we can categorize, organize, recite the new requirements, enumerate the potential or real problems with each in order to assemble a comment letter to CMS by June 29th. But in so doing, it is easy to lose sight of the forest, the bigger picture – the Medicare Hospice Benefit.

The Medicare Hospice Benefit (MHB)

In the NPRM, as in all other rules written by CMS, there is a lengthy preamble discussion of the history of the MHB.  In addition, throughout the NPRM, CMS references the MHB as established in statute, with implementing regulations finalized in December 1983.

And therein lies the problem – 1983 – 32 years ago.  The MHB was developed in a world very different from the one we live in today and the one in which hospices provide end of life care. Think about it. Where were you and what were you doing in 1983? For the youngsters among us – were you even born or out of toddlerhood in 1983? There were no VCRs much less DVRs, laptops, tablets, cell phones, electronic health records, social media, email, streaming video, online education – it was such a different world back then. And it was a different world for hospices – the majority of hospice patients had a cancer diagnosis, their terminal illness followed a common trajectory, establishing a 6 month prognosis was more straightforward and the related / not related issue was not a big deal.

The MHB was wonderful back in 1983 and a testament to its “wonderfulness” is how long it has lasted.  But it needs to change and the first “incremental steps” toward payment reform of the tiered reimbursement with the SIA does not cut it. It is akin to putting a band-aid on something that requires major statutory surgery. This does not mean that the payment reform proposed in the NPRM is not good or should not happen, but it is not going to solve the problems CMS is hoping it will (i.e. long lengths of stay, frequency of skilled visits, hospice care in nursing homes, etc., etc.) It may help with the current misalignment of payment with hospice service intensity and give the long talked about U-shaped curve a fair trial – I suppose that remains to be seen. But it does not address some of the fundamental problems with the MHB that are threatening the existence of many hospices, especially the smaller ones.

I have described elsewhere some of the issues with the MHB. This NPRM, when looking through the trees, further increases the urgency I feel to look at the forest and the need to advocate for fundamental change in the MHB. Let me try to explain by discussing some of the trees.

Payment Reform

The NPRM proposes the first substantive payment reform since the inception of the MHB with a tiered reimbursement system for routine home care and a SIA.  While I do not want to minimize or diminish the extensive effort, research, and resources that have gone into CMS’ efforts at payment reform, I am concerned that many of their findings seem based on outlier hospices (of the “proceed until apprehended” variety) and their solutions are not focused on these outliers but addressed to the whole industry. CMS has the data necessary to identify aberrant providers – why don’t they scrutinize the hell out of them, penalize them and put them out of business?

I also have concerns about the data upon which CMS is basing its proposals.  Are the appalling statistics related to the lack of visits at the end of life only based on skilled RN and SW visits? If so, is it possible the stated low frequency of visits is misleading because it is not taking into account perhaps a whole vigil team of volunteers at the patient’s bedside or perhaps a chaplain was there with the patient and family offering comfort and support, or an LPN or hospice aide who are often the most present and important members of the team? Maybe there were NO visits and that truly is appalling. But CMS can’t have it both ways – affirming and emphasizing the interdisciplinary nature of hospice but only giving credence to visits by RNs and social workers.

Also, somewhat related to this, the SIA was designed in a manner taking into account “operational concerns and programmatic complexity led us to explore the concept of a SIA that could be implemented with minimal systems changes that limit reprocessing of hospice claims due to sequential billing requirements.” p. 25857

Shouldn’t something as important as the first substantive step toward hospice payment reform be designed to be the best it can be with concomitant system changes as needed?   And while the operational concerns and programmatic complexity of CMS were taken into consideration, what about that for hospices?  Certainly the first significant change to hospice reimbursement in 32 years would seem to warrant a demonstration project or, at the very least, a trial run to assure CMS’ systems can in fact manage the changes, assist hospices with figuring out managing their account receivables, billing, and avert potential unintended negative consequences.

“Clarification” of all diagnoses on claims

Now on to my favorite trees – the “virtually all,” related/not related, and prognosis vs diagnosis, which I believe are all more or less under the rubric of the clarification of all diagnoses on claims in the NPRM.

For fun, let’s parse each one of these trees separately within the NPRM and hope that by the end of it, we have a better view of the forest – the unwieldy MHB.

“Virtually all”

Here is the non-statutory, non-regulatory language that first appeared in the preambles to the proposed and final original implementing regulations of the MHB in 1983 (let me repeat, 32 years ago):

“It is our general view that hospices are required to provide virtually all the care that is needed by terminally ill patients. …Therefore, unless there is clear evidence that a condition is unrelated to the terminal prognosis; all conditions are considered to be related to the terminal illness.” 

In the NPRM, this “tree” is referred to a number of times. Let’s look at two of the most interesting:

 “We continue to expect hospices to adhere to the long-standing policy to provide ‘‘virtually all’’ care during a hospice election as articulated in the 1983 Hospice Care proposed and final rules as well as most recently in FY 2015 Hospice Wage Index and Payment Rate Update final rule.” p. 25847

What is interesting here is that I think suddenly, the “virtually all” language has been promoted to “long-standing policy.”  I am not a lawyer but I somehow suspect that preamble language does not quite rise to the level of policy.

Then there is this:

 “As stated in the 1983 hospice final rule, and reiterated throughout hospice rules since implementation of the benefit, it is our general view that the waiver required by the law is a broad one and that hospices are required to provide virtually all the care that is needed by terminally ill patients (48 FR 56010). p.25877

It is just not true that this “virtually all” language has been reiterated for 32 years. Obviously I am not anal enough (maybe that is not obvious) to look through all hospice rules “since the implementation of the benefit” but I did look at a few prior to 2013 and could not find this language.  Given that the final rule of the “new CoPs” published in 2008 is arguably the most significant rule published since the implementation of the benefit, I did look at that and did not find any “virtually all” all language.  I find it curious that CMS has been trying to pass this “virtually all” language off as “longstanding policy” that has been reiterated forever when in fact it only reappeared (with a vengeance) in relationship to Part D a few years ago.

Enough said about “virtually all” – let’s look at the closely related, related/not related issue.

Related/Not Related

When CMS is promoting the “virtually all” concept, as it has been for the past few years, the issue of what is or is not related to the terminal diagnosis (and prognosis – more on that next) is critical.  Here is the crux of the issue with related/not related:

… Therefore, unless there is clear evidence that a condition is unrelated to the terminal prognosis; all conditions are considered to be related to the terminal illness. It is also the responsibility of the hospice physician to document why a patient’s medical needs will be unrelated to the terminal prognosis. p. 25836

NHPCO and others have tried valiantly to maintain the critical role of the physician in determining on a case-by-case basis what is related or not related to the terminal diagnosis/prognosis.  But here is the kicker from page 25878 of the NPRM within the context of CMS’ concern regarding hospices unbundling services:

We have received anecdotal reports from non-hospice providers who have rendered care and services to hospice beneficiaries in which the non-hospice provider states that the care given was related to the terminal prognosis of the individual. These reports go on to say that they have contacted hospices to coordinate the care of the hospice beneficiary only to be told by those hospices that they disagreed with the non-hospice providers’ clinical judgment that the care was related to the terminal prognosis. … These non-hospice providers stated that they disagreed with this practice, and considered it fraudulent. As such, they were unable to be reimbursed by the hospice or by Medicare for services provided that they felt were the responsibility of hospice. p. 25878

No reference to the hospice physician’s clinical judgment here.  What is critical in this is CMS’ seeming willingness to take the “non-hospice provider’s clinical judgment” regarding relatedness over the hospice’s.  This is only one of many potential examples of the slippery slope of related/not related.  It becomes even more of a quagmire when trying to determine relatedness to diagnosis, prognosis or both.

Diagnosis vs Prognosis

I wrote a whole blog post on this issue here and the topic just seems to get murkier. Within the context of the NPRM it remains confusing.  One might well ask if it is a semantics issue.  The dictionary defines prognosis as:

1.Medicine/Medical. a forecasting of the probable course and outcome of a disease, especially of the chances of recovery.  2. a forecast or prognostication.

In hospice, eligibility has been determined by whether the prognosis for, or outcome of, a patient’s terminal illness is death within 6 months if the illness runs its normal course.  What has become confusing (and critical) lately is CMS’ definition of prognosis, which is reiterated in the NPRM that hospices must consider what factors influence that terminal prognosis, or, as CMS states:

….what factors influence the terminal prognosis of the individual (that is, what conditions render the individual terminally ill and which conditions are related). p. 25880

Let’s look at a ridiculous (maybe) example.  For the rest of my life I am going to have a diagnosis of S/P TKR  (status post Total Knee Replacement).  Let’s imagine that someday in the very distant future I am on hospice with a primary diagnosis of bone cancer. That is the condition rendering me terminally ill. While I am on hospice the cancer affects the knee joint to such a degree that I am writhing in pain and need revision surgery to replace or resettle the knee prosthesis in order to palliate my pain and enhance the quality of my life for my remaining days.  Since the knee problem was potentially a result of the bone cancer and would contribute to a shortened prognosis if not addressed, would something that occurred long ago in the past (the TKR) now be considered related and the hospice be responsible for paying for the knee surgery and inpatient stay?  Let’s complicate it more – change my terminal diagnosis to cardiac disease but my knee wore out because the prosthesis was old and in need of replacement. I was in massive pain – would the hospice be on the hook then because my pain would be influencing my prognosis and need management?

We can pick that example apart but the point is that the all-inclusive prognosis with virtually all care of hospice patients being related to the terminal prognosis and therefore the financial responsibility of the hospice is unreasonable and unsustainable. This is particularly true for small hospices and, quite frankly, it is also the reason “unbundling” occurs.

Clearly, I could go on and on and I have. Let’s try to wrap this up with the clarification of all diagnoses on claim forms.

Clarification of all diagnoses on claim forms

A requirement of the NPRM is that hospices must now list all of a patient’s diagnoses on hospice claim forms.  CMS states that they are clarifying existing policy but I think there has been a lot of confusion regarding this issue.  If, as CMS states, over 77 percent of hospice claims reported only one diagnosis in 2010, and this is supposedly a bad thing, there is obviously a disconnect.  I refuse to believe that 77% of hospices would wantonly disregard a CMS directive and only self- correct to 49% in 2014.  CMS stated that in the 2013 Hospice Wage Index they said hospices should report all related (not unrelated or coexisting) diagnoses on claim forms even though coding guidelines required the reporting of all diagnoses that affect patient assessment and planning (p. 25880).  Hence the confusion. It is even confusing figuring out how confusing it is!

So the clarification is now to report all diagnoses, related and not related.  CMS states that the purpose of this is to assure hospices are doing adequate comprehensive assessments and to determine costs.  My bet is that the data will eventually be used to determine if hospices have adequately assumed financial responsibility for EVERYTHING wrong with the patient.  I would imagine a lot of “non-hospice” scrutinizers are going to lend their opinions on the relatedness issues.

There is also a new an emphasis in the NPRM that hospices assess and report if a patient has a mental illness, in addition to his or her principal terminal diagnosis. Is the assumption here that hospices have or should acquire expertise in managing mental illnesses and assume financial responsibility for psychiatric medications or inpatient psychiatric hospitalizations and/or clearly document how the mental illness does or does not contribute to or influence the terminal prognosis?

Do you see how things are getting / have gotten out of hand?


So enough.  I believe that the “trees” discussed above are the proverbial straws that are breaking the MHB’s back.  At the risk of being totally annoying with my forest/trees metaphor – let me conclude by saying that we can and should send comment letters to CMS in an attempt to put out some of the fires in some of the trees in the NPRM, but please, please, with that, let us also try to figure out how to salvage the forest because it is burning.

Posted by Heather P. Wilson, PhD, CHC  CEO, Weatherbee Resources, Inc.