Decoding Health Care

Part 20: End of life care

During my time at Aetna, I was frequently invited to participate in oncology conferences as the voice of the payer. It helped that I had practiced oncology for many years before working for Aetna. At one conference, the major topic of discussion was quality measurement in oncology. I opined that quality measures that focused on end of life care such as chemotherapy in the last 7-14 days of life and hospice enrollment were good measures. I was approached by the leader of a major patient advocacy organization after the discussion and they told me I had it all wrong; patients don’t consider an oncologist who makes it easier for them to die being “high quality”; patients consider a “high quality” oncologist one who can help them live. These two skills are not mutually exclusive.

Most patients with metastatic cancer ultimately die of their disease, so being good at end of life care is critical. The fact that metastatic cancer is still fatal might not be obvious given all the hyperbole in the popular press about advances in cancer. Every year, the American Cancer Society publishes a “report card” devoted to cancer statistics. And for several years now, the survival of Americans with cancer has steadily improved. The percent of Americans dying from cancer has dropped almost 30% since 1991. This is mostly attributed to the dramatic reduction in cigarette smoking, but early diagnosis (screening) and treatment have played a part.

From a population perspective, we are doing a lot better with cancer. But what about if you as an individual get cancer?  About 70% of patients diagnosed with cancer survive, although the survival rates vary a lot from cancer to cancer. This survival rate is largely based on whether the cancer is diagnosed at an early stage. Unfortunately, the cure rate for advanced (metastatic cancer) hasn’t improved nearly as much. Metastatic lung cancer, pancreatic cancer, and ovarian cancer are all fatal, although modern treatments have prolonged survival (particularly in lung cancer). As long as advanced cancer isn’t curable we need to think about end of life care.

In general, patients with advanced cancer derive the greatest benefit from the first treatment they receive. As you cycle through treatments, and the cancer becomes more and more resistant, the benefits decrease. In addition, the toxicities of treatment increase as does the symptom burden associated with the progression of the cancer. At some point, the small likelihood of benefit and the high likelihood of toxicity makes the patient and the doctor question the wisdom of continuing aggressive treatment. This is usually when the topic of hospice is introduced.

Hospice is not a new idea. In fact, the idea of hospice dates back over 1000 years to the times of the Crusades. Hospices were shelters along the Crusade route where crusaders who had become ill or were dying could be made comfortable. These shelters were faith based, staffed by religious orders. They reached their zenith during the Middle Ages but faded as the influence of the religious orders waned. Subsequently, other hospices (both faith based and secular) provided terminal care for diseases like tuberculosis and cancer in Europe and Australia.

The modern era of hospice care can be traced to Dame Cicely Saunders in England in the 1960’s. Saunders, who was a nurse and social worker (and subsequently a physician) focused on the palliation of symptoms in patients with terminal illness. She was interested in pain control. But she was also interested in the psychosocial and spiritual needs of her patients and their families. Many of the guiding principles of modern hospice care can be traced to her work.

Dame Saunders spread the word about hospice internationally. She spoke in the US and found her approach had broad appeal. This was accelerated when Elisabeth Kubler-Ross, a Swiss psychiatrist, published her famous book, “On Death and Dying”. This book made it okay to talk about death and dying. It described the emotional and psychological stages people experience in the process of dying, along with the coping mechanisms they employ. Dr Kubler-Ross had a profound impact on our approach to end of life care.

A major turning point for hospice care in the US came in 1982 when Congress created the Medicare Hospice Benefit. In 1993, President Bill Clinton made hospice a guaranteed health benefit. Suddenly there was a way to pay for hospice. But like everything else in American medicine, the rules for payment are a little complicated. And Medicare establishes these rules, since 90% of the patients receiving hospice care are Medicare beneficiaries (https://www.medicare.gov/coverage/hospice-care).

Hospice is part of the Part A benefit and all Medicare beneficiaries are eligible without copayment. Medicare pays the hospice a “daily rate” irrespective of what services are provided. In order to be eligible, the beneficiaries must have an expected life expectancy of six months or less. Hospice will provide some home nursing services, home health aides, medical supplies and prescriptions needed for symptom management (like pain medicine). It will also cover hospitalizations for symptom management as well as respite care. It will not pay for anti-cancer treatments or care delivered by physicians or facilities not related to the hospice diagnosis.

Although this might seem pretty straightforward, I can tell you from personal experience as an oncologist that the hospice provider has a lot of control with regards to specific services. For example, they usually will cover blood pressure meds and cholesterol lowering meds (patients on hospice usually don’t want to stop them). Hospices almost never cover blood transfusions. This seeming arbitrariness drives patients and their families nuts, as does the fact that essentially every medical decision needs to go through hospice. Some hospices are less rigid than others and recognize that the patient often has a long standing relationship with their oncologist. Other hospices feel the need to review every single decision made by the oncologist including changes in pain medications. Doctors and patients do not appreciate this conflict.

One common patient and family misconception is that hospice provides round the clock care. It absolutely does not. Over 80% of hospice patients receive the bulk of their care at home with an identified family or friend care giver. Although they are more common than in the past, dedicated in-patient hospices are not the site of care for most patients dying of cancer.

More than half of all cancer patients use hospice services, although the average length of stay in hospice is less than two weeks. Hospice utilization varies widely based on the cancer diagnosis. And utilization is significantly less in minority groups. Almost a third of cancer patients who sign up for hospice are on hospice three days or less. There is good evidence that hospice patients who do not have cancer are on program a lot longer (close to two months in fact). Why is this the case?

A recent Art of Oncology piece in the Journal of Clinical Oncology provides insight (https://ascopubs.org/doi/full/10.1200/JCO.24.00396). In this piece, the author, a PhD researcher in palliative care describes her personal experience with her father and hospice. She describes how her father, who had an advanced malignancy, suffered a rapid decline in his condition when his treatment was interrupted due to intercurrent illness. Her father’s strong desire was to get strong enough to resume treatment. Hospice seems never to have been considered except as a bridge to make him stronger in order to resume treatment. His hospice length of stay was three days. The author argues that these three days were very beneficial, and that the focus on length of time on hospice is misplaced as it doesn’t consider the needs of the individual patient. Perhaps.

Was this patient’s death a total surprise? That is a hard question to answer without knowing the clinical details, but my experience tells me it wasn’t. When cancers progress rapidly due to an unplanned break in therapy it is usually a really bad sign. So how can this sort of situation be managed better?

Bad outcomes (including death) in advanced cancer are the rule rather than the exception. Oncologists and their patients rejoice when things go well, but oncologists do their patients a disservice when they don’t take the long view. Studies have suggested that a simple predictor of a patient’s life expectancy is the “surprise question” posed to the treating physician: “Would you be surprised if this patient passed away in the next 12 months?”  Although the performance characteristics aren’t perfect, this question actually works pretty good in cancer patients. And most importantly, it should trigger certain actions by the care team: Does this patient have an advanced directive? Will they benefit from palliative care?

There has been a long running debate about whether discussing prognosis robs patients of hope, makes them depressed or shortens their lives. Many studies suggest the answer is no (https://ascopubs.org/doi/10.1200/JCO.2012.42.4564). In reality, the answer is probably not but that each patient is different. The medical cognoscenti have decided that these end of life discussions are a mandatory component of good care, and I tend to agree. But sometimes it can be really hard. How prognosis is discussed is very important.

One way that we as oncologists have attempted to deliver better end of life care is by earlier introduction of palliative care. The definition of palliative care is so broad and imprecise that most oncologists feel they do it for every patient every day. When I speak of palliative care in the end of life setting I am not talking about managing hot flashes in women on hormonal therapy or peripheral neuropathy in patients who received paclitaxel or oxaliplatin (although these certainly are palliative care in the broadest sense). When I speak of palliative care, I am referring to discussing goals of therapy, advanced care directives, and symptoms related to advanced malignancy or its treatment (like pain).

The value of palliative care in this setting is well established. It was brought into focus by Dr. Jennifer Temel at the Massachusetts General Hospital in her lung cancer clinic (https://www.nejm.org/doi/full/10.1056/NEJMoa1000678). She showed that patients receiving early palliative care had better quality of life, less depression, less aggressive end of life care, and actually lived longer. This ground breaking study has led to the widespread implementation of palliative care services in oncology. Although the precise components are not well defined, there is a certain consistency in the approach taken by providers who offer this intervention in both academic as well as community practices.

Is palliative care just a euphemism for hospice? I am not sure it really matters. Clearly palliative care providers are very good at what they do. The trick is getting them involved early. The hospice discussion gets really tough when the patient isn’t doing well. It becomes a much less loaded discussion if it is called palliative care and occurs well before the often inevitable downhill slide. The real problem is that the reimbursement for these services isn’t very good. A lot of practices hire palliative care providers knowing that at best they will break even financially. But they make everyone’s lives better, including patients.

One huge advantage palliative care has over hospice is the ability to continue active treatment while discussing goals of care and addressing symptoms. Hospice has been absolutely steadfast in its refusal to consider any anti-cancer therapy while on hospice. This is certainly due to the cost of active treatment. Given that the average daily reimbursement for hospice is about $200 a day,  there is simply no way to afford drugs that cost $15K a dose. There have been discussions of doing a pilot in which hospice is reimbursed for the cost of drugs but that has not materialized.

There has been renewed interest in end of life care with the development of alternative payment models. Many of the quality measures in these programs focus on end of life. Why would payers care?

Caring for patients at the end of life can be very expensive. For a long time the focus was on chemotherapy in the last 14 days of life but this actually doesn’t occur very often (probably fewer than 10% of patients get chemotherapy in this time frame). But hospital care is extremely common, and more importantly ICU stays are not uncommon. About 30% of patients spend a day in the ICU in the last month of life. Almost 30% of people die in the hospital. This is in spite of the fact that the vast majority of American want to die at home.

It is a huge mistake to underestimate how hard dying at home can be. An excellent piece by Sunita Puri in the WSJ describes the challenges very clearly (https://www.wsj.com/health/healthcare/more-people-are-dying-at-home-is-that-a-good-thing-79bd964e). Dying at home isn’t always appropriate. But as health care providers taking care of patients with terminal illness we must do our best to help our patients at the end of their lives. I am not certain what a good death is, but I have seen my share of bad ones.

When I was in practice, the hospice where I referred most of my patients used to tease me that I was their number one referring physician (hinting that might not be something I would necessarily brag about). But I have personal experience with hospice.

My father had end stage heart disease and was doing poorly at home. His care giver was overwhelmed and I lived 1400 miles away. I suggested he consider hospice care. I carefully described what hospice offered, and he quickly agreed. Neither my father nor I was burdened with any prejudices about what it meant to be on hospice. I made the referral after discussing it with my father’s primary care physician. The relief expressed by his doctor was palpable. This is a hard discussion for every physician, not just for oncologists.

My father thrived on hospice. His CHF got much better when his medical regimen was simplified, he was given a hospital bed (that allowed him to elevate his legs), and supplied with oxygen. Nursing supervision that optimized compliance didn’t hurt. And he loved the health aides and nurses. After a couple of months he was doing so well he graduated from hospice (which actually happens to 10-15% of people who go on hospice). He did well for over a year (not hospitalized once) until he fell and broke his hip shortly after which he passed away. I am a personal fan of hospice.

Before I close, I want to tip my cap to President Jimmy Carter. Every week some celebrity boasts that they are going to “beat cancer”. Some do, some do not, but it has very little to do with their willingness to fight. President Carter has shown everyone how recognizing one’s mortality, enjoying each day to its fullest, and using hospice to its greatest potential can help you make the most of your life when the end is near.

In my next post I will shift gears and discuss innovation in health care and how we pay for it. We will start with the role of Venture Capital and Private Equity, and then discuss health care start-ups.