Rebuilding Choice Architecture for Incurable but Treatable Lung Cancer: Rethinking the Future of Hospice Care in the United States

Rebuilding Choice Architecture for Incurable but Treatable Lung Cancer: Rethinking the Future of Hospice Care in the United States

Supportive & Palliative Care
Dec 11, 2019
By Maggie Salinger, MD, MPP, and Arif Kamal, MD MBA, MHS, FAAHPM, FASCO
Posted: December 11, 2019

IN REFERENCE TO: Mor V, Wagner TH, Levy C, et al. Association of expanded VA hospice care with aggressive care and cost for veterans with advanced lung cancer. JAMA Oncol. 2019 Jun 1;5(6):810-816.

Dr. Maggie Salinger


Hospice is a nationally known interdisciplinary program that provides comfort care for the terminally ill. Yet, despite being lauded for its positive effects on both quality of life and healthcare costs, hospice continues to be an under-utilized resource.1-3 Fewer than half of Medicare beneficiaries leverage hospice’s end of life (EoL) services, and, of those who do, nearly one-third postpone comprehensive palliation until death is less than a week away.3 The issue of low and delayed hospice uptake is linked to unfavorable choice architecture in the lead-up to enrollment, where simplistic eligibility criteria mandate that patients select either disease-directed therapies or comfort-focused ones—a distinction that has become all the more obsolete with the evolution of therapies that are better targeted and better tolerated, particularly in lung cancer. Thus, with an eye toward reduced spending and improved patient autonomy, policy makers have been considering ways to eliminate this false dichotomy between medical and palliative care at the EoL. Much of the innovation in this space has occurred within the Veterans Health Administration (VHA). Since 2009, the VHA’s Comprehensive End of Life Care Initiative (CELCI) has invited veterans to receive hospice care in conjunction with disease-directed therapies. Through a recent study in JAMA Oncology, Mor and colleagues leveraged this large-scale programmatic shift and the fact that its rollout would vary across time and space to perform a quasi-experimental examination of CELCI’s effects.4 The study should be applauded not only for its clever design and promising results, but also for its significant contribution to broader economic, political, and philosophical discussions.

Using a cohort of more than 13,000 patients with newly diagnosed stage IV NSCLC at VHAs across the country (years 2006 through 2012), Mor’s team constructed a modified difference in differences regression to compare patients with high levels of hospice exposure to those with low levels of exposure according to the rate of palliative care consults at each site. The outcomes they examined in the 6 months following diagnosis were per capita costs and patterns of healthcare utilization, including receipt of concurrent and/ or aggressive treatment.4

Their design stands out among other studies on the topic both because of its large sample size and its ability to circumvent the selection bias inherent in a comparison of individuals opting into or out of hospice. These strong suits amplify the significance of their findings, which show that promotion of patient autonomy through palliative integration is not only possible, but also practical.

Mor et al.’s analysis revealed that people in high hospice exposure groups received more concurrent care (e.g., palliative consult plus radiation therapy) and less aggressive care (17.5% vs. 7.4% and 28.3% vs. 35.5% in the highest vs. lowest quintiles, respectively). The quintile with highest exposure was also the least expensive, yielding savings on the order of a couple hundred dollars per day.4

These differences in cost were evident despite the fact that hospital length of stay was similar among quintiles.4 Thus, savings seemed to come from a hospice-associated reduction in people’s demand for aggressive interventions. Framed differently, financial gains were not predicated on restricted access to cancer therapies, nor were they dependent upon allowance of only one therapeutic modality at a time.

Implications for Systemic Change
The implication for policy makers is that we need not marshal patients into silos; we may still be able to provide affordable care when hospice is offered as a complementary service rather than solely as a substitute for active anti-neoplastic treatment. One risk of generalizing these findings, however, is that fewer than 20% of veterans received concurrent care in the highest quintile, and the magnitude of savings associated with hospice exposure diminished over time.4 Therefore, it would be difficult to predict the financial outcomes in a broader Medicare context where concurrent care could someday become more widely available, including in earlier stages of disease.

Further expansion of concurrent care would likely be accompanied by major policy changes, most notable among them being the possibility of Medicare Advantage (MA) swapping its carve-out model for a carve-in operation. This move toward a more integrated approach would be advantageous for a variety of reasons, but so too it might it also have deleterious effects.

As noted above, such policy changes would afford patients and providers greater degrees of autonomy and therapeutic flexibility. Alongside this ethical benefit, we would expect to see an increase in hospice uptake because at least some of its palliative services would become available by default. This would be a significant deviation from the current status-quo, in which patients must opt in to the program, agree to forgo medical treatment, and have a life expectancy of less than 6 months.

Shifting to a carve-in approach would also affect the quantity and quality of EoL services. Importantly, it would no longer be the case that hospice enrollees are transitioned out of MA plans and into traditional Medicare. Because integration would heighten MA’s financial responsibility at the EoL, its contracted payers would fold hospice-related costs into their capitated payment plans. Replacing the current per diem basis of payment would alleviate concerns about the duration of hospice care and instead direct attention to its quality.5 However, the associated need for payers to recalculate risk scores would be complicated by the fact that any change in the choice architecture for hospice enrollment would bring with it a change in the demand for its palliative services, the extent and cost of which would be difficult to estimate.

There is also general uncertainty about which hospice interventions reap the most benefit. As a result, policymakers have no clear guidance for how to go about streamlining a program that, historically, has been so adept at tailoring its care. And although it is true that Mor et al.’s study implies we can have our cake and eat it too, we know that some traditional elements of hospice are actually going to be cost inefficient in the eyes of insurance payers. As an example, carve-in plans are unlikely to offer year-long bereavement support to family members of decedents.

The potential for mismatch between the socially determined value of an initiative and its economically calculated worth extends far beyond the topic of hospice redesign. But because death is a destiny we all share and an event of great import to all cultures, perhaps this particular policy issue could bring unity to a divided Congress and serve as a roadmap for navigating other philosophical complexities in the healthcare reformation of the United States. ✦

About the Authors: Dr. Salinger is an internal medicine doctor and resident at Duke University Hospital, Internal Medicine Residency Program. Dr. Kamal is an associate professor of Medicine and an associate professor in Population Health Sciences at Duke Cancer Institute.

1. Kelley AS, Deb P, Du Q, Aldridge Carlson MD, Morrison RS. Hospice enrollment saves money for Medicare and improves care quality across a number of different lengths-of-stay. Health Aff (Millwood). 2013;32(3):552-561.

2. Wright AA, Keating NL, Balboni TA, Matulonis UA, Block SD, Prigerson HG. Place of death: correlations with quality of life of patients with cancer and predictors of bereaved caregivers’ mental health. J Clin Oncol. 2010;28(29):4457-4464.

3. Report to the Congress: Medicare Payment Policy. Medpac. Published March 2018. Accessed August 31, 2019.

4. Mor V, Wagner TH, Levy C, et al. Association of expanded VA hospice care with aggressive care and cost for veterans with advanced lung cancer. JAMA Oncol. 2019 Jun 1;5(6):810-816.

5. Stevenson DG, Huskamp HA. Integrating Care at the End of Life: Should Medicare Advantage Include Hospice? JAMA. 2014;311(15):1493-1494.