How Do We Identify Who Needs Palliative Care Most?
By Larry Beresford
Anyone with a serious illness can benefit from palliative care interventions, according to palliative care advocates. But because palliative care resources are finite, the question becomes, who should receive those limited services? What do we really know about the need for palliative care across medical conditions, across groups, and across populations—or, for that matter, which patients could benefit the most from this service? Can we devise more nuanced models, definitions, and criteria for deciding whom our services optimally should try to reach?
If we could define serious illness in ways that ensure palliative care support for those who need it when they need it most, that might help us identify which patients to screen for palliative needs and how to triage for specialty-level palliative care. As our healthcare system rapidly evolves toward new value-based and population health models, how we identify this population has significant implications for the present and future delivery of palliative care. We know that professional staff and other resources in palliative care are finite, and we also know there are misconceptions about palliative care and when it should be used—despite best efforts by advocates to address those misconceptions.
There is no single best way for identifying those who might most benefit from palliative care in all circumstances, but leaders in the field and within AAHPM are engaging with the challenge on a number of fronts with the aim of defining better ways to target, deploy, and utilize this finite palliative care workforce for maximum benefit to patients and families when they need it most.
AAHPM has advocated for advancing basic or primary palliative care, a model of training and support for frontline providers like primary care physicians and nurse practitioners to enable them to address such palliative needs for their patients through tactics like effective pain management and facilitation of conversations about the patients' goals of care. In this model, specialist interprofessional palliative care teams, with their advanced skills, would be reserved for more difficult and complex cases.
But the challenge remains: how do we identify those who could benefit the most and then reduce the barriers that stand in the way of accessing this service? "How do we find the patients who should get palliative care?" asked Katherine Ast, MSW LCSW, AAHPM's director of quality and research. "This question comes up at the patient care level, but also at the level of health systems, health plans, and other groups responsible for managing the health of defined populations."
What Do We Mean by Serious Illness?
Serious illness has been defined as "a health condition that carries a high risk of mortality and either negatively impacts a person's daily function or quality of life or excessively strains their caregivers."1 That is a definition rooted in the patient's experience, but is there a method to identify patients with serious illness using algorithms programmed into the electronic health record (EHR) and other data sources?
"If you could prospectively identify a seriously ill patient population, how would you use that information to improve care at a population level? Who would you connect with basic or with specialized palliative care? What we are talking about, ultimately, are measures of care quality," Ast said. So part of the challenge is to advance care quality by advancing the science of quality measurement in palliative medicine.
AAHPM has a 3-year cooperative agreement (#1V1CMS331639) from the Centers for Medicare & Medicaid Services (CMS) to develop quality measures for community-based palliative care that are intended for use in the government's value-based quality initiatives. It also has fostered the creation of the Palliative Care Quality Collaborative with support from the Gordon and Betty Moore Foundation to develop a unified registry to capture both program- and patient-level quality data for patients with serious illness, including those receiving palliative care, and has partnered with the Hospice and Palliative Nurses Association on "Measuring What Matters," an initiative that highlights 10 measures that matter to our field. (See aahpm.org/quality/measuring-what-matters for more information.)
AAHPM also is collaborating with OptumLabs® and AMGA on an AARP-funded serious illness denominator project utilizing the OptumLabs Data Warehouse to examine how big healthcare databases might be used to help identify the patients who need palliative care. This project aims to use data from EHRs and healthcare claims to find populations of patients who are seriously ill but have not been identified as such so they can be offered palliative care services and enable clinicians to capture their data in palliative care quality measurement.
That has involved convening subject matter experts, engaging stakeholders, and then testing and refining the denominator of appropriate patients. "We have explored the use of diagnostic codes, utilization codes, and durable medical equipment orders to get at the type and severity of illness. We have looked to define and identify patients with multiple chronic conditions in these databases," Ast said. "We'd like to leverage the information that exists in the EHR, such as patient falls, caregiver stress, or other nuggets that might be found in provider notes, by using natural language processing."
This technology is still a ways off, but researchers like Charlotta Lindvall, MD PhD, a palliative care physician at Dana-Farber Cancer Institute in Boston, MA, already are exploring machine learning to identify triggers for a palliative care consultation. (See the fall 2019 issue of AAHPM Quarterly for more information.)
As the researchers dig deeper into the serious illness denominator, it's clear that not all of the needed data can be found in existing databases of health plans and health systems. And that's just one piece of the puzzle, Ast said. "Right now, our work is purely a research initiative. We're not ready yet to identify actual triggers for palliative care or to identify the patients who should be targeted. We still don't have perfect equations for capturing patients who would benefit most and respond the best. Multiple denominators will be needed for various purposes and unique population groups. But AAHPM keeps trying to push the envelope."
The Importance of Function
Another way to frame these issues is around maximizing efficient use of resources, said Amy Kelley, MD, associate professor of geriatrics and palliative medicine at Mount Sinai School of Medicine in New York City. "As a field, we have the broadest concern that people with serious illness deserve high-quality palliative care." A condition could have a high mortality risk, but it also may impact activities of daily living and quality of life, she said. "It's not enough to know the diagnosis. When you add information about functional limitations, or the fact that the patient's husband has dementia, you have a better basis to make these determinations. That's the reality of our patients—and how we should be trying to identify who should be on the list to receive a palliative care referral."
Kelley said there is not a single denominator for palliative care need, either primary or secondary. "I believe the healthcare system needs to be measuring functional status in some systematic fashion and systematically recording it. Functional status is a critical indicator of well-being. It gets at disease severity, but also at the impact of the disease on the person's life. I'm working with the Center to Advance Palliative Care on a collaborative initiative for the National Quality Forum advocating for a national strategy to make palliative care more accessible to consumers. One of our goals is to get functional status routinely measured. At a conceptual level, function provides another fundamental layer to understanding serious illness."
Christine Ritchie, MD MSPH FACP FAAHPM, director of research in the division of palliative care and geriatric medicine at Harvard Medical School, Boston, MA, agrees that functional limitation is an important part of the conversation because it has been implicated in both prognosis and outcomes—and there are validated measures for it. "Our work is [focused on] showing how functional status interacts with underlying chronic illness. It is a critical indicator of well-being and of prognosis," she said. "But the EHR doesn't tell us all the information we need in a consistent, constructive form. How will you decide who needs palliative care if you don't have clear definitions? How are you going to know if high-quality care was given to patients?" According to Ritchie, these issues are increasing in importance for policy makers and for payers.
Who Has Access to Palliative Care?
Another crucial part of this conversation for leaders in the field involves barriers to palliative care use. These include racial/cultural differences and the barriers to access that disproportionally impact some patient groups. Misunderstandings and misinformation about palliative care, some of which may be exacerbated by these cultural differences, include a belief that palliative care is intended for the end of life only. Other barriers include lack of insurance coverage; geographical barriers, particularly between rural versus urban settings or by region; and financial disparities and other social determinants of health.
"As a field, we have moved from emphasizing equality of access to equity, with a focus on responding to actual patient needs," said Jane Loitman, MD MBA MS FAAHPM, a private practice palliative care physician in St. Louis, MO. Equity means those who need the service most are most likely to receive it—not that everyone has equal access to it. "We just have to keep doing what we can do and recognize that not everyone needs palliative care. We keep trying to define our niche, and for most of us, like it or not, that niche has turned out to be potentially life-limiting illnesses," Loitman said.
"Who needs palliative care today and isn't getting it?" asked Joe Rotella, MD MBA HMDC FAAHPM, AAHPM's chief medical officer. "Where are the gaps? Can you identify a population with serious illness and [that is] likely to have palliative needs and make that your denominator for quality measures, or use it to power a predictive modeling algorithm? How can we use tools at the system level to make sure the best care is delivered at a personal level?"
To answer these questions, Rotella said, the field needs to ask what we mean by benefitting the most from palliative care. "Is it defined by the individual? Relative to what goal? Is it a better patient experience or better clinical outcomes based on measures of quality of life? Or is the outcome defined by payers and their quest for value—trying to make sure that they are not paying for unnecessary and unwanted medical interventions across covered populations?" he asked. The criteria thus may differ depending on whether they reflect the perspective of patients, populations, or payers.
Rotella urges caution in focusing on limits in palliative care capacity. "As a field, palliative care is still in the mode of raising awareness about the value of our services. As demand for palliative care grows, we could develop better training, service, and payment models to meet the needs and trust that the capacity could follow," he explained.
"If we could do a better job of identifying who benefits most from palliative care and what that actual benefit is, would it be possible to attract more investment in palliative care infrastructure?" Rotella asked. "The real excitement around palliative care denominator research is what it offers for value-based coverage. We would like to find these people even before their first disastrous hospitalization."
Other Perspectives on Palliative Care Need
AAHPM members and other leaders in the field offer additional perspectives on how to bring clarity to these questions. Somewhere in the interaction of identified need, access, barriers to service, and enhanced understanding of what palliative care can offer patients lie some important insights that could help advance the field.
Steven Baumrucker, MD FAAFP FAAHPM, system medical director for palliative medicine in the Ballad Health System of Upper East Tennessee, points out that CMS's Oncology Care Model (OCM), a pilot program designed to improve the quality of cancer care at the same or lower cost through a voluntary managed oncology approach, also mandates that every patient who starts palliative chemotherapy at participating oncology practices must be provided access to palliative care services.
OCM is a 5-year model from 2016 to 2021 to test innovative payment strategies, including real-time monthly payment models; advanced service models; and retrospective, performance-based payment based on quality and shared savings. "Our palliative care service grew from a half-day to 4 days per week after the system decided to comply with this rule through consultation on day one of chemotherapy," Baumrucker said. The oncologist can simply click a button on the EHR to trigger the referral.
Marie Bakitas, DNSc CRNP, professor of nursing at the University of Alabama at Birmingham, has been involved in several research initiatives offering further insights on how to understand these questions. "We have learned a lot over the years about who does and does not need palliative care," she said.
For example, Project ENABLE (Educate, Nurture, Advise, Before Life's End) tried to include every patient who was newly diagnosed with advanced cancer and found that those with high depression scores responded better—with greater improvements in quality of life and on survival outcomes—than those who did not.
Subsequent research by Bakitas studied palliative care for patients with cardiovascular disease. "We looked at 415 patients with heart disease, half of them African American. We found that, unlike other studies, we did not achieve dramatic improvements in their quality of life using a heart failure–specific measure," she said. Perhaps the study population was already fully optimized from a cardiac perspective and not sick enough to realize measurable improvements on that scale. Or, according to Bakitas, perhaps there are other, unmeasured aspects of care where palliative care makes a bigger difference.
"One possible conclusion is that rather than offering palliative care services, for example, to every newly diagnosed patient with advanced cancer, perhaps we should instead target those who we believe would benefit the most, such as those with depression or those who have some threshold of symptoms," she observed.
"We were all encouraged with results showing that early palliative care benefits patients, such as ENABLE and Dr. Jennifer Temel's landmark study.2 But at the end of the day, if we say, as the American Society of Clinical Oncology has proposed, that everyone with metastatic disease and symptom burden should get palliative care, are we recommending something that we'll never be able to achieve, especially in low resource areas and smaller hospitals?" she asked.
"We don't have the best instruments for measuring what palliative care does to benefit people and for identifying those who need specialized palliative support before it becomes obvious. Wouldn't it be better to identify some basic, evidence-based triggers to target our efforts?"
Regardless of how this change manifests, however, it's clear that to ensure palliative services are reaching the patients who can benefit from them most, more work is needed to answer these—and many more—questions.
- National Consensus Project for Quality Palliative Care. Clinical Practice Guidelines for Quality Palliative Care. 4th ed. Richmond, VA: National Coalition for Hospice and Palliative Care; 2018. www.nationalcoalitionhpc.org/ncp/.
- Temel JS, Greer JA, Muzikansky MA, et al. Early palliative care for patients with metastatic non–small-cell lung cancer. New Engl J Med. 2010;363:733–742.
Read the next article or go to the table of contents.