Palliative Care: a Data Scientist's Perspective
- Natasha Hagemeyer
Palliative care access has significantly expanded across clinical settings over the last 20 or so years. In 1998, only about 15% of hospitals had a palliative care program of any kind. By 2019, that figure rose to about 72% in the United States. Certain regions, larger hospitals, and nonprofits are even approaching the 100% mark.2 There has also been a large uptick towards specialization in the area of pain management across multiple clinical programs and open acknowledgement that the notion of “quality of life” must be addressed on equal footing with “quantity of life”.
This increase in access, however, is not evenly distributed across the United States. Southern, rural, and/or lower income communities continue to have low levels of access, with only about 17% of rural hospitals with fifty beds or less reporting a palliative care program of any kind.2 Also, even today, the notion of palliative care continues to be conflated with hospice and end of life care rather than matters to be addressed earlier in a patient’s treatment trajectory.
During this COVID-19 crisis, the focus on and need for expert-level delivery of comfort measures and pain management has surged commensurately with the need for all other care delivery. Palliative care offers a unique set of desirable treatment characteristics that are especially meaningful during periods of increased stress on the healthcare system and a large surge of high-risk patients who are both suffering and afraid. It adds a calming voice, a sense of dignity, and control for both patients and providers when all these may be in short supply. A strong foundation in palliative care delivery can reduce overly aggressive end-of-life care (e.g., avoidable hospitalizations, unnecessarily lengthy hospital stays, unnecessarily lengthy intubation and other forms of life support, or expensive and risky procedures), thus freeing capacity and scarce resources.
The general population's desire for palliative care measures has increased significantly—over 90% of people would expect themselves or their loved ones to receive palliative care measures if they were facing a serious illness.3 Proper identification and prioritization of patients who would benefit from palliative care measures is a struggle in the best of times. One can only imagine what a challenge it is today. Understanding who should receive palliative care, when they should receive it, and what amount of care is appropriate is a significant skill, requiring a non-trivial amount of training and guidance.4
The recent COVID-19 related surge aside, the need for appropriate delivery of compassionate, palliative care will only continue to increase in the United States and across the globe as the world population lives longer and is more likely to experience progressive chronic disease. Given that most people value palliative care, would expect to receive it if severely ill, and are more likely to actually need that type of care during the course of their lives, physicians need an increased level of confidence and understanding of how to effectively identify and deploy comfort care measures. Research suggests that training for the typical, non-specialized provider continues to lag, and the ability to properly identify the appropriate level of care continues to be a challenge for many doctors and other prescribing clinicians. A significant need exists to continuously increase the clinicians sense of skillfulness, knowledge when to appropriately refer for consults, empowerment, access to support resources, and ongoing education about palliative care.5
Given that most physicians have only a rudimentary level of palliative care training, many organizations have increased reliance on technology to assist physicians. The most common approach is through “Clinical Decision Support” systems (CDS). Such technologies generate a prescriptive care plan for physicians to follow, at the point of care, based on what is usually a complex and often opaque algorithm. While this approach has merit and is especially beneficial when the clinician is a mid-level provider such as a nurse-practitioner or physician assistant, many physicians resist this approach as they feel like decision making has been taken out of their hands.
Palliative care has never been more important than it is currently, both in the short term due to COVID-19 and long term due to ongoing demographic changes. More clinicians are charged with making sophisticated decisions on how to appropriately prescribe palliative care measures, while there continues to be a lag in education on best practices. In the next blog posting, I will discuss the approach we designed to close the knowledge gap, educate and empower through peer comparison, and help physicians ensure that they are continuously improving and aligning with best practices in palliative care delivery. It is our mission at Agathos to surface opportunities toward valuable care through great clinical decision making, without taking the power of the decision out of the hands of the physician.
- https://bulletin.facs.org/2019/11/practicing-primary-palliative-care-a-call-to-action/Fast-Track Treatment in the Emergency Room, January 1, 2011. Healthcare Design Magazine. Available at:
Natasha is a published academic, researcher, and a Data Scientist working for Agathos. She cut her teeth conducting research in behavioral ecology. She currently leverages her data expertise to identify patterns and insights in healthcare data for improved patient care.