Feature

High-Tech Health

New Technological Approaches Are Transforming the Delivery of High-Touch Palliative Care

By Larry Beresford

Information technology is evolving rapidly in health care. Hospice and palliative medicine clinicians may feel dizzy contemplating the many advances in predictive modeling, artificial intelligence, machine learning, and telehealth—and their patients’ engagement with new devices, personal applications, and “gizmos.”

Medicine is sometimes slower to take advantage of technological advances than other industries, said Joe Rotella, MD MBA HMDC FAAHPM, AAHPM’s chief medical officer. “We can get so caught up in our own world that we forget all of the other developments going on. Any of these new technologies could be helpful in our work, but they sometimes come with unintended consequences and hidden burdens.”

Take the electronic medical record (EMR). Once promoted as a time saver and tool for quality improvement (QI), for many clinicians, it instead has been a difficult process of adaptation and even a contributor to burnout.

“For some, the EMR is seen as the machine that interferes with high-touch care,” said Arif Kamal, MD MBA MHS FAAHPM, at the Duke University Cancer Center in Durham, NC. “There can be friction between technology and caring. But there has to be a middle ground, where technology can be beneficial to clinicians in the delivery of palliative care, especially when it allows us to identify patients with particular needs in advance of when those needs arise.”

Inspiration Begins On Rounds

Even simple solutions, like an online brochure, website, or YouTube video, can improve patients’ engagement with their care, Kamal said.

Duke’s palliative care program strives to solve real-world problems using applications for common devices like smartphones, he said. “Our inspiration starts on rounds. One day in clinic, I realized there were holes in our schedule because of no-shows. Patients were coming to oncology appointments on the same day and then leaving, skipping their appointment with palliative care. They didn’t recognize that palliative care is different than oncology. They didn’t have the knowledge or language to put value on it. So we developed [an] app called PCforME.”

The app introduces patients to palliative care, prepares them for upcoming appointments, and encourages them to use the Palliative Care Passport—a printable document for sharing their responses with their caregivers, loved ones, or the clinical team.

“In palliative care, the way we engage patients is to encourage them to ask questions,” Kamal said. With a free app called Prepped, the patient can record information about their own clinical situation and receive pertinent questions to ask based on that condition and on crowd-sourced answers from other patients, helping prepare them for the initial palliative care consult.

Other apps with palliative care implications being developed at Duke include*

• Extra Layer of Support (ELOS), a web-based tool to engage newly diagnosed cancer patients in shared decision making with their oncologist by offering information about appropriate supportive resources available in the cancer center, which may include nutrition, physical therapy, chaplaincy, clinical trials—and palliative care.
• BiteSizeQI, a web-based tool to help clinicians and other frontline staff develop QI skills and apply them to a QI project. The content of the program is not palliative care specific, Kamal said, “but we recognize many palliative care programs struggle with the skills and leadership needed for QI. This app offers the equivalent of several hours of classroom time.”
• HOPE—the Hospice Preparation and Education mobile app, which, much like PCforME, introduces patients to hospice care and connects them with their care team and collects patient symptom and caregiver distress data.

Early Warning Systems

Some of the new technologies will impact and improve AAHPM members’ practices, said Dana Lustbader, MD FAAHPM, chief of the department of palliative medicine for ProHEALTH, an accountable care organization in Lake Success, NY, and a leader in the use of telehealth for palliative care.

There are many new devices that can be helpful in monitoring single-organ function, like remote heart monitors—including the simple-to-use Fitbit—or Bluetooth-enabled patient weighing scales. “You can even use an [electrocardiogram] app on an Apple device, where the patient’s heart rhythm gets interpreted and transmitted to a provider. Similarly, for diabetes, there are skin sensors and smart contact lenses for glucose monitoring,” Lustbader said.

For frail, homebound older patients, wireless or Bluetooth-enabled sensors could note if the toilet hasn’t been flushed for a while, suggestive of early dehydration or reduced oral intake, or indicate that a patient isn’t moving around as much and transmit the data to a remote site, provider, or concerned family caregiver. Hospitals increasingly use modified early warning systems, alerting clinicians to potential deterioration in a patient’s condition or change from their baseline status. “I think we’ll have that in the home before too much longer,” Lustbader said.

“I think technology is a great thing. It will add a lot of value. We’ll eventually get better at using it. But there’s also a risk,” she added. “Some of these gizmos and toys are a lot of fun, but are they helping people with multiple chronic conditions or serious illness? Not unless they are coupled with a robust palliative care model that provides interventions during off hours, so the patient can reach a clinician who can respond in real time to address an urgent medical issue, pain, or symptom so the patient is protected from having to go to the emergency room. Health-related apps, gizmos, and toys are only as good as what you do with the information.”

“We do virtual visits often and have made it really easy for our patients and their caregivers. We connect with them right on their own smartphones,” Lustbader continued. “I have come to believe that in certain situations, telehealth visits might even be better than in-person visits. The patient is more comfortable at home on their bed or couch. They’re more relaxed and sometimes more honest. You can see what they’re wearing. You can ask them to pan the room with their phone. When you visit in person, you are a guest.”

Telemedicine technology can deliver compassionate, person-centered care with just a click of a handheld device, Michael Fratkin, MD, of the ResolutionCare Network in Eureka, CA, wrote in a recent blog post for the Center to Advance Palliative Care.¹ Video-enabled medicine now supports more than 40% of the center’s palliative care visits with patients. But in the absence of population-based contracts, healthcare reimbursement has not caught up to providing adequate support, he notes.

Video technology, the most rapidly growing area of telemedicine, also can be used in creative ways to expand access to palliative care, confirms a new report in the Journal of Palliative Medicine.² According to Lustbader, another application of telemedicine in palliative care is enabling provider-to-provider connections; with the help of video, it also enables clinicians to include geographically distant family members in meetings and expand rural access.

Expanding Use of Predictive Modeling

Artificial intelligence, predictive modeling, and similar technologies soon will be able to predict what may happen to patients in their near future, Lustbader added. “You can look at claims data for people with serious illnesses, recent hospitalizations, and the need for more help.” Health information exchanges also could provide the team with an alert when a patient is in the emergency department. “In palliative care, many of us are using predictive modeling today. When it comes to population health and accountable care organizations, we have gotten better at finding the right patient in a timely manner,” she said.

Charlotta Lindvall, MD PhD, a palliative care physician at Dana-Farber Cancer Institute in Boston, has been working with machine learning to improve serious illness care—while also safeguarding its use. “There are fantastic opportunities to improve serious illness care,” she said. “Our machine learning initiative, part of a 3-year clinical pilot at Dana-Farber, is built on the recognition that important clinical information is written into the medical record.”

Free text in electronic charts is a rich source of that information, but right now, it’s difficult to access. That’s why Lindvall is using natural language processing and a method of advanced computing to identify information about the patient’s symptoms, such as shortness of breath, anxiety, or depression, that otherwise are not easy to track. She and her colleagues also have studied how deep-learning algorithms could identify in the EMR whether there has been a documented serious illness conversation with a patient during their stay in the intensive care unit.³

“If we used these methods to analyze notes from 50,000 cancer patients, we could identify symptoms and recognize emerging needs,” Lindvall said. “Within 3 years, we’ll be using this kind of method in palliative care, much as we now do in the [emergency department].” Mortality predictor models also could use the health data collected to help inform prognostication or even predict hospice eligibility.

"Most of the mortality predictors are used with good intention, but I think we need to be careful so we don't end up back in the death panels debate," Lindvall said.

“A lot of these technologies will soon trickle down to palliative care and hospice—but our role will be to advocate for patients with serious illness. Think about the incentives in the technological space. Is it pushed by the healthcare system or by us? Right now, it’s lacking transparency,” she said.

Starting the Conversation

The challenges of patient engagement in palliative care often are seen in the context of advance care planning and the reluctance of many adults to complete advance directives. There are barriers to starting this conversation—including that advance directive forms are scary, said Rebecca Sudore, MD FAAHPM, a professor in the division of geriatrics at the University of California, San Francisco.

Sudore and her team developed PREPARE for Your Care, an online resource to help people begin the process of advance care planning for future medical decision making. Launched in 2013 and available in English and Spanish, the program uses how-to video stories to guide users through five planning steps focused on values, preferences, and communication.

PREPARE also offers written pamphlets, a question guide, state-specific advance directives, and toolkits for group movie events and medical visits. PREPARE materials can be completed at home and can be given to patients before an appointment or as homework to bring to their next palliative care appointment. They also can be made available at the health system’s check-in locations and at a kiosks in clinic lobbies.

“We went with a Web-based platform for wider dissemination,” Sudore said. “This is a patient-centered tool—not clinician focused—created with and for diverse older adults. We set up PREPARE and the advance directives to be easy to use, easy to read, and legally valid in all 50 states.”

At the Salt Lake City, UT, Veterans Affairs (VA) Medical Center, Shaida Talebreza, MD HMDC AGSF FAAHPM, section chief for geriatrics and academic palliative medicine, and her team worked together with the VA National Center for Ethics in Health Care (VANCEHC) and three other pilot sites to develop and implement the VA Life-Sustaining Treatment Decisions Initiative. It is designed to document patients’ preferences for different life-sustaining treatments and store that information as a permanent note with durable orders in the national VA EMR so it is accessible when it is needed.

“Now, when we have the conversation with the patient, the note documenting it is easily found in the medical record using a template for capturing the content of the conversation. It can also automatically generate relevant medical orders,” Talebreza said. In other settings, these kinds of conversations aren’t captured and the patient has to repeat them multiple times. “We have captured all of it, patient goals, values, preferences, and treatment decisions. The next time the patient comes to the hospital, we have access to it.” The information is comparable to a POLST form—but this data is captured electronically for the entire VA system.

“We have tried to elevate the level of the conversation. It’s about the quality of the interaction, finding out what matters to patients, what’s important in their lives, their goals and values—not just treatment decisions,” Talebreza said. “We need to be trained in how to have these conversations.” VANCEHC worked with VitalTalk, a national initiative to promote better end-of-life conversations, in the development of a national curriculum for healthcare providers, which has been disseminated to all VA facilities.

Since the initiative launched nationally in July 2018, a total of 217,464 goals-of-care conversations about life-sustaining treatment have been conducted and documented in the VA’s EMR. “It’s a relatively simple application but with wide implications,” Talabreza said. “In Salt Lake City, our home-based primary care team (census of 150 frail veterans) fully implemented the initiative. We did a review for all deaths to see if the patients were offered a goals-of-care conversation. One hundred percent were, 84% accepted, and of those, 91% who died received end-of-life care congruent with their stated preferences. It’s inspiring to work on a project meant to ensure that patients receive end-of-life care that is consistent with their wishes.”

References

1. Fratkin M. Telemedicine 3.0: how video technology can support people with serious illness. Center to Advance Palliative Care. https://www.capc.org/blog/telemedicine-30-how-video-technology-can-support-people-serious-illness/. Published June 12, 2019.
2. Calton BA, Rabow MW, Branagan L, et al. Top ten tips palliative care clinicians should know about telepalliative care [published online ahead of print June 25, 2019]. J Palliat Med. doi: 10.1089/jpm.2019.0278
3. Chan A, Chien I, Moseley E, Salman S, et al. Deep learning algorithms to identify documentation of serious illness conversations during intensive care unit admissions. Palliat Med. 2019 Feb;33(2):187–196.

Larry Beresford is an Oakland, CA-based freelance medical journalist specializing in hospice and palliative care. Contact him at This email address is being protected from spambots. You need JavaScript enabled to view it. or see his website at www.larryberesford.com.

*These various apps are owned by Duke University. Kamal reports a conflict of interest related to their commercialization. For more information about them, contact him at This email address is being protected from spambots. You need JavaScript enabled to view it. .

Read the next article or go to the table of contents.