The 2020 Innovator Awards Co-Third-Place Winning Team: Memorial Sloan Kettering Cancer Center
Care management programs have become ubiquitous in the area of primary care delivery, particularly for patients living with such chronic illnesses as diabetes, congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), coronary artery disease (CAD), and asthma, among others. But what about the intensely complex area of oncology? Given that the U.S. healthcare system remains very much focused overall on the highest levels of acuity, including inpatient hospital care delivery, it turns out to be quite a challenge to upend how the care of cancer patients is managed in reality—including the care of those patients undergoing chemotherapy.
At New York’s renowned Memorial Sloan Kettering Cancer Center, a multidisciplinary team of clinicians, care delivery designers, and data and IT experts has been moving forward with a program that is sure to be looked at nationwide. The program called “InSight Care,” is focused on “enabling seamless connection with [the hospital’s] patients wherever they are, whenever they are in need.” Specifically, in addition to the broad goal of enhancing overall quality of care delivery, the programs goals include efforts to “focus on keeping patients, their caregivers, and our care teams connected at all times through various communication techniques; deliver predictive, anticipatory, and proactive care for patients on active treatment;” and “maximize the physical, psychological, and emotional comfort of patients,” program officials say. And all of those goals have come up against the unpredictability of patients’ care needs when they are receiving chemotherapy.
On average, cancer patients undergoing chemotherapy experience one unplanned hospital admission and two emergency department (ED) visits a year, often as the result of patients proving to be ill-equipped to manage side effects at home. Meanwhile, patients often assume that little can be done, and wait to seek assistance until their symptoms worsen; and clinicians find that they aren’t in the loop early enough to intervene proactively.
Thus, Memorial Sloan Kettering leaders went to work, bringing together clinician, operational, and data analytics and IT representatives, to build the program together. Among those leading this initiative have been Ophelia Chiu, director, design strategy and innovation; Bobby Daly, M.D., an assistant attending oncologist at MSKCC; and Isaac Wagner, senior director, analytics, at the hospital organization. They and several others have been core team leaders in the initiative, with participation by a large number of colleagues across the organization.
The leaders of the program note that, on average, patients complete more than 50 percent of the daily symptom assessments, and have helped to generate red alerts for pain and for other issues. And the pilot cohort of patients has experienced fewer hospitalizations than patients in a similar unenrolled high-risk patient group.
Looking back at the origins of the initiative, Daly notes that “Ophelia, Isaac, and I have been working with a number of fellow colleagues to implement this initiative. The core questions we asked ourselves were the following: how can we better monitor our chemo patients and manage them at home, so we can proactively intervene when they’re having symptoms? And, can we build a program where we can help our patients by providing more care for them when they’re outside of the clinic?”
Importantly, Daly notes that, in fact, “There are more similarities between the care management of patients with chronic diseases like COPD and CHF, and the care management of chemotherapy patients, than there are differences. As with those diseases, it’s good to identify symptoms early. Similarly with oncology patients on chemo, the goal is to intervene earlier and provide more engagement. So there are a lot of similarities, and I think we can learn from each other.”
Speaking of learning, the present initiative had as its foundation in a clinical trial that earlier had focused on patient-reported incomes. “Patients with advanced cancer were categorized by standard-of-care reporting by patients, where they’re calling in, or proactive symptom reporting,” Daly recalls. “Improvements in quality of life, fewer ED visits, and longer lives,” were the results of the effort to collect patient-supplied data. “That was really a breakthrough in understanding that patients could live longer and have better outcomes, through proactive patient reporting of symptoms,” he adds. “And because of the results achieved, we decided to implement the program for our patients outside the clinical trial setting, and to make it a part of the standard of care,” via the InSight Care program. What’s more, he adds, “Since the study was conducted, patients have gained more access to mobile devices, so we had the added opportunity to monitor patients more intensively, using their mobile devices.”
Chiu explains that “There were three things that came together to drive this work forward. There was work happening around trying, from an analytics perspective, to understand clinically if there were things driving patients into urgent care, for example. So, which patients might be more at risk for urgent care visits? So there was strategic work to dig into, around that idea. In addition, there was work happening on my team to start to think about our vision for what we were calling visual patient care. How could technology help us to think differently so that we weren’t limited to things happening during those inpatient visits?”
Chiu continues, “There’s a lot that happens in between visits that we often didn’t have transparency on. So there’s a bit of a pun in the name ‘InSight’: on the one hand, the data might be able to provide insight and perspective, and there’s also the idea of ‘in site’—referring to patients actively in our care. And a third element was inpatient availability. For us, the urgent care center is the primary doorway into our inpatient beds. So if we were proactive, in terms of understanding what was happening with urgent visits, that that would subsequently reduce avoidable inpatient stays; we could intervene earlier. So it were those three things that came together as the impetus for the program.”
Wagner, coming at the initiative, with his data team, from the analytics side of the table, says that one of the core issues to be worked out was, “What should the relationship between a clinician and a model be? You really don’t want to take control or power away from a physician. So how do these models assist or empower a doctor to get to the right choice? Who should enter this program, and who’s more at risk? Those are the key questions. We wanted to do this in a way that was valuable for the clinicians, in terms of helping them to assess levels of risk.”
With regard to that, Wagner emphasizes that “Data models cannot help you make the decisions; the doctors make the decisions. These models help doctors and nurses make decisions. This is different from machine learning, as applied to other clinical areas. And that’s one of the biggest lessons. This really aims to help and assist and support, and not to control. Any attempt to control behavior, such as, use this drug and not that drug, is destined for failure,” he emphasizes. “The value is in the fact that a machine can sort through tons of data and provide insights to doctors. The machine can’t sort through softer issues such as what a patient’s personal goals are and what they can do for themselves; a human can do that better.”
And, when asked what the biggest challenge has been from her perspective, Chiu says that “We’ve really been trying to innovate within our existing care system. We created a new clinical program with new roles, processes, teams, and technology, all at the same time. And not surprisingly, some of the challenges upcoming are around how we scale and increase adoption of this beyond our initial pilot group. It’s something that we’re actively working on. And on the flipside, there’s the scaling from a clinical and operational perspective. But we’re also working on the fact that we had cobbled together a suite of tools that our extended care team was using to manage patients; but we also knew that that was a starting place in terms of the available tools. So we’re trying to assess what the needs are for that team.”
“I think we are finding from our patient interviews that patients really appreciate this type of intensive monitoring in the home; they feel it provides them with a support network and a low barrier to access, because the InSight team really acts as an extension of the primary oncology team, as it acts as a form of 24/7 support,” Daley says. “This provides insights for treatment planning, and it helps us figure out how to integrate this data back to patients, and communicate it back to them, so they can understand their symptomology over time. That’s another learning. And we’ve started to introduce televisits; but could there be a way to get vital signs from the home, and more intensive telehealth capabilities in the home?”
Ultimately, Chiu and Daly note, the goal is to extend the same care management methodologies to all Memorial Sloan Kettering patients, as the leaders of the initiative create the foundations needed to scale up. And, in addition to improving patient outcomes, the initiative is enhancing the patient experience. “The patients really like it, and that improves their adherence,” Daly notes. “They feel someone’s reading these assessments and is responding quickly.” And for cancer patients with often-intensive care management needs, that indeed appears to be a game-changer.
