Authors

  1. Eastman, Peggy

Article Content

COVID-19 has shined the light of urgency on the need for effective digital health care tools in oncology. Even before the pandemic began, such tools as telehealth were growing in importance in cancer practices. To examine the scope and role of digital applications in cancer care and research, the National Cancer Policy Forum (NCPF) of the National Academies of Science, Engineering and Medicine (NAS) sponsored a virtual workshop in Washington, D.C.

  
digital health, heal... - Click to enlarge in new windowdigital health, health care informatics. digital health, health care informatics

The online workshop, which also explored critical regulatory and reimbursement challenges, was held in collaboration with the NAS Forum on Cyber Resilience. A dissemination report from the meeting will be forthcoming.

 

"This is an evolving field," said workshop chair and NCPF member Lawrence N. Shulman, MD, MACP, FASCO. "Slowness was not an option...We had to be more agile than we've ever been" when COVID-19 struck.

 

Although the workshop was organized before the pandemic hit, it has served as a catalyst to greatly heighten the interest in and use of digital tools, noted Shulman, who is Deputy Director for Clinical Services and Director of the Center for Global Cancer Medicine at the University of Pennsylvania's Abramson Cancer Center. He said the workshop was the beginning of a discussion on digital health tools in oncology. To move the field forward, "we need national will" from patients, clinicians, regulators, and payers, along with scientific methods of research and validation of digital tools.

 

There is no question that the pandemic has accelerated the use and adoption of digital tools in the clinic such as telehealth for patient visits and remote monitoring, said Mia Levy, MD, PhD, Director of the Rush University Cancer Center, System Vice President for Cancer Services at the Rush System for Health, and Associate Professor of Medicine in the Division of Hematology and Oncology at Rush University.

 

For example, she described the transition from paper consent forms to remote consent forms with a digital signature, and stressed "this is a pretty big evolution."

 

On the plus side, Levy noted that telehealth can offer patients personalization and convenience of care, along with minimization of wait time.

 

"Many of our patients were incredibly appreciative of this shift to telehealth," she said, noting that videoconferencing allows consultation without a mask, thus letting health professionals pick up on patients' facial cues, and also allows family members to participate in consultations if patients desire it.

 

On the downside, Levy said patients vary widely in their technical abilities, failed attempts at videoconferencing produce frustration, telehealth has clear limitations for physical examinations, and it is harder for health professionals to make a connection with patients. Levy and her colleagues developed clinical guidelines for virtual visits. While they are now back to seeing patients in person, she said that they are still using telehealth for about 15-20 percent of patient visits.

 

"Digital health is critical for precision medicine, patient centricity, and learning systems," said Amy Abernethy, MD, PhD, Principal Deputy Commissioner of Food and Drugs and Acting Chief Information Officer at the FDA.

 

She noted the FDA has a Digital Health Innovation Action Plan, and the agency is considering whether it has the right regulatory pathways for digital health technologies.

 

"Oncology is leading the way, which is both humbling and makes me proud and reminds me how hard this work is," said Abernethy, a hematologist/oncologist by training. FDA's digital health action plan, an initiative of its Center for Devices and Radiological Health (CDRH), has a threefold vision: 1) issuing guidance to provide clarity on the medical software provisions of the 21st Century Cures Act; 2) launching an innovative pilot precertification program to develop a new approach to digital health technology oversight; and 3) building FDA's bench strength and expertise in CDRH's digital health unit.

 

A number of speakers described innovative digital tools for use in oncology practice and research, which is an actively developing area. These include the following.

 

1. A bidirectional communication artificial intelligence (AI) tool called Penny is proving useful for remote monitoring, said Samuel U. Takvorian, MD, MS, Instructor in the Division of Hematology and Oncology at the University of Pennsylvania Perelman School of Medicine. "It's an automated chatbot with a human feel," said Takvorian. Patients report their symptoms to Penny, which they can do at any time. He said the AI tool has improved adherence, reduced call volume, and improved provider triage.

 

2. An individualized digital feedback tool for reporting and understanding symptoms called the Recovery Tracker developed at Memorial Sloan Kettering (MSK) Cancer Center has helped patients feel more connected and cared for and cut down on calls to the nursing staff, according Andrea L. Pusic, MD, MHS, FACS, FRCSC, who formerly was at MSK and is now Chief of the Division of Plastic and Reconstructive Surgery and Co-Director of the Patient-Reported Outcomes, Value, and Experience Center at Brigham and Women's Hospital. "Patients don't have to wait for the doctor to tell them they're OK," she noted. "A lot of it was clinical common sense." Pusic is the developer of BREAST-Q, a patient-reported outcomes measure for breast surgery.

 

3. A three-dimensional deep learning-based imaging algorithm to detect the presence of cancer in lymph nodes yielded results as good or better than those from radiation oncologists, said Sanjay Aneja, MD, Assistant Professor in the Department of Therapeutic Radiology at Yale School of Medicine. He said quantitative imaging methods do appear to have utility in clinical oncology, but also stressed that validation of these methods is required. Aneja completed post-doctoral training in machine learning at the Center for Outcomes Research and Evaluation (CORE).

 

4. The Stanford-Palo Alto Medical Foundation Oncoshare Project is a neural network model in the San Francisco Bay area that uses "big data" to improve breast cancer care. It integrates data from local and national sources to detect unwarranted variations in care at institutions in the same area and can inform solutions to avoid such variations. "Cancer data sources are fragmented; we aim to integrate data sources across health care settings," said Allison W. Kurian, MD, MSc, Oncoshare principal investigator and Associate Professor of Medicine (Oncology), Associate Professor of Epidemiology and Population Health and Director of the Women's Clinical Cancer Genetics Program at Stanford University School of Medicine.

 

5. An analytical algorithm predicting 6-month mortality in cancer patients can be used to help guide a serious illness conversation. Using the algorithm, which incorporates clinician views, was helpful in saving physicians' time and resulted in a major increase in discussions on serious illness, said Ravi B. Parikh, MD, MPP, Assistant Professor of Medical Ethics and Health Policy at the University of Pennsylvania, Staff Physician at the Corporal Michael J. Crescenz VA Medical Center, and Senior Clinical Advisor at the Coalition to Transform Advanced Care. "We as oncologists often get predictive prognoses wrong," said Parikh, noting that these prognoses are frequently overly optimistic, which delays conversations on serious illness and end-of-life care and can lead to overuse of treatments. "Overly optimistic prognoses are not benign," stressed Parikh, who specializes in observational data analysis and pragmatic clinical trials.

 

6. The Jvion predictive analytics tool identifies patients who may be high-risk but are less obvious to the clinician, said Sibel Blau, MD, President and CEO of the Quality Cancer Care Alliance Network, Medical Director of the Hematology-Oncology Division and Head of Precision Medicine at Northwest Medical Specialties, and Clinical Associate Professor in the Hematology Division at the University of Washington. Using Jvion, which has a 30-day mortality vector, can improve care, decrease use of the emergency department, and increase hospice use, said Blau, who is Immediate Past President of the Washington State Medical Oncology Society.

 

Peggy Eastman is a contributing writer.

 

Important Emerging Digital Issues

 

* Digital health tools should focus on and incorporate the patient voice and patient-reported outcomes. "Early in my career I began to believe that clinicians underreport their patients' symptoms," said Ethan Basch, MD, MSc, Director of the Cancer Outcomes Research Program at the Lineberger Comprehensive Cancer Center at the University of North Carolina at Chapel Hill. Basch leads a longstanding digital health research program which has established that up to half of patients' symptoms go undetected during routine cancer care and in clinical trials, and that patient-reported outcome-based digital strategies substantially improve detection. He created a system for the National Cancer Institute to collect patient-reported adverse events during cancer trials called the PRO-CTCAE. Dasch recommended that payers reimburse for the use of validated tools-wearable and mobile, for example-that enable patient-reported symptom reports and remote monitoring.

 

* Digital health tools must be rigorously validated before being used in the clinic, and they must make the busy clinician's life easier, not harder.

 

* A stable, consistent digital health reimbursement platform which overcomes restricted use is needed; this will require federal and state legislation. Also unclear is whether the parity legislation for telehealth reimbursement put in place at the time of the pandemic medical emergency declaration will expire or continue post-COVID. Seema Verma, MD, Administrator of the Centers for Medicare & Medicaid (CMS), has stated that CMS wants to continue reimbursement for some expansion of telehealth beyond the pandemic.

 

* Telehealth reimbursement should be for telephone visits as well as video visits, as some patients only have phones. Verma has noted that many patients may not have access to or feel comfortable with video technology. To leverage telehealth for all Americans, legislation should enable free cellular minutes for telehealth visits for low-income patients.

 

* Stringent rules on privacy and security to protect patients are needed for the use of digital health tools in oncology. "Our data is not an abstraction to us; it represents generations of suffering for my family," said Andrea Downing, Community Data Organizer and President and Co-Founder of the Light Collective. In 2018 Downing discovered a security vulnerability affecting closed groups on Facebook. "We need a better federal privacy law for data," stressed Kristen B. Rosati, JD, a partner in Coppersmith Brockelman, PLC, and a past president of the American Health Lawyers Association. There can be taxonomies of harm from data breaches, warned Fred Schneider, PhD, Chair of the NAS Forum on Cyber Resilience and the Samuel B. Eckert Professor of Computer Science at Cornell University. "We strive to anticipate problems before they arise," added Schneider, who is now focusing on system security.

 

* Digital health tool applications should be interoperable across health systems. What is needed is a guiding coalition around any use of a digital health application; that coalition should address specific use applications and set usage standards.

 

* Developers of digital applications should take international considerations into account. "Cancer initiatives have gotten increasingly international in nature," said Yann Joly, PhD, FCAHS, AdE, Associate Professor, Department of Human Genetics & Biomedical Ethics Unit, Faculty of Medicine; and Research Director for the Centre of Genomics and Policy at McGill University. Joly noted that the advent of cloud computing has led to geographically dispersed data centers, and there is a need for harmonization at the global level and alignment of data privacy laws among different countries.