Opportunities and Challenges for Using Digital Health Applications in Oncology Proceedings of a Workshop (2021) / Chapter Skim
Currently Skimming:
Proceedings of a Workshop
Proceedings of a Workshop
Pages 1-48
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 1... ...
Digital health encompasses health content, digital health interventions, and digital applications, such as communication tools connecting patients and clinicians (e.g., secure email in the patient portal, text, chat, video visit) , remote monitoring tools, clinical decision support tools, and systems for exchanging health information.
|
From page 2... ...
Many workshop speakers said that the opportunities presented by digital health tools are particularly compelling for oncology; however, capitalizing on these opportunities necessitates careful attention to the design, implementation, and use of digital health technologies. Lawrence Shulman, deputy director for clinical services at the Abramson Cancer Center at the University of Pennsylvania, and Fred Schneider, professor of computer science at Cornell University, outlined the goals of the workshop.
|
From page 3... ...
• Evaluate how digital health tools affect the quality and effi ciency of care delivery and the impact on patient outcomes to inform appropriate reimbursement models. (Bradley, Strawbridge)
|
From page 4... ...
(Fuchs) Participants discussed myriad opportunities for using digital health tools to enhance cancer care by expanding access to high-quality care; using scarce medical resources more efficiently; improving patient–clinician communication; and enabling patients to safely and effectively manage more aspects of their care at home, especially during the ongoing pandemic.
|
From page 5... ...
Like any technology or medical tool, digital health tools pose potential risks for the safety and privacy of patients and the security of health data; they also have associated ethical, legal, and financial ramifications. While they offer many exciting possibilities to improve cancer care, several participants noted, these technologies also have the potential to create access barriers and exacerbate disparities unless they are carefully designed, implemented, and evaluated.
|
From page 6... ...
OVERVIEW OF DIGITAL HEALTH APPLICATIONS IN ONCOLOGY While digital health tools are being implemented across many settings of care and disease areas, Shulman noted that the oncology setting presents unique considerations and potential benefits. The cancer care continuum spans prevention and early detection; diagnosis; cancer treatment, involving an array of diverse and complex treatment options (including clinical trials)
|
From page 7... ...
At home, digital health technologies have enabled patients to interact with clinicians directly via telephone, online chat, and video telehealth visits; use remote monitoring tools such as patient-reported outcomes and biosensors to feed data to their health care team; and download health data from various sources onto personal devices. Shifts in Clinical Trials and Evaluations of Novel Therapies Shah noted that many clinical trials in oncology have been affected by the pandemic: Patient enrollment slowed, and inves tigators are unsure how the COVID-19 pandemic will affect data collection and trial completion (Upadhaya et al., 2020)
|
From page 8... ...
First, FDA-issued guid ance encouraging use of telemedicine or other tools in clinical trials to reduce the risk of COVID-19 exposure associated with in-person clinic visits. Shah noted that this is especially important for patients with cancer, who are likely to be immunocompromised.
|
From page 9... ...
He added that decentralization could also facilitate the involvement of more diverse, representative populations in clinical trials. Kluetz encouraged evaluation of data from clinical trials that deployed a hybrid decentralized approach during the COVID-19 pandemic to help determine which remote assessment tools work best and to inform clinical cancer research going forward.
|
From page 10... ...
tools such as chatbots or diagnostic algorithms can never fully replace human care; their role is to augment and inform clinical decision making. Shulman, Levy, and several other participants emphasized the importance of ensuring that digital health tools are safe and meaningful for patients and helpful for clinicians, can improve patient outcomes, and are appropriately reimbursed.
|
From page 11... ...
CANCER CENTER OUTPATIENT VISITS DURING COVID PANDEMIC NEW OFFICE VISIT NEW VIDEO VISIT RETURN OFFICE VISIT RETURN TELEPHONE VISIT RETURN VIDEO VISIT Pre-COVID Surge Preparation Reopening & Recovery National Deferring care Reschedule canceled appointments emergency Telehealth Increase clinic capacity Work from home (23%) Staff redeployment No visitor policy Illinois shelter Screening & testing in-place order Rush System for Health | 7/22/2021 1 FIGURE 1 Outpatient visits at the Rush University Cancer Center at the start of the COVID-19 pandemic.
|
From page 12... ...
Using new digital health tools also requires staff training and adjustments to clinical workflow. Levy noted that it is helpful to rehearse how a new digital health tool will be used in clinical practice, prior to its implementation in patient care.
|
From page 13... ...
4 are a key part of cancer care and integral to many digital health tools, yet Levy described challenges with integrating them into the EHRs. It can be difficult to develop questions to meet every possible clinical scenario, and the relevance of particular questions may vary substantially depending on a patient's cancer type, disease stage, and type of cancer therapy.
|
From page 14... ...
Workshop speakers discussed the use of digital health tools for remote monitoring, collection of PROs, and improved patient access to their health data. Remote Monitoring Technologies Peterson discussed the use of wearable, mobile, and other remote monitoring technologies in cancer care and cancer research.
|
From page 15... ...
trials started annually that include a connected 15 SOURCES: Peterson presentation, July 13, 2020; Marra et al., 2020. digital product.
|
From page 16... ...
She urged the development of infrastructure to support their implementation in diverse populations, particularly among underserved individuals. Peterson noted that health equity challenges -- such as disparities in patient access to cancer care, varying digital health literacy rates, and inadequate resources among some health systems -- need to be addressed in order to avoid inadvertently exacerbating existing health disparities when implementing digital health tools.
|
From page 17... ...
PROCEEDINGS OF A WORKSHOP 17 FIGURE 3 Conventional physician-led symptom monitoring (blue line) leads to underreporting compared to patient-reported outcomes (orange line)
|
From page 18... ...
Penny was built in response to several trends in cancer care: a shift of cancer care from the hospital setting to outpatient clinical practice and in the home, as well as an increase in the use of oral cancer therapies. Takvorian noted that while these trends can be beneficial for patients with cancer, they also transfer much of the responsibility for managing side effects, monitoring symptoms, and following treatment plans onto patients and their caregivers.
|
From page 19... ...
Preliminary analyses from a randomized clinical trial of more than 2,500 patients found that those who used the Recovery Tracker reported less anxiety and a greater sense of connection with their care team. In addition, nursing staff were handling 0.4 fewer nursing calls per patient, which Pusic said was a meaningful reduction given that the center performs more than 10,000 surgeries per year.
|
From page 20... ...
, when incorporating PROs into digital health tools for routine clinical care. He noted that high-quality ePRO assessments in clinical care can generate a robust stream of structured real-world data that could inform cancer therapy development, pharmacovigilance, and optimal care strategies for patients with cancer.
|
From page 21... ...
By bringing these data together in a single clearinghouse, the organization can offer researchers access to large data sets and can help to match patients with relevant clinical trials. Clinician-Oriented Technologies Amid growing pressure to improve value in cancer care, Shulman said that digital tools present an opportunity to improve the quality of care while also improving efficiencies in care delivery.
|
From page 22... ...
The algorithm for lymph 11 Extranodal extension occurs when cancer cells in a lymph node have broken through the capsule and spread into the surrounding tissue. See https://www.mypathologyreport.ca/ extranodal-extension (accessed June 14, 2021)
|
From page 23... ...
Aneja stressed that the intention is not for ML to replace radiologists; rather, ML tools can be incorporated to support their work, with the potential for performance gains, improved patient outcomes, and a reduced need for invasive surgeries. Fuchs and Aneja were asked about the potential for integrated diagnostics -- or the convergence of imaging, pathology, and health informatics -- to improve cancer care.
|
From page 24... ...
Clinicians were then prompted to initiate these conversations in three ways: They received an email summarizing how often they held serious illness conversations compared to their peers; they were provided with a list of their patients who had been identified as high risk; and they were sent a text reminder (with the option to opt out) on days when patients with a high risk of dying within 6 months were scheduled for clinic visits.
|
From page 25... ...
Researchers, clinicians, and technology developers are exploring the opportunities -- and pitfalls -- of accessing, integrating, and using health data in a variety of research and clinical contexts. Leveraging Electronic Health Records and Artificial Intelligence to Improve Cancer Care Two speakers focused on opportunities to leverage EHR data to bridge the divide between cancer research and cancer care: Neal Meropol, vice president and head of Medical and Scientific Affairs at Flatiron Health, and Allison Kurian, professor of medicine, epidemiology, and population health at the Stanford University School of Medicine and director of the Stanford Women's Clinical Cancer Genetics Program.
|
From page 26... ...
, suggesting an unprecedented opportunity to decentralize clinical trials and create a platform for a learning health system14 that incorporates digital health data. Kurian described the Oncoshare Project,15 which integrates EHR data from two health care systems serving the San Francisco Bay area -- Stanford University Medical Center, an academic tertiary care system, and Palo Alto Medical Foundation, a community-based system -- along with data from the California Cancer Registry and genomic data from outside laboratories.
|
From page 27... ...
Using Artificial Intelligence to Improve Care Delivery Sibel Blau, medical director of hematology-oncology at Northwest Medical Specialties and president and chief executive officer of the Quality Cancer Care Alliance Network, discussed how ML and AI technologies can be leveraged to improve care delivery, describing her organization's use of the predictive analytics tool Jvion.16 Blau said that Northwest Medical Specialties is an 16 For more information, see https://jvion.com (accessed June 7, 2021)
|
From page 28... ...
In short, Blau stated that predictive analytics tied to interventions has led to improved patient outcomes and has become a key component of continual practice improvement. Making Data Usable The use of health data for research and practice improvements increasingly depends on the ability to access, understand, and combine data from different sources.
|
From page 29... ...
Digital health applications require vast training data sets; constraining data sets according to national borders therefore limits the effectiveness of these tools, especially for rare cancers. "The more data you get, the more data you fill into your tools and applications, the better they become," he said.
|
From page 30... ...
Bertagnolli stated that these standards will make critical health data accessible for AI retrieval and analysis, advancing progress in clinical care, clinical research, and health care policy. The first version of mCODE has 73 data elements that are essential for oncology care.
|
From page 31... ...
Bertagnolli noted that early mCODE applications show promise for enhancing patient care and the conduct of cancer clinical trials. For example, the ICAREdata collaboration used mCODE to obtain EHR data for clinical trial case reports.
|
From page 32... ...
She said that some -- but not all -- types of digital health applications require FDA clearance or approval prior to implementation in clinical practice. Abernethy noted that many digital health components that do not require FDA review are nonetheless of interest to the agency; for example, they may be important to the design of clinical trials that FDA may review for regulatory submissions.
|
From page 33... ...
She noted that the initiative 23 For more information, see https://www.fda.gov/media/106331/download (accessed June 1, 2021)
|
From page 34... ...
Legal Considerations Kristen Rosati, partner at Coppersmith Brockelman PLC, discussed legal considerations in digital health, including data security and the complex laws governing privacy protections for patient health data. The Privacy Rule requires HIPAA-covered entities28 to provide patients with access to their "designated record set," which includes their medical records, insurance claims, and other health information, and also gives patients the right to direct disclosure of these data to third parties.
|
From page 35... ...
As an example of this complexity, the HIPAA Privacy Rule governs de-identification standards for health information, including genomic data, which is highly relevant for oncology, but there is no government guidance on whether genetic data can ever be considered de-identified. Health information is only protected if it is "individually identifiable," and the Office for Civil Rights has concluded that not all genetic information is individually identifiable.
|
From page 36... ...
When she was having difficulty navigating the health care system, she sought support from other members of the BRCA community, who shared deeply personal information within a closed group on social media. However, Downing raised concerns that the privacy settings could allow individuals outside of a closed group to have access to group members' information.35 Arguing that health data deserves strict privacy controls, Downing suggested that all digital health applications should be subject to enforceable patient rights, a position she summarized as "no aggregation without representation." Downing, McGraw, and Rosati noted that the protection of patient health data once it leaves the control of a HIPAA-covered entity is often unclear; outside of HIPAA, the patchwork of state and federal laws can leave many patients' data unprotected.
|
From page 37... ...
Belmont also noted that patient-generated data are not easily included in EHRs, and clinicians will need to adjust workflows to incorporate real-time patient data for patient care. Belmont noted that remote digital telemetry applications used to monitor patient vital signs could potentially expose large numbers of patients to new risks, such as software errors.
|
From page 38... ...
"If our goal is to put telehealth infrastructure in place, remove restrictions for clinicians to practice across states, and incentivize through alternative payment models and payments for adoption, now is the time to put those in place permanently for patients." However, she cautioned that more evidence is needed to assess the impact on quality of care and patient outcomes, and the cost-effectiveness of telehealth. Lara Strawbridge, who directs the Division of Ambulatory Payment Models at CMS's Center for Medicare & Medicaid Innovation, suggested that there are many options for creating payment models and said that models need to offer flexibility and autonomy to clinicians to provide the best patient care so that payers are not micromanaging decisions.
|
From page 39... ...
Ethical Data Use Nancy Kass, deputy director for public health and professor of bioethics and public health at Johns Hopkins University, highlighted ethical considerations for digital health applications, including transparency, consent, patient privacy, data security, data access and sharing, ownership and governance, justice, and benefits. She noted that organizations constantly make decisions about where health data go and what data will be de-identified, as well as who can access the data under what circumstances.
|
From page 40... ...
In addition, she called for incentives and penalties to ensure that proper ethical protocols are followed. Several participants discussed opportunities to facilitate patients' consent, and even compensation, for sharing their health data.
|
From page 41... ...
For example, a lack of visual representation of people of color in digital health tools may support the false belief that people of color are less at risk from cancer (Allicock et al., 2013)
|
From page 42... ...
Holding developers accountable, having robust and comprehensive policies, and highlighting equity issues in multiple places will lead to better AI tools, she said. REFLECTIONS AND SUGGESTIONS Throughout the workshop, participants discussed examples of digital health tools in action; examined the opportunities for digital health to advance cancer treatment, improve care delivery, and support oncology research; and identified critical questions, needs, and potential challenges as the oncology field moves toward their broader adoption.
|
From page 43... ...
, Shulman said changes made in response to the COVID-19 pandemic have demonstrated that rapid change is possible, and he urged the oncology community to use this opportunity to advance digital health tools and technologies. Shulman stressed that a national push is needed from the entire oncology community to promote a better, evidence-based, patient-centered system of care delivery and research.
|
From page 44... ...
2016. Use of gene expression profiling and chemotherapy in early-stage breast cancer: A study of linked electronic medical records, cancer registry data, and genomic data across two health care systems.
|
From page 45... ...
2019. Two-year survival comparing web-based symptom monitoring vs routine surveillance following treatment for lung cancer.
|
From page 46... ...
2020b. The role of digital health technologies in drug development: Proceedings of a workshop.
|
From page 47... ...
2020. Impact of COVID-19 on oncology clinical trials.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.