3

STATEMENT OF THE PATIENT AND FAMILY LEADERS

To improve health care generally and their own health specifically, patients, families, and caregivers want to be able to access, contribute to, and share their health and personal data, particularly with respect to information and outcomes that are important to them (Haug, 2017). They also want transparency in all cases in which their data are aggregated, whether for public health or for commercial purposes (e.g., drug development). This is because the unauthorized access and/or disclosure of health data jeopardizes trust in the health system and could be used to discriminate against patients in insurance access or rates or employment (Kravitz and Allen, 2018; The New York Times Editorial Board, 2016). In addition, patients want access to research results that may directly impact their care.

The extent to which patient, family, and caregiver involvement in the sharing, linkage, and use of data can improve health and health care is not widely recognized by stakeholders in the health care system. The resulting underuse of data can hamper disease self-management, clinical decision making, and research involving health data.

This statement describes key barriers and solutions for leveraging and sharing health data from the perspective of the patient and family leaders working group (see Box 3-1).

CONCERNS AND BARRIERS

From the perspective of this stakeholder group, the key concerns and existing barriers to advancing data sharing, linkage, and use can be divided into four categories: (1) cultural barriers, (2) organizational barriers, (3) pragmatic and operational barriers, and (4) regulatory barriers.

CULTURAL BARRIERS

The organizations and individuals that generate, collect, and store health data (“data holders”) tend to lack awareness and understanding of the ways in which the participation of patients, families, and caregivers in health data collection can improve health care, despite growing research that points to the potential benefits of such involvement (Castle-Clarke and Imison, 2016; Toll et al., 2019). Electronic health record (EHR) data are largely encounter-based data, which are filtered and entered by clinicians. Contextual and personal information about patients is often missing; however, patients serve as an important source of data about their home and environment, their social and behavioral information that could be key to treatment (Zulman et al., 2016), and their own patient-reported outcomes (Jim et al., 2020). Providers can underestimate patients’ “invisible work” as well as patients’ expertise about their own health issues and self-care (Klasnja et al., 2010; Valdez et al., 2015). Without awareness of the myriad types of data patients can supply, stakeholders may not fully appreciate the many ways in which patients and caregivers could support efforts in care delivery and health improvement. This leads to underestimating how patient-generated data can assist with reconciliation of health data, enhance health record completeness and accuracy, and improve population health outcomes (Lavallee et al., 2020).

In addition, there is an imperfect understanding about individuals’ rights to their data under the Health Insurance Portability and Accountability Act (HIPAA), the federal law that puts forth national standards for the protection, use, and disclosure of an individual’s sensitive health information. Clinicians and health systems are not sufficiently responsive to patient, family, and caregiver requests for health data. They have been reluctant to support full clinical data transparency with patients or promote data-sharing systems, in part owing to concerns about additional workload, disrupted workflows, or a lack of understanding regarding legal requirements for sharing medical or mental health data. Clinicians and health systems generally do not seem to recognize and appreciate that patients, families, and caregivers are major users, managers, and contributors of data, even though this is the case in other industries (e.g., travel, banking, fitness, and web-based retail). Banking, retail, and travel have fundamentally altered how consumer data are used to facilitate smooth and easy transactions. Yet, health care, as an industry, has resisted a similar shift. Clinicians and health systems believe that health data are useful for informing clinician decision making but not necessarily for patient decision making and care self-management, despite evidence to the contrary (Cohen et al., 2016; Hixson et al., 2015).

These factors collectively contribute to a less-than-ideal culture of data sharing across the health care system, one that does not embrace patients, families, and caregivers as core members of that system (Topol, 2016). Some health care systems do recognize the value of patient-generated data and are integrating these data into their strategies and operations. Yet, barriers continue to exist, preventing the evolution of such initiatives and limiting their scale and dissemination. Some patients mistrust the health systems and other firms that collect data, and other patients often have difficulty accessing their own health data, which can undermine their willingness and ability to provide data.

ORGANIZATIONAL BARRIERS

Health information technology (IT) systems (e.g., EHRs) tend to focus on business processes and billing. They generally devote fewer resources to providing patients with convenient access to their data or to patients’ ability to directly provide data electronically. Other aspects of health IT systems, such as patient portals, are often not sufficiently designed to work well for patients (Archer et al., 2011; Farber et al., 2015; Irizarry et al., 2015; Ryan et al., 2016; Taha et al., 2013). Despite their name, “patient portals” only provide a fraction of the patient’s information that resides in EHRs. Some systems limit the types of data available to patients and families. Data may be available but

not conveniently accessible, which means that the patient often has to navigate several screens before being able to download his or her data. In general, health systems may underestimate the burden and degree of work required by patients to find, access, compile, update, and exchange health data, which negatively impacts patient experience and willingness to engage (Ancker et al., 2015; Gordon et al., 2018).

Clinicians often lack an established and routine workflow with which they can comprehensively capture all patient information, including patient-reported data (Cohen et al., 2016; Genes et al., 2018). Features of EHR systems also can contribute to clinician burnout, which can lead to negative effects for patients (e.g., resistance at the organizational level to implement additional technology solutions for patients) (Kroth et al., 2018).

Health IT vendors are typically incentivized to ensure the efficiency and effectiveness of their systems for the organizations using those systems rather than to consider the applicability of the systems and the data they contain for patients. Those vendors often are not incentivized to effectively enable data sharing with other vendor systems—creating technological silos. This is compounded by issues of data blocking by health systems or data holders, which have not been adequately addressed.

PRAGMATIC AND OPERATIONAL BARRIERS

Patients, families, and caregivers face many obstacles in providing health data. Ready- and easy-to-use web-based and mobile channels that solicit and allow patients to provide data that can be directly accessed by clinicians and integrated into clinical IT systems such as EHRs are lacking. The reconciliation of data represents an additional obstacle. Clear mechanisms that patients can use to suggest corrections to erroneous or outdated data collected about them by health systems, payers, or others are absent. Patients encounter cumbersome processes when requesting to update or correct their health data. Inaccurate information can lead to patient safety hazards and negatively impact a myriad of factors—for example, entities sometimes use patient data to generate risk scores, which dictate the pricing of services (Bell et al., 2017; Bourgeois et al., 2019; Landi, 2019; Ravindranath, 2019). Evidence shows that patients’ feedback and input can improve the accuracy of their health record data (Dullabh et al., 2014).

EHRs typically do not allow providers to catalog patients’ preferences about with whom they would like to share their information. In addition, patients generally are not able to confirm the accuracy of information in their medical records or provide additional information (Nyrop et al., 2019).

REGULATORY BARRIERS

A variety of regulatory barriers inhibit the sharing, linkage, and use of health data. Furthermore, these barriers are greater for some subpopulations, including adolescents, racial and ethnic minorities, and members of marginalized groups—many of whom struggle with challenges related to accessing and using devices and tools—even though some research suggests that these individuals have the most to gain from access to health data (Walker et al., 2019).

The legal requirements associated with data sharing often lack clarity. For example, regulatory policies and institutional review board decisions are inconsistent about the return of genomic research results to patients when tests are not conducted in a Clinical Laboratory Improvement Amendments–certified laboratory. Hospitals often are not in compliance with federal and state regulations regarding the release of information in medical records (Lye et al., 2018). Patients lack simple methods to complete identity proofing and provide permission for low-barrier, bidirectional data sharing.

The management of rare diseases provides a striking example of how regulatory barriers impede care. Patients with rare diseases are especially eager to contribute to research by donating data and tissue, and patient-partnered research is becoming more common (Kirwan et al., 2017). However, data generated on a specific patient are not easily linked across research studies, even when that patient is willing to consent to such linkage. Current regulatory policies, research practices, and EHR design all hinder patient-partnered research.

PRIORITIES AND ACTION STEPS

From the above description of concerns and barriers, the stakeholder group representing patients and families identified five priority issues that need to be addressed to facilitate widespread data sharing. When prioritizing the barriers, the working group members considered which barriers represented the most significant issues preventing widespread data sharing and linkage to improve patient care and which could be either wholly or partially addressed in the next 2 to 3 years. Some of these priorities combine several of the themes from the previous section.

1. Patients wanting but not being granted access to health data

All stakeholders in the health system should be aware of patients’, families’, and caregivers’ rights and desires to benefit from and be involved in the

provision, receipt, management, and use of health data. Though patients have not been seen as primary users of health data, evidence documents the benefits of supporting patient and family data-sharing preferences (Avram et al., 2018). Patient demand for data sharing is only likely to increase in the coming years because younger patients who have grown up in the digital era are especially likely to demand access to their data and will be propelled by the Centers for Medicare & Medicaid Services’ (CMS’s) Interoperability and Patient Access and Office of National Coordinator 21st Century Cures Act: Interoperability, Information Blocking, and the ONC Health IT Certification Program rules (Nelson, 2019) (see Chapter 5).

In addition, an action step achievable within 2 to 3 years is the development of a national campaign led by a consortium of advocacy organizations designed to increase awareness among data-holding entities such as health systems and EHR vendors about the benefits of bidirectional data exchange, which will de facto educate patients about the potential uses for data and the rights they have with regard to data sharing and access. This campaign should tie together different disease communities and target crosscutting groups like the National Health Council, PatientsLikeMe, or the National Patient Safety Foundation, along with representatives of health systems, insurers, clinicians, pharmaceutical companies, and other data holders.

Another action step is for OpenNotes, CARIN, and FasterCures to champion guidelines for how to engage patients, payers, caregivers, clinicians, hospital administrators, and compliance officers in seeing the value of transforming data into useful information to be used virtually at the point of care.

A longer-term goal is to pursue a “patients first” approach to data access. Such an approach would prioritize data transparency and reciprocity while recognizing that interest in and preferences for data sharing may vary among individual patients and among groups sharing particular conditions (Nelson, 2019). An action step would be to convene patient advocacy groups and networks representing chronic and acute conditions, pediatric and adult conditions, and rare and common conditions to develop statements on patient data access needs and preferences. The questions these statements could address include the following:

• What types of information do patients want to access, and what are the preferred methods of getting that access?
• What does bidirectional data exchange entail?
• What is the appropriate reimbursement structure for covering the costs of exchanging, reviewing, and using data for care delivery?
• How do data link to patient outcomes?

At the national level, the Office for Human Research Protections should develop guidelines that clarify the type of sharing permitted by HIPAA. These guidelines should be available in patient-friendly language and should be accessible through appropriate media.

In the area of mental health, policy makers and patient advocacy groups should work together to address policy barriers to data sharing. The Veterans Health Administration (VHA) shares mental health notes, and the VHA has created a course to help veterans understand how to access and understand mental health and other sensitive topic notes in their EHRs (Dobscha et al., 2016). This could serve as a model for various health care settings.

2. Insufficient appreciation among stakeholders that patients and families are key providers of patient-generated data

Evidence-based arguments for sharing health-related data, including the potential harms of not sharing data, could demonstrate to all stakeholders the value of data sharing. Case studies demonstrating these benefits should also be made available (Bell et al., 2017; Kreimer, 2015; Nelson, 2019; Schulte and Fry, 2019). Data sharing can be a component of patient safety, self-management, and quality improvement, contributing to a new standard of care. Evidence also shows the benefits to scientific studies of inviting patients to contribute their data for research purposes (CancerBase, 2020; Count Me In, 2020).

A systematic review of research could characterize patient preferences related to data sharing at the point of care as well as the benefits of data sharing for research purposes. Use cases from organizations that have facilitated bidirectional data exchange could include Geisinger, Dartmouth, the VHA, and Kaiser Permanente (Anderson et al., 2017; Bourgeois et al., 2019; Geisinger, 2018; Mafi et al., 2018; VA, 2018). Disseminating research could help address cultural barriers to widespread patient data access, especially if the link is made between data sharing and value-based care results.

Opportunities to contribute data to publicly funded research projects, particularly to projects related to outcomes and preferences such as the All of Us Research Program, should be a core feature of health data management nationally. This would provide a real-life context for patients to understand the power and the value of data aggregation and sharing. Many organizations could provide examples of the value of bidirectional data exchange to both care and research, including the epilepsy and the amyotrophic lateral sclerosis communities, the metastatic breast and prostate cancer communities, the

angiosarcoma and gastroesophageal cancer communities, and the hereditary cancer communities (BRCA Exchange, 2020; Broad Institute, 2020; Harvard Medical School, 2020). If commercial entities are provided with access to patient datasets, there should be a central fund aggregation pool that can put those commercial payments to use in funding public research and in defraying the costs to patients and families of participating in research. Perhaps, a portion of the fees collected from commercial entities by CMS’s Research Data Assistance Center could be re-invested into making data useful and eventually lowering the cost to access (ResDAC, 2020a).

3. Lack of understanding among stakeholders of how patient-generated data can improve workforce efficiency, enhance data coverage and accuracy, and yield better population health outcomes

An action step is to present certifying and specialty boards with data about the barriers to data sharing and engage them in developing tools that clinicians can use to determine how to inform patients about data-sharing opportunities. One successful example is the partnership between PatientsLikeMe and the American Academy of Neurology (AAN) to improve quality care guidelines for epilepsy and the training of future neurologists. This work resulted in the addition of two quality measures informed by patients and included in the “Epilepsy Update: Quality Measurement Set” approved by the AAN Quality and Safety subcommittee (Wicks and Fountain, 2012).

4. Financial disincentives to data sharing

Most health care systems and clinicians can only bill for time spent providing information to patients during appointments and not by communicating with patients through online portals, e-mail, text, or other digital media. Providers may be concerned about being overburdened by new workflows and technologies to support bidirectional data sharing, especially if EHRs are poorly designed. A commitment to develop a compensation strategy for asynchronous care and electronic communication could help allay these concerns. In the wake of the coronavirus disease 2019 (COVID-19) pandemic, insurers and federal agencies have relaxed regulations around virtual care. A specific action step is to create reimbursement codes for bidirectional communication among providers, staff overseeing clinical data systems, and patients. A related longer-term opportunity for modifying clinician and patient behavior is to change the reimbursement structure to one that pays for patient-centric value, outcomes,

and population health. Considerable work is ongoing in this area, but a majority of reimbursement remains fee-for-service.

Additionally, patient advocacy groups and organizations serving disease-specific communities could help reframe notions about the risks and benefits of data sharing, changing perceptions so that the risks associated with not sharing data are greater than those associated with sharing data. The most significant risks associated with not sharing data are misdiagnosis, late diagnosis, repeat tests, poor care coordination, and medical errors, with consequent effects on health, care quality, and costs to patients, payers, and the U.S. health care system. Advocates for data sharing could help demonstrate that these risks are greater than those related to the potential harms of data sharing.

Another action step is to develop a communication campaign targeted at health systems and physician professional societies highlighting evidence that data sharing in health settings decreases risks for clinicians and health organizations (Kreimer, 2015; Quinn et al., 2012; Tai-Seale et al., 2019). Organizations to be engaged in developing such a campaign include the Council of Medical Specialty Societies, the Institute for Patient- and Family-Centered Care, the Society to Improve Diagnosis in Medicine, and the Institute for Healthcare Improvement. Among the examples that could be cited are the Virginia Mason health system in Seattle, which has adopted a position of data transparency and full accountability for mistakes involving data, and the national Culture of Safety movement, which called for admitting to errors, figuring out root causes, addressing associated costs, and reducing immediate impacts (Kamo et al., 2019). At the same time, patient advocacy organizations could work with malpractice insurance companies to integrate expert patient participation into their work and to increase organizational knowledge of patients’ interest in bidirectional data exchange.

5. Lack of workflows and technologies that make it easy for providers to incorporate patient-generated and patient-held data

Apps, patient portals, and other tools, such as personal health records, could enable patients to easily gather and share their own data. Such tools could also help patients understand and shape the ways data are used; the ability to turn on and off access to different categories of data would increase trust and data accuracy (Kreimer, 2015). However, linkages between some apps and patient portals are such that each time a piece of information is added or changed in the EHR, the patient must manually initiate a new download of data to his or her app. The ability to “set and forget” automated data feeds between EHRs

and apps will be critical to making patient workflows easier. For example, a probable solution is one demonstrated by the Health Level Seven International Fast Healthcare Interoperability Resources Argonaut Project technical standard, which includes a provision for “persistent token” access. This provision enables ongoing connectivity without requiring a manual data request. CMS has implemented a policy of persistent token access for 1 year, unless revoked, in its Blue Button 2.0 application programming interface.

Design challenges in technology need to be addressed using a human-centered approach, especially with regard to transforming data into information that is useful for patients and providers. Efforts are under way at the Patient Family-Centered Care Partners—a group consisting of patient family advisors, health care administrators, clinicians, and social workers—to design a patient portal that facilitates increased communication among patients, families, and care providers.

In addition, new technology is needed to make data sharing between patients and providers easier and more beneficial for clinicians in a way that alleviates the burden of data collection for clinicians. For example, technology that would facilitate the comparison, selection, and exchange of data between a patient’s device and an EHR, much like exchanges between smartphones, could make it easier and faster for clinicians to integrate patient-held data, whether those data are generated by other clinicians and specialists, pharmacies, laboratories, or the patients themselves. (Genes et al., 2018; Nkoy et al., 2019). This also would facilitate the transfer of medical data when patients see new or multiple providers. CMS’s new Data at the Point of Care program enables this very scenario and could serve as a model for other entities (CMS, 2019a).

In partnership with patients, clinicians, health system leaders, HIPAA experts, and others, technology companies should continue to develop solutions that allow patients to access relevant information in EHRs, provide feedback on errors and completeness, and contribute their own health data. An action step would be for The Office of the National Coordinator for Health Information Technology or another entity to issue a challenge in which companies would compete to develop and present the most viable way to instantaneously upload and use patient data in EHRs and for data from the clinical record to be meaningfully visualized for patients’ use.

SOLUTIONS IMPACTING MULTIPLE PRIORITY BARRIERS

An action step that would address many of the barriers identified above is to measure and publicly report data on the level of access health care providers give

to patients, as well as the level of bidirectional exchange with patients and with other providers. These measures could be process or structural measures and could be reported on Hospital Compare, Physician Compare, and individual ratings sites managed by payers, accrediting bodies, private entities, and others (e.g., Consumer Reports, Checkbook).

The federal government will need to make an initial investment in determining appropriate measures. Measures should be co-designed with patient advocates.

RESPONSIBLE ORGANIZATIONS

Many groups will be responsible for implementing the action steps described above, including

• providers,
• patient advocates,
• researchers,
• health disparities advocates,
• malpractice lawyers,
• medical ethicists,
• health IT developers,
• payers,
• employers,
• HIPAA compliance officers,
• consumer advocacy organizations,
• informaticists,
• digital inclusion experts,
• data privacy experts,
• app developers, and
• policy makers and regulators.

Specific organizations to involve include the National Patient Safety Foundation, the National Health Council, the Institute for Patient- and Family-Centered Care, the Institute for Healthcare Improvement, PatientsLikeMe, FasterCures, and disease-specific organizations and communities.

This page intentionally left blank.