HTI-1 Proposed Rule Public Comments

Jun 22, 2023

National Coordinator Micky Tripathi
Office of the National Coordinator for Health Information Technology
Mary E. Switzer Building
330 C. St SW, 7th Floor
Washington, DC 20024

Re: RIN 0955–AA03; Health Data, Technology, and Interoperability: Certification Program Updates, Algorithm Transparency, and Information Sharing

June 20, 2023

Dear Dr. Tripathi:

As an individual, board-certified patient advocate, the primary carepartner to two disabled adults, and a patient myself, I appreciate the opportunity to provide comments on the Health Data, Technology, and Interoperability Certification Program Updates, Algorithm Transparency, and Information Sharing Proposed Rule (HTI-1).

As a patient advocate for over 25 years, each day I work with patients and their families as they navigate life-altering, life-limiting diagnoses, such as cancer, while navigating our complex healthcare system. Each day, I connect patients and their loved ones to the information, tools, and technology they need to make informed decisions about their care.

Most of my work is concentrated in helping patients gain access to and understand the power of their health information and how to utilize that power to navigate barriers, inequities in access to care, and fragmentation in coordination of care.

Over the years, I’ve had the privilege to serve as a member of the HITAC United States Core Data for Interoperability (USCDI) Task Force, the Interoperability Standards Workgroup, as an invited subject matter expert to the HTI-1 HITAC Proposed Rule Task Force, and a contributor to various ONC efforts. I’ve had the opportunity to amplify the patient and carepartner perspective in collaborations with like of the American Medical Association (AMA), the American Information Management Association (AHIMA), SHIFT, HIMSS Public Policy Committee, the Sequoia Project Consumer Voices Project, HLTH Foundation Techquity Coalition, HL7 Patient Empowerment Workgroup, the Society for Participatory Medicine, Digital Medicine Society (DiMe), and most recently, as an advisor on financial toxicity and equitable access to care for CancerX. I’m grateful for the opportunity to provide comments from the intersection of these experiences and my daily patient advocacy work.

Key Themes:

From the patient and carepartner perspective, there are 4 key themes I’d like to highlight:

1) First and foremost, the future of our information sharing ecosystem and health IT policy framework must recognize patients, their carepartners, caregivers, authorized advocates, and consumers as end users of certified health IT by way of patient portals, apps, and APIs.

2) It is paramount to not only advance the exchange of accurate, comprehensive patient health information to best support inclusive, compassionate, safe, trusted, and affordable care but to do so while striving to protect patients and our care providers from prosecution and criminalization for practicing evidence-based medicine. The sanctity of the doctor-patient relationship must be protected from efforts to weaponize patient health information, especially in the post-Dobbs landscape.

3) I encourage ONC and all healthcare stakeholders to recognize and prioritize the importance of co-creating the future of health IT and predictive decision support interventions (DSIs) in collaboration with the patient stakeholder.

4) As we dream bigger about the future of information sharing and AI-powered clinical decision support, we must recognize the importance of patients having access to the recommendations (outputs) of predictive DSIs, especially when the recommendations (outputs) are used to inform decisions about an individual’s care or coordination of care. When DSI outputs are utilized to make decisions about an individual’s care or coordination of care, such predictive DSI outputs should be recognized as EHI. Patients need significantly more transparency about the role that AI-powered clinical decision supports may, or may not, play in their care. Currently, patients have no way of knowing if a predictive DSI has been utilized to make decisions about their care or coordination of care.

Comments on USCDI v3

I support ONC’s proposal to adopt USCDI v3 to help capture more accurate and complete patient characteristics that are inclusive of individual and community diversity as well as to help identify and address disparities. ONC should collaborate with key stakeholders to advance granular data segmentation functionalities to prevent the weaponization of data, i.e., pregnancy status, and prevent the criminalization of practicing evidence-based medicine.

From the patient and carepartner perspective and as someone who has the unique privilege of frequently working in end-of-life care and emergency circumstances, I strongly recommend ONC consider including Advance Directive documentation, with an immediate priority focused on establishing an on-ramp for access to currently available, unstructured Advance Directive documents (i.e., PDFs and scanned images).

At the April 12, 2023 meeting, the HITAC approved recommendations to the National Coordinator developed by the Interoperability Standards Workgroup on USCDI Draft Version 4, containing recommendations on Advance Directives. As per the transmittal letter offering the final report from the HITAC with recommendations submitted:

“End-of-life care information is critical to patient care to prevent harmful, unnecessary medical treatment against an individual’s specified wishes. Inclusion of Advance Directive documents meets USCDI draft v4 prioritization criteria of: (1) mitigated health and health ONC HITAC Recommendations on Draft USCDI Version 4 - April 12, 2023 care inequities and disparities; (2) address the needs of underserved communities; and (3) represents important additions over previous USCDI versions. The COVID-19 pandemic taught us that a lack of Advance Directives may cause suffering and inequity when making critical end-of-life care decisions.”

The Joint Commission also highlighted the importance of including Advance Directives in their submitted comments.

From the patient and carepartner perspective, it would also be helpful to explore a modular approach for USCDI, as suggested in the commentary from Apple, where data elements may be grouped to better support specific use cases, i.e., appealing insurance denials, prior authorization, demonstrating medical necessity for Social Security disability benefits. This is would reduce patient and physician administrative burden, improve access to timely care, and reduce costs.

Comments on Predictive DSIs:

Page 23776 of HTI-1 states, “Within healthcare today, predictive models are increasingly being used to inform an array of key decision-makers, including clinicians, payers, researchers, and individuals, and to aid decision-making through CDS.”; I respectfully disagree that predictive models are increasingly being used to directly better inform individuals, like patients, their carepartners, caregivers, and their authorized advocates. There is a stark asymmetry in access to predictive models and insights directly to patients. Patients desire greater access to not only predictive DSIs as part of their care, but also access to information about DSIs and DSI outputs.

Patients face an asymmetry of availability of and access to fair, appropriate, valid, effective, and safe (FAVES) information. It is important to note that clinicians are not the only end-users who need access to this information for informed and shared decision-making purposes and for building trust and transparency. Imagine a world where patients and consumers may select their care based on the availability of an offered DSI at a provider’s office, hospital, or institution.

Each component of FAVES information should be:

inclusively defined from the patient perspective through co-creative processes
readily available to patients as end-receivers AND end-users of DSIs

End users of predictive DSIs, including patients, their carepartners, caregivers, and their authorized advocates, must have access to source attributes in the public domain. The patient voice should be inclusively represented when consensus decisions are made about what source attributes are selected as a priority.

I applaud ONC for revisiting that source attribute information is not routinely available to users at point of care, and for proposing requiring developers of certified health IT to make available the source attributes information referenced via drill display, drill down, or link out from their modules.

Information about which datasets have been used to train and evaluate predictive DSIs should be made readily available in the public domain. Details about training and evaluation datasets can provide valuable information when considering whether an algorithm is fit for purpose. Including information about which datasets have been used for training and evaluation of predictive DSIs may reduce the risk of bias and harm by way of biased variables that may have been included.

Since errors may occur at any point in the DSI lifecycle, it is critical that the patient stakeholder voice is represented to reduce the risk of bias and model risk.

In thinking about model risk and the potential of inaccurate outputs, how can patients determine if predictive DSIs may have negatively impacted their care, especially when patients currently do not have access to DSI recommendations (outputs)? On a similar note, the use of DSI recommendations at point of care are currently not documented in our electronic health record.

Pages 23781-23782 of HTI-1 state, “We believe that these additional source attributes would better support the transparency of predictive DSIs and that such information is necessary for users to decide whether and how to use the predictive DSI, including whether to apply the predictive DSI to individual patients.“; what if a patient does not want a predictive DSI applied to their care? Currently, from the patient and carepartner perspective, there is no transparency about what the financial impact is, whether positive, negative, no impact, or currently unknown, of applying a predictive DSI to an individual patient’s care. Alternatively, what if based on source attributes and other information, an available predictive DSI is not applied to an individual’s care? What if a patient wishes to appeal that decision?

From the patient and carepartner perspective, there is currently no transparency around this decision-making process. I encourage ONC to ensure that these workflows and decision-making processes are better documented for greater transparency and shared decision-making. Stronger documentation could be achieved by way of documentation in clinical notes or through the creation of a USCDI DSI Recommendations data class for sharing and exchange of predictive DSI recommendations (outputs) to support patient access to their health information and informed decision-making.

I applaud ONC proposing to adopt a new functionality that would require a Health IT Module to be able to record at least one action taken, and by whom it was taken, when a CDS intervention is provided to a user.

I applaud ONC for proposing that a Health IT Module be able to generate a human-readable display or report of the responses and actions taken and by whom.

Page 23791 of HTI-1 states, “Many such approaches exist; however, there is no universal best process to ensure fairness.“; I highly encourage ONC and key stakeholders to consider Patients Included best practices to ensure the patient stakeholder voice is a driving force in matters of fairness and in the co-creation of the future of predictive DSIs.

I applaud ONC for proposing a new functionality to enable users to provide electronic feedback data based on the information displayed through the DSI and to have the ability to export such feedback data. These feedback loops are important from the patient and carepartner perspective.The patient voice should be diversely represented during the consensus process for identifying the minimum proposed feedback data for export.

Comments on Patient Requested Restrictions:

I applaud ONC’s commitment to believing that individuals should be provided a reasonable opportunity and technical capability to make informed decisions about the collection, use, and disclosure of their EHI.

From the patient and carepartner perspective, it is disappointing to routinely see pushbacks on necessary advancements to interoperability and achieving more granular, segmented privacy controls in comments submitted from various vendors and stakeholders intimately involved with advancing interoperability. From my experience, I can appreciate the challenges associated with robust standards development, real-world testing, complex implementation requirements, and staffing shortages. We should not lose sight of patient unmet needs, the negative impacts on patient outcomes, and patient safety.

For example, there are concerns about the timeframe associated with the Patient Requested Restrictions Certification Criterion. Meeting aggressive timelines will need aggressive operationalization through public-private-patient partnerships. Advancing the future of health IT can’t continue to be hindered by our acceptance of the speed of the status quo. The pandemic showed us the power of public-private partnerships to expedite the advancement of, for example, the deployment of telehealth and virtual care as well as vaccine development. How can we approach standards development, real-world testing, and implementation more innovatively and collaboratively to accelerate traditional timelines? Are there opportunities for accelerators and connect-a-thons to catalyze progress?

Certain portions of HTI-1, i.e., Patient Requested Restrictions, cannot continue to wait for years before we attain successful implementation, especially considering the active weaponization of health data and criminalization of the practice of evidence-based medicine post-Dobbs in certain regions of the country. This is a matter of patient and provider safety and health equity. Pushbacks and requests for implementation delays will likely mean delays in essential care, lives lost, inequities perpetuated, and continued harm, suffering, and burdens unnecessarily imposed on individuals receiving care and their care providers. I ask ONC to leverage its authority and leverage the power of multi-stakeholder partnerships, including partnerships with patients, to expedite advancing technical and functional capabilities to support doctors and patients with the tools and technology to protect sensitive information. I support the American Academy of Family Physicians (AAFPs) comments urging “HHS to work with health IT developers and health data management platforms to advance meaningful data segmentation capabilities. Given the lack of industry progress in this area, the AAFP urges ONC to examine how it can spur action to respond to growing threats to patient privacy, the patient-physician relationship, and patient and clinician safety.”

Patients want control over more than just specific data elements and classes; we want accountability, transparency, and control over how our data is used and for what purposes. Patient communities have expressed they are more willing to share their health information with their doctor and care team as it pertains to their care and are less comfortable sharing their health information with employers, big tech, life insurance companies, social media, data aggregators, and annual medical reporting companies.Prioritizing Patient Requested Restrictions to enable restrictions that align with real-world patient concerns and unmet needs, such as those highlighted in the most recent patient survey, Patient Perspectives Around Data Privacy , conducted by the American Medical Association (AMA) in collaboration with Savvy Cooperative.

Individuals must be able to trust that the health information generated during the sanctity of the doctor-patient relationship is protected from being weaponized and free from contributing to discrimination and exacerbating health inequities.

Comments on Prior Authorization

Patients and their families are overwhelmed with administrative burdens, including prior authorization requests and appeals. It’s well documented that prior authorizations frequently drive significant delays in patient access to care and are known contributors to physician burnout. I support the recommendations of the AAFP urging ONC to include electronic prior authorization functionalities in CEHRT as soon as possible to alleviate the challenges associated with prior authorizations required by payers for prescription medications and other patient care.

Comments on Manner Exception

ONC is proposing to add a TEFCA condition to the information-blocking manner exception. If an EHI request is fulfilled by TEFCA, even if a requestor preferred another means, if the EHI that is provided to the requestor is incomplete, the requestor, especially if it is a patient or authorized representative, should be able to report the incomplete EHI provided and escalate the initial request for expedited fulfillment using a non-TEFCA preferred means.

Closing Remarks:

I applaud and support the efforts ONC has taken through NPRM HTI-1 to improve transparency, enhance trustworthiness, support consistency, and advance health equity by design.

The team at ONC has done an exceptional job of addressing a robust spectrum of industry-wide concerns that exist in our information-sharing and interoperability arena within this NPRM. The team at ONC should be commended for the number of public-facing resources that have been provided leading up to and throughout the public comment period, including blogs, webinars, fact sheets, a Word version of the NPRM, and a public comment template.

From the patient and carepartner perspective, I applaud ONC’s role in advancing transparent and trustworthy predictive technology in health care as well as transparent risk management for and governance of predictive DSIs.

The success of predictive models and machine learning algorithms in health care must be rooted in inclusive co-creation and symmetry in access to information for ALL stakeholders.

I highly encourage that matters addressed within HTI-1 proactively, continuously, and longitudinally include diverse patient perspectives and expertise. The Sequoia Project Consumers Voices Project is a strong example of a Patients Included model that demonstrates a reproducible model and foundation for ensuring equitable and diverse patient and carepartner voice representation.

Other related questions not covered include:

How are errors currently in patient medical records going to be addressed? Will there be a greater emphasis on correcting errors in medical records to improve data quality of data powering DSIs?
Will there be a patient/consumer DSI complaint portal for reporting concerns, bias, discrimination, and harm?

As an industry, we are cautious so as to not overburden vendors, developers, and health systems. We are mindful not to further overburden our doctors, clinicians, and providers who are already experiencing extreme stress and duress on this side of the pandemic.

What we don’t talk about enough are the realities that patients and their families face daily. Patient burnout is a silent public health crisis. Reducing Patient Administrative Burden (PAB), or the work that individuals need to do to get the care they need, also needs to be carefully considered in this NPRM and the future of our health IT.

Our current information-sharing landscape has propelled patients past being satisfied in simply being able to “see” certain parts of our health information. We want access for action’s sake.

Powering interactivity with our health data to make informed decisions in our care is of the essence and can be realized through HTI-1. The time for patients to help co-create the future of health IT and AI-powered DSI is now.

I thank and applaud the ONC’s continued commitment to achieving the full implementation of the 21st Century Cures Act. I look forward to continued collaborations and continued efforts to support greater inclusion of diverse patient and carepartner voices and expertise as we co-create and navigate the future of health IT and health IT policy together.

Respectfully,

Grace Cordovano

Grace Cordovano, PhD, BCPA
Founder, Enlightening Results
Co-founder, Unblock Health

HTI-1 Proposed Rule Public Comments

Related Article

The Five Key Points Missing from the Majority of Digital Health Transformation Strategies