5

Inspiring Collective Action: Perspectives from Federal Stakeholders and Reflections from Individual Workshop Participants

The aspiration of achieving a national capability to monitor and assess MCM use is a shared responsibility among stakeholders, Lushniak observed, which starts with leadership and direction from within the federal system. In the final panel session of the workshop, federal government stakeholders reflected on the workshop discussions and shared their observations on the gaps, challenges, and opportunities around monitoring and assessing MCM use in the midst of PHEs. Panelists included Stacey Arnesen, chief of the Disaster Information Management Research Center in the Specialized Information Services of the National Library of Medicine (NLM) at NIH; Greg Burel; Redonna Chandler, deputy director of the Division of Clinical Innovation at the National Center for Advancing Translational Sciences (NCATS) at NIH; John Fleming, deputy assistant secretary for Health Technology Reform at ONC; Joseph Larsen, deputy director of the Division of CBRN Medical Countermeasures at BARDA; and Carmen Maher. The panel discussion was followed by an open discussion with workshop participants, facilitated by Lushniak.

Following the federal panel session, participants divided into breakout groups to discuss priority issues and potential actions for moving forward. Upon reconvening in plenary session at the close of the workshop, breakout group moderators Lushniak, Cobb, and Bakken reflected on the main points of discussion from their groups.

FEDERAL STAKEHOLDER PERSPECTIVES ON MOVING FORWARD

Biomedical Advanced Research and Development Authority

BARDA works toward filling gaps in preparedness and seeks to make investments that will be transformative in increasing the capacity to respond, Larsen said. In reflecting on the workshop, he observed that hurdles to collecting data on MCM use are more policy related than technological. Larsen recognized that there is a need to engage industry in discussions because medical product sponsors can provide valuable feedback on the design and conduct of post-marketing (Phase 4) studies. He also emphasized the importance of determining what type of information is absolutely necessary to collect and what information would be nice to have. Accountability is needed, said Larsen, and one agency (possibly CDC or FDA) should be designated as the lead agency responsible for collecting information. Larsen suggested piloting an MCM data collection system around an existing surveillance program, such as the routine monitoring of botulism cases in the United States. Botulism is going to continue to occur naturally, he said, and a pilot program for monitoring and assessing use of an approved MCM (heptavalent botulism antitoxin) for this indication could be launched.

Larsen also noted the need to determine costs associated with data collection, identify funding sources, and consider the sustainability of the enterprise. As an example of potential costs, he said that BARDA recently funded an expanded access protocol for the Ebola treatment ZMapp, including funding for data collection in two countries in West Africa and 11 sites in the United States. Capping the study at a maximum of 35 potential patients, it will cost $3 million per year for data collection, even if the capability is not used. This investment is not insignificant when it must be applied to 10 or 15 MCMs. Larsen noted that BARDA is already facing challenges in sustaining the investments it has made in companies and product development. Companies need to be assured of continued procurement of their product in the future, which drives commercial returns and enables them to continue to devote resources to MCM development.

Strategic National Stockpile

When discussing monitoring and assessing MCM use, the focus is generally on efficacy and safety data, Burel said. However, there is a spectrum of information needed throughout a PHE to inform the ongoing response and enable rapid changes to meet emerging needs. He noted, for example, that antiviral drugs were distributed to states during the 2009 H1N1 influenza pandemic, but there was no rapid feedback on what happened after the states received the products.

Various systems can be leveraged for monitoring the use of MCMs administered in a more traditional medical setting. When MCMs are administered in nonmedical settings, there is currently a dearth of any kind of useable data, Burel said. As discussed over the course of the workshop, however, data on MCM use might be found in non-traditional places, such as social media. Burel asked: Are there data to suggest people have confidence in the MCM? Are they receiving the product, and if so, what is the level of compliance? The data to answer such immediate questions about the response might already exist, he said.

Burel called for collaborative efforts among federal, state, and local authorities and industry to develop simple, easy-to-use reporting. For example, could a mobile app be developed that integrates into dispensing operations at a POD? Patients could enter follow-up information on their use of the MCM they received, providing valuable data on compliance and experience.

We do not know everything we need to know, Burel said, but there is a big block of data that we know we need. It is important to start now, and not wait until the day of an event to decide what questions to ask.

National Center for Advancing Translational Sciences

Part of the mission of NCATS is developing innovations that aid health research of all types, particularly tools that speed up the process of clinical research and data capture, Chandler said.¹ NCATS is an excellent resource for the clinical networks aspect of MCM monitoring and assessment efforts, she said. For example, the NCATS Clinical and Translational Science Awards (CTSA) Program supports a network of more than 50 academic medical centers across the country. CTSA facilitates collaboration, allowing researchers across institutions to come together as a clinical research enterprise. NCATS is currently working to create interoperable data warehouses across CTSA program awardees, which Chandler said could be invaluable for addressing many different types of research questions. In addition, the data could help to facilitate rapid recruitment for clinical studies.

All CTSA program sites have signed an agreement to streamline IRB review for multisite studies by relying on a single IRB.² The concept was tested in a pilot study and used by the Harvard-affiliated CTSA programs following the Boston Marathon bombing to rapidly initiate a clinical study, Chandler said. A group of investigators was interested in studying hearing impairment and repair in individuals who were at the exact site of the blast and in geographical circles around the blast. The ethical review process for

___________________

¹ For more information, see https://ncats.nih.gov/index.php (accessed August 23, 2017).

² See https://ncats.nih.gov/expertise/clinical/smartirb (accessed August 23, 2017).

the multisite study, which might normally take months (or longer), was completed in a matter of days because the sites had previously formed a network and agreed to a single IRB review. This exemplified a point that individual participants raised often during the workshop: there are certain research questions that can only be answered in the context of a PHE.

Chandler noted that there is a tension in a PHE between the government responders and local community. Another valuable aspect of the CTSA network, she said, is that NCATS has existing relationships with the local centers of excellence it funds and an understanding of the resources they have available, the service delivery system in their area, and the needs of their patient population.

Regarding monitoring and assessment of MCM use, Chandler said there is always room for more collaboration and engagement and suggested looking at what capacity across NCATS might be applied during a PHE. For example, NCATS intramural researchers have significant expertise in subjects such as screening libraries of approved compounds for repurposing existing drugs.

Office of the National Coordinator for Health Information Technology

One area of focus for ONC is information transfer, particularly in emergencies, Fleming said. One of the lessons learned from the response to Hurricane Katrina was the need for a system that can retrieve and share patient information from a range of sources during a disaster response. Residents of nursing homes and assisted living, for example, were sent to care facilities in Mississippi, Texas, and elsewhere, and the nurses and doctors receiving these patients were without information regarding medications or diagnoses for these patients. To address this unmet need, ONC is currently pilot testing the Patient Unified Lookup System for Emergencies, which provides connectivity with health information exchange systems.

The United States is lagging behind the rest of the world when it comes to interoperability of health care data systems, Fleming noted. EHR includes structured data and unstructured data; structured data are more interoperable than unstructured data (such as imaging files and free text). ONC is focused on improving the interoperability of data systems without compromising privacy and security, which requires vigilance when developing policies and standards to ensure that records cannot be hacked or misused, he said.

From a funding perspective, ONC is working on implementing rules and regulations for the 21st Century Cures Act.³ Fleming explained that

___________________

³ See https://www.congress.gov/114/plaws/publ255/PLAW-114publ255.pdf (accessed October 23, 2017).

ONC serves a coordination function that includes coordinating distribution of funding; establishing standards, quality measurement, and measurement of interoperability; and overcoming deliberate information blocking.

National Library of Medicine

Arnesen reiterated an earlier comment that the field of data science requires a broad range of skill sets and specific health subject-matter expertise (see Data Science Training in Chapter 3). She suggested that the NLM Disaster Information Management Research Center could help to address that need. Some themes that consistently emerge in discussions of disaster preparedness and response are data, information, and communication, she said. Information and communication are core components of library sciences, Arnesen added, and data science will feature significantly in the forthcoming NLM strategic plan.

NLM has evolved from simply collecting information, to organizing and indexing the information, to discovering what can be done with data and considering issues around data sharing, open data, data analysis tools, and other areas. NLM has a lot to offer in the area of “data preparedness,” Arnesen added, and in helping to develop tools for response operations that can be ready-to-go for data collection and data analysis.

Although NLM is often not thought of when discussing response operations and MCMs, it has been engaged, for example, in working with ASPR on Radiation Emergency Medical Management and Chemical Hazards Emergency Medical Management toolsets, which provide information on MCMs. Arnesen noted that she looks forward to increased engagement of NLM by those working on MCM monitoring and assessment.

Discussion

In the discussion that followed the federal stakeholder perspectives, the federal panelists further discussed the data needs and resources available to decision makers, as well as the funding, costs, and interoperability of the data systems proposed throughout the workshop. John Tegeris of BARDA noted that the task at hand is not to design a theoretical solution and put it on the shelf. It must be built, exercised, and refined, he said. He suggested incorporating data collection infrastructure activities into a study of a product in development, thereby accomplishing several goals across agencies, industry, and other stakeholders.

Operational Questions, Data Needs, and Collection Methodologies

FDA, BARDA, and SNS are federal decision makers in need of data and information on MCMs, Lushniak said, and NCATS, ONC, and NLM are in positions to help provide access to that data. He prompted Maher, Burel, and Larsen, of FDA, SNS, and BARDA, respectively, to identify questions they need answered with regard to MCM use. Key questions for CDC and SNS, Burel said, are whether and how people are making use of an MCM that has been provided to them and whether the MCM is having the immediate desired effect.

Some MCMs are approved and entered into the SNS based on limited human effectiveness data, Maher added, and a PHE is likely the only time the MCM will be used in its intended population. Therefore, it is important to gather requisite data to quickly assess whether the performance expectations extrapolated from preclinical studies are, in fact, correlated with how the MCM is performing in the real world. Decisions can then be made by state and local authorities regarding whether to continue use of the MCM or to stop deployment, she said. These data are also needed to enable regulatory decision making, said Maher. For products administered under an EUA, for example, data are needed to ultimately support FDA regulatory review and approval, said Larsen. For an approved product, data are needed to provide greater confidence that the product is safe and effective.

Maher raised several questions regarding leveraging existing tools and systems for data collection. Who owns each of these systems or tools? How can independently owned tools be coordinated? Who coordinates them? Who are the decision makers, and what information do those decision makers need from these systems? How is that information normally obtained, and are there data gaps created by using these existing systems? What might need to be built into the existing systems? How is the information prioritized and communicated back to stakeholders?

Leveraging Existing Resources to Supply Data Needs

Lushniak asked Chandler, Fleming, and Arnesen what capabilities NCATS, ONC, and NLM, respectively, might have to fulfill the needs of MCM decision makers. Fleming observed that an “information highway” is needed—an infrastructure for MCM data. He noted that it is not the job of ONC to build that highway, but to see that it is built through ONC’s coordination function. What is most essential, he said, is standardizing information transfer and protecting privacy and security so that information can get from point A to point B for the purposes of providing care and conducting research.

NLM is partially responsible for helping build that information high-

way and considering what data are collected and what data are stored, Arnesen said. She mentioned the FAIR principles of data management, which emphasize that scientific data should be findable, accessible, interoperable, and reusable (i.e., collected once and used multiple times, by different researchers). NLM is looking across the spectrum of research as it pertains to storing, accessing, and sharing data, and Arnesen suggested that MCM monitoring and assessment could be used as a pilot or test case. NLM is also working closely with ONC in developing standards both for interoperability and in establishing common ontologies, she added, and the development of standards could be an area in which MCMs could be used as a test case.

Chandler elaborated on the CTSA Program at NCATS as a resource for the integration of clinical care delivery and research and the collection of data. CTSA researchers understand both data collection and the delivery of clinical care, she said. Structurally, the CTSA Program provides funding to academic medical centers to serve as hubs within their localities; they are then networked to hospitals, federally qualified health centers, and other points-of-care delivery locations. NCATS is beginning to work on data interoperability across its academic medical centers hubs and ultimately plans to incorporate all of their network organizations. She explained that these institutions provide different types of care and thus have the potential to be involved in the collection of different types of data.

Entrepreneurship exists across the CTSA Program, Chandler said. CTSA awardees, in partnership with industry, bring medical products through development, regulatory approval, and post-approval monitoring, and then feed data and information back in support of a learning health system. Bakken added that each CTSA Program hub has extensive resources for engaging the local communities, and she suggested that these resources could be brought to bear for MCM monitoring and assessment.

Funding and Costs

Cooper and Lushniak prompted the federal panelists to consider how the changes and strategies being discussed could realistically be implemented in the context of a rapidly changing fiscal environment, which includes threatened cuts to public health funding. Burel noted that funding is a persistent issue, and discussions of developing critical monitoring and assessment capability do not become theoretical simply because of declining budgetary resources. There is a need to think innovatively about funding, he said, instead of relying solely on federally appropriated funds. For example, Burel said, there could be cooperative funding among states to develop a tool to meet broad stakeholder needs, which would also leverage foundation support.

Burel pointed out that it is not yet known what the actual costs of an MCM monitoring and assessment system will be, which is essential information when discussing funding. He suggested defining the essential data elements, data that would be nice to have but that are less essential, and information that would be difficult to obtain. After this exercise, one could identify existing sources that might contain this information, potentially looking beyond the health care system. He suggested engaging the National Academy of Public Administration to better understand what other departments, agencies, states, and localities have done to collect information and perhaps identifying how to leverage those resources. Fleming noted the potential of leveraging IT to lower costs and deliver more. Other industries, particularly in the private sector, have lowered costs by using IT, he observed.

Maher reiterated the importance of defining ownership or coordinating responsibility for the tools and solutions. Who is empowered to assemble the different pieces from where funding already exists? Additional funding would be ideal, but potential solutions exist that can be adopted in the absence of substantial additional funding. Maher highlighted the need for public health to become much more efficient at bringing the PHEMCE perspective to stakeholder activities that are already being funded. For example, how might MCM data collection be incorporated as part of the recent implementation of a new EHR system at the U.S. Department of Veterans Affairs? In other words, show other stakeholders how this problem is also their problem, and how helping to solve it serves them as well, she said.

PRIORITIZING ISSUES FOR ACTION: REFLECTIONS FROM INDIVIDUAL WORKSHOP PARTICIPANTS

Following the final panel session, workshop speakers and attendees were invited to discuss potential strategies and actions in the areas of CONOPS, clinical networks, and sources of data (both electronic health data and big data). At the end of the workshop, Lushniak, Cobb, and Bakken shared their individual reflections on the discussions held in these breakout sessions and throughout the workshop.

Concept of Operations⁴

In the context of CONOPS, Lushniak shared his observations on what could be done to make the system ready to determine the safety and effectiveness of a given MCM when the next PHE happens, as well as potential solutions and action steps. He likened CONOPS to a machine with a set of levers that one must pull to instruct the machine what to do. Similarly,

___________________

⁴ All comments in this section are directly attributable to Lushniak.

determinations on the type of CONOPS needed for monitoring and assessment of MCMs during PHEs involves an understanding of key parameters that will define the scope of the problem, including the scale of the PHE, the type and regulatory status of the MCM, and how and in what locations the MCM will be distributed.

Lushniak noted there could be resistance or opposition by certain stakeholders to operational aspects of MCM monitoring and assessment. For example, state and local partners may resist some CONOPS activities because they may already be overburdened and have limited or insufficient resources (e.g., time, staffing, funding, or supplies) to participate in these operations. There may also be concern raised from groups on specific issues (e.g., opposition to vaccinations or data privacy advocates).

A primary challenge to CONOPS for MCM monitoring and assessment is funding, said Lushniak. Other challenges, he observed, include politics and political will; personnel and workload issues; siloing of data or lack of collaboration; communication; education and training; bureaucracy; scalability and flexibility of plans; differing missions, priorities, and authorities across stakeholders; technology; and exercising and evaluation of CONOPS.

There are present opportunities for CONOPS in MCM monitoring and assessment, said Lushniak, including the potential to validate regulatory approaches for regulating medical products (e.g., the Animal Rule); collecting data and information to inform, refine approaches, and set priorities for PHE responses; standardizing information; potentially decreasing the risk of products if positive safety and efficacy data were collected; demonstrating return on investment; building systems for future scenarios; and saving lives.

Lushniak described steps that could be undertaken within the coming year related to CONOPS for MCM monitoring and assessment, including

• Explore the use of clinical trial networks for MCM monitoring and assessment and share lessons learned (potentially as part of a BARDA tabletop exercise);
• Understand and define the data needed for monitoring and assessing MCMs, which could be informed by exploring a specific MCM as a use case (potentially as an SNS workshop);
• Develop communications strategies to illustrate the challenges and opportunities for MCM CONOPS (potentially at upcoming forums such as the Health 2.0 conference);
• Leverage industry resources and expertise;
• Conduct a landscape evaluation of existing systems at the state and local levels (potentially working through the National Association of County and City Health Officials); and

• Build or expand partnerships with stakeholders that are often underrepresented or not represented at workshops such as this one (e.g., CMS, private payers, or IRBs).

Clinical Networks⁵

Cobb shared his observations on how an extensible clinical system for monitoring and assessing MCMs that is rapid, adaptable, scientifically robust, and minimally disruptive to the course of care during a PHE could be developed. Cobb noted that considerations for realizing such a system could include

• The developmental stage of the MCM and level of evidence available (e.g., investigational product, approved under Animal Rule, or FDA approved);
• The type of research to be conducted (e.g., interventional or observational; acute, real-time data collection or ongoing study of long-term outcomes);
• The nature and scale of the PHE (e.g., national versus local; infectious disease versus natural disaster); and
• The class of MCM (e.g., drug, biologic, device, or combination product).

There are several strengths of the current system and potential opportunities to consider when developing a clinical trial network for MCM monitoring and assessment, said Cobb. From a resource standpoint, there is an existing clinical trial infrastructure that could be leveraged; expertise that could be brought to bear; new innovations on the horizon; a movement toward public–private partnerships; and global experience with past PHEs, domestically and globally, that can inform the process. A variety of opportunities are available, including case studies from past PHEs (e.g., Ebola); testing protocols in benign, predictable threat situations (e.g., seasonal influenza); socializing the issue to engage and educate the public; and the opportunity to conduct annual exercises.

Cobb also highlighted several weaknesses of the current system and threats that could affect progress, including a lack of leadership or ownership; lack of commercial models or incentives; fragmentation of the health care system; funding cuts and uncertain sustainability of existing structures; the risk of unknown public health threats; and dissemination of misinformation.

___________________

⁵ All comments in this section are directly attributable to Cobb.

Several potential solutions and next steps could be undertaken within the next year to advance a clinical trial of networks for MCM monitoring and assessment, noted Cobb:

• Draft a playbook of potential scenarios and a framework for evaluating them against available data sources.
• Begin annual exercises and tabletops to test the feasibility of pilot efforts and provide proof of concept of a “network of networks” approach.
• Identify and address leadership gaps, including designating an organization or agency to be tasked with the authority to create or leverage clinical trial networks to conduct research during PHEs.
• Build private–public partnerships and mechanisms for conversation among stakeholders, including product sponsors.
• Create an advisory board for administrative preparedness that would address issues around human subjects protection in public health practice versus research.
• Leverage existing surveillance systems.

Data Sources⁶

Bakken presented her perspective on how existing data sources, systems, and technical and human infrastructures can be leveraged to meet the needs of MCM monitoring and assessment for safety and effectiveness. She noted some key parameters, stakeholders, context, and potential solutions and action steps that could be considered.

One could consider this topic from the perspective of a broad set of stakeholders and their needs, Bakken said, and indeed some stakeholders are both data suppliers and data users. A vast range of stakeholders collect and use data that could potentially be leveraged for MCM monitoring and assessment, said Bakken, and the particular stakeholders involved in any one effort are driven by the scenario. There is a particular need for representation for special or vulnerable populations, whose data are not always captured, as well as those still facing technology barriers to inclusion in research, Bakken added. For example, certain areas of the country, including tribal areas, continue to lack broadband Internet access, which is a hindrance to capturing or delivering data.

Bakken noted that challenges to adequately leveraging data in monitoring and assessing of MCMs could be categorized as technical or nontechnical. She highlighted a lack of adequate standards for semantic operability across data systems as a main technical challenge. Other techni-

___________________

⁶ All comments in this section are directly attributable to Bakken.

cal challenges include text mining; terminology; timeliness of data delivery; time to enact a change of data system vendor; the volume of available data; governance of data sharing; time needed to set up data use agreements and renewals; training of data collectors; pockets of data that are not being collected; and configuration of databases.

Other primary, non-technical challenges, she added, are workforce development in data science and the lack of a prospective plan to test the readiness of data collection and transfer processes. The volume of available data presents a challenge, but it is also an opportunity, said Bakken, though there are certain populations that are not presently included in datasets, including some international and tribal populations.

Based on the workshop discussions, Bakken listed several potential opportunities, solutions, and next steps that could be leveraged, built upon, or undertaken within the next year:

• Data preparedness framework and testing plan—Consider which data sources, and what level of data veracity, are needed for which purposes.
• Data sharing governance—As part of a data preparedness plan, establish priority data use agreements with identified partners, such as pharmacy benefit managers or insurance systems, and exercise the agreements ahead of time to establish relationships and build trust.
• Data standardization—Identify and address gaps in current standards and develop a common ontology as they relate to data collection for MCMs.
• Data storage and use—Encourage continued refinement of models for mapping datasets and maintenance data coordinating centers.
• Commercial pharmacies to fill unmet data needs—For example, consider placing “pop-up pharmacies,” which have existing data collections capabilities, in underserved areas.
• Patient- and consumer-generated data collection—Take advantage of non-traditional data sources such as wearables, self-reported data, social media information, and search engines.
• Data science—Leverage the current heightened interest in data science as an opportunity to build a data science workforce. Look to existing bodies of knowledge, such as the U.S. Digital Service, for support in this area. Promote financial incentives, including student loan forgiveness, for health care workers.

CLOSING REMARKS

Lushniak closed the workshop by thanking the various groups that made it a success: the planning committee, for volunteering their time for several months to plan and convene the workshop; the workshop speakers, who volunteered their knowledge and experience; the workshop sponsor, FDA, for bringing to light this important topic; the workshop participants for taking part in the discussions; National Academies staff; and workshop facilitator Laura Runnels.

Many potential solutions were discussed throughout the workshop, said Lushniak, including a wide array of challenges and potential solutions and next steps. He added that when it comes to MCMs, the health and safety of our nation, and the national security of our country, the work does not stop here.

This page intentionally left blank.