Highlights of Main Points Made by Individual Speakers1
- The health information technology aspects of health care delivery reform can strengthen public health preparedness, response, and recovery.
- Data can enable a broader understanding of community needs that can help to improve the structure and delivery of both day-to-day and emergency care.
- Readily accessible health data can enable real-time situational awareness and response.
- “Social-health” information exchange also includes community-based service organizations and can enable providers to focus on the whole person during a disaster response, addressing acute medical needs as well as housing, shelter, and other needs that impact health.
A major theme throughout the workshop was how the collection, analysis, use, and sharing of health data is a driving force in the transformation of U.S. health care. Lurie said that the expansion of health information technology (IT) will provide better, faster, safer care. Already, electronic health records (EHRs) have contributed significantly to the continuity of care in disasters. Data from EHRs, syndromic surveillance, and other sources can facilitate modeling, predictive anal-ytics, and real-time situational awareness that can aid effective planning and execution before an event, and provide decision support during an event.
1This list is the rapporteurs’ summary of the main points made by individual speakers and participants and does not reflect any consensus among workshop participants.
The Affordable Care Act (ACA) is a catalyst that can move the health care focus from sick care within a hospital environment to health promotion and disease prevention within the community. The challenge is securing the funding to implement this across the country, especially in underserved areas. Every locality needs to understand its population, the segments of that population, and the risks to that population in that area. Just as DeSalvo explained earlier in the report, Embrey suggested that as more people gain coverage under the ACA and participate in health systems, there will be volumes more data available that can be used to better understand the health and preparedness needs of the community, plan for those with specific health needs (e.g., children, elderly, disabled who will need specialized care), and foster individual and community resiliency.
Lessons from New Orleans
DeSalvo and Lurie shared an example from when Hurricane Isaac left New Orleans without power for 5 days in the fall of 2012. DeSalvo said there were significant concerns about residents who were dependent on electronic durable medical equipment who did not evacuate. In the absence of knowing who was where, public health officials went door to door to identify those with medical special needs. This was clearly not efficient, so they also used Medicare claims data, in a way that complies with the Health Insurance Portability and Accountability Act (HIPAA), to map the location of oxygen-ventilator–dependent residents in the city. Matching those addresses against some of the high-rise apartment buildings, they were able to prioritize which buildings the power company should work to restore power to first, and then to evacuate or bring back-up batteries to others. Understanding the population is also essential for managing evacuation. Lurie noted that while the convention center in New Orleans was equipped to handle evacuees arriving with durable medical equipment, the federal medical station that was set up in Baton Rouge could not provide care to as many people as planned because the circuit breakers could not handle the load of the medical equipment.
In the summer of 2013, the Assistant Secretary for Preparedness and Response (ASPR) conducted a pilot program in collaboration with the Centers for Medicare & Medicaid Services (CMS) and the city of New Orleans to identify and map (rapidly and in a HIPAA-compliant way) all the people living in New Orleans who have electrically dependent durable medical equipment. A door-to-door exercise showed that the information harvested from CMS records correctly identified the right people and their equipment 93 percent of the time. If CMS data could be combined with data from private insurers and accountable care organizations (ACOs), Lurie said, then a picture of the whole population of a community and their needs could be developed.
Embrey said that understanding the current community demographics means also having a way to communicate to population segments. There are sophisticated tools that facilitate the use of cell phones to collect public health information on the ground, and to use that information for decision support in an emergency; however, there is no common framework and no resources dedicated to developing them for use in disaster response. Because of this and other reasons, communication with medically vulnerable individuals in disasters is done differently in cities across the country and could benefit from standards or coordination.
Kevin Larsen, medical director of Meaningful Use in the Office of the National Coordinator for Health Information Technology (ONC) provided a brief introduction to the evolution of the health information infrastructure. Prior to the ACA, the Health Information Technology for Economic and Clinical Health (HITECH) Act enacted as part of the American Recovery and Reinvestment Act of 2009 (ARRA) funded the Meaningful Use program. The program has two components, a CMS component that includes incentive payments to hospitals and doctors for implementing EHRs, and an ONC component that involves certification of EHR technology, development of industry standards to ensure interoperability, and regulation of what IT vendors produce. In many ways, the HITECH Act was set up to support the implementation of the ACA. About 80 percent of hospitals and more than half of the physician practices in the country are now using health IT and EHRs. Interestingly, Larsen said,
these are close to evenly distributed across urban and rural practices, and large and small practices. Health information exchange (HIE) supports the ACA, ACOs, and other new payment models such as patient-centered medical homes, helping individual patients, but it also supports resilience within communities.
Larsen highlighted the impact of the expansion of health IT that occurred between Hurricane Katrina and Hurricane Sandy. During Hurricane Katrina, paper records were destroyed and many people were displaced and had no access to their health information. Many people with chronic disease could not access their current medications. By contrast, there was a robust HIE in New York at the time of Hurricane Sandy, and the same infrastructure that is used every day allowed patients to have continuity of care and access to their medications during and after the storm, wherever they received care. Lurie shared a similar example from the tornadoes in Joplin, Missouri. All patients evacuated from the regional medical center that was destroyed were able to receive relatively seamless care wherever they were transferred to because EHRs were available, and patients being treated at outpatient substance abuse centers that were destroyed were able to receive their methadone at a clinic 1 hour away because remotely hosted EHRs were accessible. Larsen said that EHRs are a resilient system, a routine working system that is used every day. When a disaster strikes, a whole new system or process is not needed.
In addition to provider-mediated exchange, there is also what Larsen called “consumer mediated exchange.” As part of the Meaningful Use program, ONC’s Blue Button Initiative2 is a portal that allows individuals to access their own record and to download a standardized file. Another key part of the Meaningful Use program is exchange with public health, including public health reporting and syndromic surveillance reporting.
Justin Barnes, vice president of Greenway Medical Technologies and board member of the CommonWell Health Alliance, said a lot of money has been dedicated to support the development of health IT infrastructure for data exchange, from both the ACA and the ARRA, and there are many opportunities to build onto existing infrastructures. For example, $21 million was awarded through the ACA to 43 Health Center Controlled Networks to support the adoption and meaningful use of EHRs. Through the ARRA, the State Health Information Exchange
Cooperative Agreement Program awarded $564 million to fund state and regional health information exchanges. Other investments from the ACA that can be leveraged for health IT resilience include the ACA Prevention Public Health Fund, with more than $10 billion to award for a variety of activities, including public health research and tracking, as well as immunization expansion, and $5 billion invested in the Patient-Centered Outcomes Research Institute (PCORI), which includes funding to establish data infrastructures to create, disseminate, and implement research evidence and practices. Barnes said there is an opportunity here to encourage PCORI to conduct research on emergency health care and IT.
Many speakers and participants stressed the need for standards-based interoperability of health information systems. Barnes noted it is unlikely there will ever be one big centralized database for the country, but standards would allow access to and sharing of information across different databases. There are opportunities to leverage existing interoperability efforts, Barnes said, referring to the ONC eHealth-Exchange and S&I Framework as examples.3 Beyond standards-based interoperability of health record systems, there are also opportunities for open architecture integration between EHRs and application programming interfaces that allow public and private partners to share data (e.g., to link federal or state emergency preparedness centers with local emergency departments [EDs] and emergency responders). Doing this requires a minimum standard for data content and for the transport of that data. Mobile EHRs, patient portals, state and regional HIEs, and personal health records (e.g., Blue Button mentioned by Larsen) can all be combined to support public health, emergency preparedness, and both individual and community resilience. Importantly, local, state, or proprietary standards or interfaces should be discouraged, he said, as this would create more disjointed programs. In addition, he noted that the lack of a national patient identifier and the lack of comprehensive syndromic surveillance standards add to the challenges of sharing across databases.
Integrating Systems to Understand Data
Trying to integrate data across disparate clinical systems is challenging, said Roland Gamache, senior director for informatics for the National Association of County and City Health Officials. Comprehensive clinical information comes from a broad set of data sources, including communicable disease surveillance and syndromic surveillance that is required for health preparedness tasks. Population-based clinical data are captured and stored in many independent databases and information management systems, with different patient identifiers. The aggregate of information about an individual across such data sources could be used for many uses including the routine delivery of care, public health processes, clinical and comparative effectiveness research, preparedness activities, and other public health and health care–related processes. These data sources are often referred to as silos, and informatics allows us to build bridges among these silos, Gamache said.
Larsen said that moving public health reporting to an IT-embedded infrastructure has faced challenges that include the inability of state systems to receive the data and lack of standards across departments. Barnes concurred and said that uptake requires education of state leaders to create these standards and to enable care providers to share more information. The standards to achieve Meaningful Use in a comprehensive way currently do not exist. Barnes advocated for measures endorsed by the National Quality Forum (NQF)4 for the vendor communities.
Several participants expressed concern about the lack of inclusion of community pharmacies in the integration of EHRs and access to that information in times of disaster. Most community pharmacies are not an integrated part of health systems in regions. They are part of the small business model of health care. Larsen said that the bulk of prescriptions in the United States are done through e-prescribing, which follows very tight standards. However, prescription information remains very decentralized and distributed. Larsen mentioned that the large e-prescribing vendor, Surescripts, has been leveraged in disasters, and Barnes cited the extensive Walgreens EHR network, but preparedness planning needs to go beyond relying on a vendor for that solution. Recalling Lurie’s earlier
4NQF is a nonprofit, nonpartisan, public service organization committed to the transformation of the health care system. NQF reviews, endorses, and recommends use of standardized health care performance measures.
mentioned point on involving all partners in the development of health care coalitions, keeping community pharmacies involved in planning, but not relying entirely on one vendor, may help to maximize efforts in this area.
Gamache shared an example of how applying standards and being able to use syndromic surveillance data across jurisdictions allowed for the system to distinguish ED visits related to a disaster event (e.g., a tornado) from other ED visits for other conditions (e.g., respiratory, gastrointestinal, or fever events). This is important because it allowed for monitoring for bioterrorism or other types of emergency events even while there was an emergency event already occurring. Once an outbreak has been identified, data can be used for outbreak management and mitigation efforts. Information can also be disseminated to providers in the community about who is most at risk or how to allocate limited resources (e.g., vaccines).
Challenges of Privacy and “Change”
Larsen concluded by saying health IT implementation is really about “change management” more than it is about the technology itself. We need to identify key infrastructure and help organizations in achieving standards and upgrades for information sharing, he said. Change can be difficult because organizations all have systems they know and are used to, and they may need to compromise to be able to be interoperable. However, combining all of these evolving systems of data could augment continuity of care, both routinely and in a disaster. A challenge for interoperability is respecting and honoring the diversity of privacy and security of patients. Rules across health systems make this more complex, because each state has individual rules regarding privacy and security. However, some IT systems have eliminated part of this challenge by obtaining patient authorization for sharing information. For example, Epic Care Everywhere provides a framework for interoperability between Epic and non-Epic participating providers so that wherever the patient goes—between health care systems in the same town or across state and national borders—the clinicians providing care can have the details they need.5 Protected health information, including certain sensitive health information, can be shared with
5For more information on Epic, see https://www.epic.com/software-interoperability.php (accessed May 10, 2014).
patient authorization and in accordance with Epic’s “rules of the road” to which all participating providers agree.
In addition to the now well-known HIPAA regulations and state-level security rules, 42 CFR Part II also provides special privacy protection for some kinds of mental health care and substance-abuse addiction care.6 Enacted in 1970, 42 CFR Part 2 prohibits disclosure that an individual is in care or has been in care unless the patient signs a consent authorizing the release of the information for a specific reason and to a specific person. The regulations effectively prohibit routine release of information to physicians, inhibit inclusion of addiction treatment information in electronic medical records, and isolate addiction treatment programs (McCarty and Hartnett, 2013). Although the health care system has changed a lot since 1970, 42 CFR Part II has not been amended to account for changes in technology and integrated care. Even with the passage of the ACA, there is concern that the privacy limitations of 42 CFR Part II could restrict the integration of addiction treatment into other medical health settings and prevent “whole patient” coordination of care (McCarty and Hartnett, 2013).
Benjamin pointed out that there seem to be very different mindsets for what people accept relative to the gadgets used in their daily life and their health. The private sector has crafted systems (e.g., cell phone applications or websites for entertainment, shopping, and travel) that know who people are, where people are, what people need, and when they need it, and can push information out in a way that they have adopted and accepted into their lives. However, having people tracked across health systems has been a challenge mainly because of privacy concerns. Related to this, the conversation about a national patient identifier, while somewhat controversial, is essential because it impacts data-driven decision making. Barnes said it was a priority for the CommonWell Health Alliance, the trade association representing health IT developers. One example where the patient identifier could improve care is preventing duplicate data in a health registry. Although not at the national level, Martin discussed use of patient tracking software at the Boston Marathon in 2013. Because several emergency medical services were involved during the race, along with several doctors along the route and in various medical tents after the finish line, their medical intelligence center uses technology to track the patients they receive, which helps in providing a high level of coordinated care. Although the
logistics to implement this concept at a small “race level” such as the marathon seem manageable, scaling the concept to a higher city, state, or regional level may become more difficult and require additional levels of privacy and approval. Benjamin added that there is political sensitivity to tracking people through a national patient identifier.