33

Informatics Solutions for Emergency Planning and Response

Elizabeth (Betsy) Weiner / Lynn A. (Slepski) Nash

INTRODUCTION

Unfortunately, both natural and man-made disasters have catapulted us into a world that has resulted in making emergency planning and response a high-priority need. There has been a documented rise in terrorism incidents as well as natural disasters worldwide. Natural events have ranged from earthquakes, tsunamis, floods, hurricanes, typhoons, forest fires, to pandemic disease events affecting billions. Conflicts and nuclear disasters have added to the complexities. In addition to natural disasters, political and social upheavals massively disrupt the lives and livelihoods of populations and result in the forced displacement of millions of people. The World Health Organization (WHO, 2016) reported that during 2016 they reorganized as the new WHO Health Emergencies Programme using new emergency-management processes for risk assessment, grading of emergencies, and incident management. Their action was in response to a record 130 million people in need of humanitarian assistance, and disease outbreaks becoming a constant global threat (WHO, 2016). This new organization addresses the full cycle of disasters, working with countries to plan before an event happens, to respond in an emergency, and once the initial event has passed, to work on recovery activities. Central to their mission is the use of appropriate informatics tools. The purpose of this chapter is to explore the intersection between informatics and emergency planning and response in order to determine current and future informatics contributions.

The United States is not immune from this increased emphasis on emergency preparedness. The events of September 11, 2001, forced the United States into the realization that the country was not adequately protected from terrorism. Then, within a short window of time, the anthrax outbreaks stressed the public health infrastructure to the point that bioterrorism arose as an additional deadly threat. As a result of these two experiences, the government of the United States responded at an unprecedented pace to better prepare and manage terrorist events. Furthermore, the pandemic H1N1 incident in 2009 created more unique data collection challenges that caused public health officials to creatively provide solutions for meaningful data acquisition in order to be able to effectively manage the event. Mass shootings in the United States continue to rise, with major implications for school safety. Shootings such as the one at Marjory Stoneman Douglas High School in Parkland, Florida in 2018 have sparked activism on the part of the survivors, but no gun control legislation has currently taken place. School safety plans have, however, been updated to reflect better security and registration systems for outsiders. Other safety precautions have taken place in response to terrorism events such as the bombing at the Boston Marathon in 2013. Hemingway and Ferguson (2014) reflected on lessons learned during the Boston Marathon bombing and concluded that emergency preparedness plan updates must reflect the changing types of disasters, changing communication technologies, and the changing workforce. Just as this book was going to print, a new, evolving global threat, SARS-CoV-2—the causative agent of the COVID-19 pandemic, emerged and spread rapidly in just weeks to wreak havoc on the world’s population. The pandemic continues to expand, despite mitigation efforts. Perhaps no better example exists of the importance of data-driven decisions than the handling of this event.

Early contributions by the informatics community focused on surveillance of threat detection. However, as informaticists became more familiar with emergency planning and response, it became clear that contributions toward efficiency, analysis, remote monitoring, telemedicine, and advanced communications would be valued. The most consistent challenge for emergency and disaster response continues to be communication and information management. Effective response requires high situational awareness analyzing real-time information to assess needs and available resources that can change suddenly and unexpectedly. There is a critical interdependence between data collected in the field about a disaster incident, casualties, healthcare needs, triage, and treatment and the needed community resources such as ambulances, emergency departments, hospitals, and intensive care units. Concurrently, information from the various inpatient facilities and ambulance resources alters the management and disposition of victims at the scene of a disaster. Opportunities abound for new telecommunication technologies. Smart devices, wireless connectivity, text messaging services, social media, and positioning technologies are all advances that have application during disaster events. These technologies are being used and evaluated to improve patient care and tracking, foster greater safety for patients and providers, enhance incident management at the scene, coordinate response efforts, and enhance informatics support at both the scene of the disaster and at the community resource levels. Mesmar et al. (2016) reviewed the literature on digital technologies used during crises, their impact on affected populations, and subsequently described opportunities for and challenges faced by users. They identified 50 digital technologies and classified them according to the determinants of health they affected. These technologies were found to facilitate communication, coordination, and collection and analysis of data, enabling timely responses in humanitarian contexts. A lack of evaluation of these technologies, and issues of privacy and equity constituted major challenges. Newer types of technologies found by them included instant digital classrooms (loaded with e-books and other electronic resources), electronic voucher cash cards for mobile money, gaming, credit card– sized computers that can be inserted in monitors preloaded with educational material and information portals, redesigned portable medical equipment, robotics, spoken language translation, and 3D printing.

The 2004 earthquake and tsunami that devastated parts of Southeast Asia illustrated the uncoordinated invasion of people and organizations that resulted in unnecessary duplication, competition, and failure to assist many of the victims in need (Birnbaum, 2010). The World Health Organization took on the challenge of increased global coordination and response, working with 194 Member States, across six regions, and from more than 150 offices (WHO, 2019). This organization at the global level has been aimed at discouraging individual and organizational response efforts that were not part of this coordinated response. The United States has also organized their planning and response efforts for the same reasons, and informatics is increasingly taking on more important roles in these efforts.

Central to any discussion of a pandemic event is the Case Fatality Rate (CFR) (percent of deaths from the total number of people who are identified as infected with the disease). However, CFR alone does not tell the entire picture. Equally important is the R0 (pronounced Rnaught). R0 is the number of people on average, the infected person will in turn infect. The unmitigated RO for COVID 19 estimated from data from Wuhan is 5.7 (Sanche, et al., 2020). Additional data used on a daily basis are new case counts, death counts, and recovery statistics. Furthermore, controversy has arisen about the use of masks to help control disease transmission. Chu et al. (2020) determined an approximate probability of infection, when exposed at close proximity for a considerable period of time to be approximately 17% when not wearing a mask. At the end of May, 2020, the US experienced the “Great Clips Incident” (Hendrix, Walde, Findley & Trotman, 2020; Washington Post, 2020). Two symptomatic hairdressers at a Great Clips salon in Springfield, Missouri cut the hair of 139 patrons before testing positive. A 17% probability would create an expectation of about 23 infections. When some of those people were tested and the rest closely followed, the actual result was 0 infections. While 0 infections is a possibility by chance alone, it is extremely unlikely (p=.000000000004). Clearly hygiene, particularly mask wearing by hairdressers and patrons, mitigated the spread of the disease. It is hopeful that using data-driven decisions will assist in the eventual eradication of COVID-19.

THE FEDERAL SYSTEM FOR EMERGENCY PLANNING AND RESPONSE

Most disasters and emergencies are handled by local and state responders. The federal government provides supplemental assistance when the consequences of a disaster exceed local and state capabilities.

Under the Homeland Security Presidential Directive 5 (HSPD5) (White House, 2003), the Secretary of Homeland Security, as the principal federal official for domestic incident management, coordinates federal actions within the United States to prepare for, respond to, and recover from terrorist attacks, major disasters, and other emergencies. Coordination occurs if and when any one of the following four conditions applies: (1) a federal department or agency acting under its own authority has requested the assistance of the Secretary; (2) the resources of state and local authorities are overwhelmed and federal assistance has been requested by the appropriate state and local authorities; (3) more than one federal department or agency has become substantially involved in responding to the incident; or (4) the Secretary has been directed to assume responsibility for managing the domestic incident by the President. Further, HSPD5 directs federal department heads to provide their full and prompt cooperation, support, and resources to the Secretary in protecting national security.

In 2006, the Pandemic and All-Hazards Preparedness Act amended the Public Health Service Act and tasked the Secretary of the Department of Health and Human Services (HHS) to lead all federal public health and medical responses to public health emergencies (Department of Health and Human Services [HHS], 2010). Also included in this legislation were many requirements to improve the ability of the nation to respond to a public health or medical disaster or emergency. Specifically included was a near real-time electronic nationwide public health situational awareness capability to enhance the early detection of, rapid response to, and management of potentially catastrophic infectious disease outbreaks and other types of public health emergencies.

The National Response Framework, enacted in January 2008, established a comprehensive, national, and all-hazards approach to respond to disasters and emergencies (Department of Homeland Security [DHS], 2008a). Built on its predecessor, the National Response Plan, it includes guiding principles that detail how federal, state, local, tribal, and private sector partners, including the healthcare sector, prepare for and provide a unified domestic response, improving coordination and integration. The framework emphasized preparedness activities that included planning, organizing, training, equipping, exercising, and applying lessons learned and assigned lead federal agencies to each of 15 Emergency Support Functions (ESF) (DHS, 2008b). The document was updated with a third edition (DHS, 2016) to implement the requirements and terminology of the National Preparedness System and emphasize the need for the involvement of the whole community. The fourth edition of the NRF was recently updated to address the many challenges encountered in 2017 during the U.S. government’s response to three major hurricanes in quick succession while at the same time responding to historic wildfires in California. Specifically, the 2017 Hurricane Season Federal Emergency Management Agency AfterAction Report called for a revision to the NRF to emphasize stabilization of critical community lifelines (e.g., safety and security; food, water, sheltering; health and medical; energy [power and fuel]; communications; transportation; and hazardous material) during severe and widespread events that occur over weeks and months. Because community lifelines rely on multiple government entities, businesses and infrastructure sectors, they have interdependencies. Failures in any one lifeline will cascade across to others. This fourth update includes a new ESF #14, Cross-Sector Business and Infrastructure, charged with monitoring interdependencies by conducting cross-sector analyses and coordinating multi-sector response operations between (or across) the government and private sector for natural or human-caused catastrophic incidents that jeopardize national public health and safety, the economy, and national security (DHS, 2019).

The ESF group functions are used to provide federal support during a response (Table 33.1), and assigns an ESF coordinator for each functional area. The Department of Health and Human Services coordinates public health and medical responses, including biosurveillance.

CASE STUDY 33.1: INFORMATICS AND 2009 H1N1

The 2009 H1N1 influenza pandemic, the last complete incident, best illustrates how informatics can contribute to an emergency response. Although the current COVID-19 pandemic also demonstrates informatics applications, that pandemic has yet to reach conclusions for effective analysis. While we await the conclusion of the current pandemic and it’s review, this discussion thus centers on the 2009 H1N1 influenza pandemic.

Initially concerned that a circulating H5N1 virus (Avian Influenza A) was mutating and could cause a human pandemic, global experts had focused efforts over the last several years on rapidly developing catastrophic plans even though a pandemic virus had not emerged. There were significant concerns, given that during the 20th century three flu pandemics were responsible for more than 50 million deaths worldwide and almost a million deaths in the United States (HHS, 2005) (Table 33.2). The CDC estimates that 43 million to 89 million people had H1N1 between April 2009 and April 2010, and they estimate between 8870 and 18,300 H1N1-related deaths (Shrestha et al., 2011). On August 10, 2010, the WHO declared an end to the global H1N1 flu pandemic (WHO, 2010).

Many governments had believed that if a pandemiccapable virus emerged, there would be rapid worldwide spread as the 1918 pandemic had spread across countries and continents in less than one year in a time without commercial air travel to facilitate the spread of disease (HHS, 2005). It was understood that a worldwide influenza pandemic occurring in this century could have major effects on the global economy, especially travel, trade, tourism, food, consumption, and eventually, investment and financial markets and could lead to widespread economic and social disruptions. As a result, many countries engaged in detailed pandemic planning and prepared to adopt draconian-like measures to delay but not stop the arrival of the virus, such as border closures and travel restrictions.

Here in the United States, modelers predicted catastrophic death estimates (Table 33.3). The 1918–1919 flu pandemic had caused the deaths of at least 675,000 Americans and affected about one-fifth of the world’s population. Researchers believed that if a pandemic of similar severity occurred today, 90 million Americans could become ill, quickly exceeding available healthcare capacity and result in approximately 2 million Americans deaths (HHS, 2005).

Preparedness planners assumed that all populations were at risk. They believed that disease would be widespread, affecting multiple areas of the United States and other countries at the same time preventing the redistribution of resources. The world would experience multiple waves of outbreaks potentially occurring for an extended period of time (over 18 months), affecting the entire United States for a period of 12 to 16 weeks with community waves each lasting 6 to 8 weeks (HHS, 2005). One to three pandemic waves would occur (OSHA, 2007). Further, planners believed that a pandemic could affect as many as 40% of the workforce during periods of peak flu illness, predicting that employees could be absent because of their own illness, or would be caring for sick family members or for children if schools or daycare centers are closed. They also recognized that workers would be absent if public transportation was disrupted or if they were afraid to leave home (DHS, 2007).

Adopting a “worst case scenario,” government experts rapidly developed a number of strategies to help local governments plan, stating that the federal government would not likely be able to provide any assistance during the actual pandemic. For example, the Centers for Disease Control and Prevention developed a Pandemic Severity Index (PSI) to attempt to characterize the severity of a pandemic (CDC, 2007). It was designed to predict the impact of a pandemic and provide local decisionmakers with standardized triggers that were matched to the severity of illness impacting a specific community. The severity index was based on a case-fatality ratio to measure the proportion of deaths among clinically ill persons. Recommended actions were identified in advance, and communicated to the public in hopes of increasing their understanding and compliance. Using the PSI, a severe pandemic influenza, similar to the 1918 pandemic, was defined as a category 4 or 5, with 20% to 40% of the population infected. For a severe pandemic, HHS recommended that localities be prepared to dismiss children from schools and close daycares for up to 12 weeks, as well as initiate adult social distancing, which included suspension of large public gatherings and modification of the work place schedules and practices (e.g., telework and staggered shifts).

The Centers for Disease Control and Prevention, part of HHS, monitors influenza activity and trends and virus characteristics through a nationwide surveillance system as well as estimates the burden of flu illness using statistical modelling (CDC, 2010). On April, 21, 2009, the CDC began reporting cases of respiratory infection with swineorigin influenza A (H1N1) viruses transmitted through human-to-human contact (CDC, 2009a, 2009b). It established the case definition for 2009 H1N1 as an acute febrile respiratory illness in a person and laboratory-confirmed swine-origin influenza A (H1N1) virus infection at the CDC by either of the following tests: real-time reverse transcription-polymerase chain reaction (rRT-PCR), or viral culture (CDC, 2009c). The CDC began tracking and reporting the number of cases, hospitalizations, and deaths at state, local, and national levels using standard state reporting mechanisms. It was soon apparent that using actual case counts resulted in dramatically underreported disease.

On July 24, 2009, the CDC abandoned initial case counts, when it recognized that those numbers represented a significant undercount of the actual number of 2009 H1N1 cases (CDC, 2010). They found that 2009 H1N1 was less severe and caused fewer deaths than expected when compared to the pandemic planning assumptions. As a result, existing plans, which used case fatality numbers as the trigger for initiating response actions, were not effective.

Scientists turned to other means to begin to understand the effects of disease and predict its future course. For example, because trending indicated that children and young adults were at higher risk, the Department of Education began looking at school closures and school absenteeism, examining both teacher and student absences. Each of the critical infrastructure key resource sectors held weekly calls with private sector partners to elicit whether there were trends beginning to indicate business interruption problems, which might forecast social disruptions. The National Retail Data monitoring system tracked the real-time purchase of over-thecounter (OTC) medications, such as fever reducers and influenza treatments, in over 29,000 retail pharmacies, groceries, and mass merchandise stores. This University of Pittsburgh (2014a) system is used to provide early detection of naturally occurring disease outbreaks as well as bioterrorism.

The CDC moved to using estimates. Using the influenza module from BioSense, the CDC tracked flu with data from over 500 local and state health departments, hospital emergency rooms, Laboratory Response Network labs, Health Information Exchanges, as well as the Departments of Defense and Veterans Affairs. The Real-time Outbreak Disease Surveillance (RODS) took chief complaint information from clinical encounters hospitals and classified it into one of seven syndrome categories using Bayesian classifiers. The data is stored in a relational database, used univariate and multivariate statistical detection, and alerted users of when the algorithms identify anomalous patterns in the syndrome count (University of Pittsburgh, 2014b). Interpreting these estimates, one study hypothesized that for every reported lab-confirmed case of H1N1 between April and July 2009, there were an estimated 79 total cases. The same study found that for every identified hospitalized case there were more likely 2.7 hospitalized people (Reed, et al., 2009).

Concerned that the limited capacity of the healthcare system would be overwhelmed and finite resources such as H1N1 test kits would be consumed, the CDC published updated self-treatment guidance and told the public that H1N1 testing was no longer necessary. Instead, persons with minor flu-like illnesses were assumed to be infected and encouraged to utilize advice lines staffed by nurses to obtain answers to questions rather than to seek appointments with healthcare providers. For the first time the U.S. government established a one-stop federal Web site that housed information such as frequently asked questions as well as messaging aimed at individuals and families, businesses, and healthcare professionals from across the federal interagency. The Web site contained tailored planning documents for schools and communities, and included targeted information for special populations. One particularly helpful site was a Flu Vaccine Locator, which contained a database that provided the general public with the locations of clinics that had vaccine supplies utilizing zip codes.

For the first time, HHS used social media to communicate with young people. Recognizing that large numbers of young adults were affected, they launched a Facebook application “I’m a Flu Fighter!” that allowed and encouraged users to spread information about H1N1, such as where they received the H1N1 vaccine, to their Facebook friends (Mitchell, 2010).

Other recent viruses have arisen but cannot be categorized as pandemic because they have not caused sustained and efficient human-to-human transmission. Informatics has been important in this reporting and analysis. H5N1, commonly known as avian influenza (“bird flu”), is such an example. In July 2013, WHO announced a total of 630 confirmed human cases which resulted in the deaths of 375 people since 2003, but did not meet pandemic criteria (WHO, 2013). Also in 2013, the American Academy of Family Physicians (2013) reported that the case numbers of H7N9 stalled in China, but that the pandemic potential remains. H7N9 is an unusually dangerous virus for humans with cases resulting in severe respiratory illness, with a mortality rate of roughly 30% (Li et al., 2014). H7N9 does not kill poultry, which makes surveillance much more difficult. Other recent threats include a new respiratory virus Middle East Respiratory Syndrome (MERS-CoV), which arose in 2013 in Saudi Arabia (Todd, 2014). We have also responded to other serious disease outbreaks with pandemic potential that were not influenza related, including Ebola (2014–2016) (Bell et al., 2016) and Zika (2015) (WHO, 2018) virus. While Ebola is blood-borne and Zika is mosquito-transmitted, each instance again highlighted the need to be as prepared as possible—because early detection and public awareness lead to a faster and more effective public health response. The most recent outbreak of COVID-19 resulted in the WHO Director-General declaring the global threat as pandemic on March 11, 2020 (WHO, 2020).

Going back to the 2009 H1N1 example, the U.S. government identified what strategies worked in the 2009 influenza pandemic and as well as the responses to other emerging infectious diseases and issued the Pandemic Influenza Plan 2017 Update (HHS, 2017). The Plan builds on previous successes and identifies five major work areas: (1) diagnostic testing and disease monitoring; (2) respiratory protection; (3) acceleration of vaccine and antiviral development; (4) modernizing medical countermeasure distribution and administration; and (5) healthcare seeking and surge strategies. Two of these have significant informatics potential: medical countermeasure distribution and administration can be enhanced by linking information technology and modern supply chain science to patterns of human behavior and care-seeking and surge need and capacities could be influenced by tools to aid individuals in care-seeking and decision-making so that people receive care that is safe and appropriate to their level of need. Much work is taking place with the World Health Organization to improve international preparedness and the rapid exchange of information and data.

Todd (2014) concludes that since all viral mutations are unpredictable, it is impossible to predict whether any of these viruses or yet another emerging virus will be the cause of a new pandemic. We must, therefore, continue our diligence in surveillance activities.

Healthcare Consumers Contribute to Surveillance Activities

In the H1N1 case study described above, healthcare consumer data became an important aspect of the disease surveillance model that augmented data collected by the CDC. Why was that the case? Now more than ever before consumers have the opportunity to contribute to surveillance activities. In some cases, the participation is a conscious decision, but in others consumers may be unknowingly contributing to this informatics process.

Part of the advantage of externally generated CDC surveillance mechanisms is that they shorten the typical lag time to publication for the CDC’s publicly reported data which is currently estimated to be from 10 to 14 days (Ginsberg et al., 2009). Telephone triage data are now being used to help track influenza in a specified geographic location with the added advantage that the data are real-time in nature. In addition, patient demographics and disease symptoms can also be captured in a standard format. Another new mechanism for data capture about influenza is through physician group proprietary systems. In these systems, the healthcare providers enter the data for suspected or confirmed influenza patients. By far, the most talked about trend in influenza surveillance for the 2009 H1N1 outbreak was Google’s Flu Trends. The assumption made with this system was that there was a relationship between how many people search the Internet for flu-related topics and how many people have flu-like symptoms. In studies conducted by Google. org comparing Google Flu Trends to the CDC published data, they found that the search-based flu estimates had a consistently strong correlation with real CDC surveillance data (Ginsberg et al., 2009).

Informatics and Incident Management

Information technology staff members have long been familiar with emergency planning for disaster recovery related to their systems, but find themselves in a new role as part of a more comprehensive team approach to disasters and emergencies. The incident management system (IMS) was first used by firefighters to control disaster scenes in a multijurisdictional and interdepartmental manner. The IMS calls for a hierarchical chain of command led by the incident manager or commander. Each job assignment is consistently followed by assigned personnel who refer to a specific job action sheet. This system improves communication through a common language, allows staff to move between management locations, and facilitates all responders to understand the established chain of command. The IMS has been adapted for hospital use and is called the Hospital Incident Command System (HICS).

The Emergency Operations Center (EOC) is the physical location where the Incident Management Team convenes to make decisions, communicate, and coordinate the various activities in response to an incident. Accurate, real-time data acquisition regarding patient needs, rescue personnel, and resources available is critical to overall coordination. Table 33.4 presents functions where technology can be used to capture and represent data for purposes of increasing situational awareness in the EOC for the purposes of making the most informed and efficient decisions. In addition, the informatics processing efforts that contribute to the incident management system are also described.

Stay updated, free articles. Join our Telegram channel

Join