OGC Disasters Resilience Pilot User Guide: Health SDI - Facilitating Public Health During a Disaster

According to the United Nations (UN) Office for Disaster Risk Reduction, financial damages and the cost in human lives due to natural disasters are tremendous:

United Nations Statistics on Disasters, 2005-2014

The statistics pre-date the 2017 Atlantic Hurricane Season which was one of the most active in history with 17 named Hurricanes [Federal Emergency Management Agency 2018] and with just three storms - Hurricanes Harvey, Irma, and Maria - causing $265 billion in damages. Hurricane Maria, which made landfall in Puerto Rico on September 20, 2017 as a Category 4 hurricane caused $90 billion in damages, crippled the power grid for months, impacted 100% of the island’s 3.8 million population and caused 2,975 deaths [Milken Institute School of Public Health 2018].

In the Response to Hurricane Maria, there is anecdotal evidence that the lack of medical personnel early on in the response contributed heavily to storm-related fatalities. These are lives that could have been saved if medical personnel were dispatched to the disaster impacted area sooner.

The scope of the geospatial need for medical personnel wasn’t known by Emergency Operations Managers until the response was underway. HSR’s system provides the foresight to ensure the medical and health needs of the impacted population are known to both Emergency Operations Managers and Emergency Medical Responders in advance and can be addressed.

To provide this foresight, HSR is creating a revolutionary Health Geo Platform – a Health-focused Spatial Data Infrastructure (SDI), as a National Health SDI, with an initial application to allow for the inclusion of health information in emergency or disaster response planning and execution processes. This application is named a Health Risk Index and merges data on social determinants of health as well as health data to identify an impacted population’s risk – health risk – from a natural disaster. This information can be utilized by Emergency Operations Managers and the boots-on-the-ground Emergency Medical Responders to ensure their efforts address the unique health challenges caused by experiencing a natural disaster.

Currently, emergency response operations conducted by the Federal Emergency Management Agency (FEMA) and their peers at State and Local levels have limited visibility into the health needs of the disaster-impacted populations they serve. Therefore, emergency operations managers and emergency medical responders may develop response & evacuation plans, or may enter disaster struck areas unaware of the overall health and medical needs among civilians.

The Health SDI and its Health Risk Index application will provide the necessary, geocoded public health information that may prove vital to the success of emergency response efforts and will align smoothly with the global emergency response systems. The Health SDI and the Health Risk Index are revolutionary decision support tools that hold unlimited capabilities.

Chapter 2 provides a high-level description of the architecture of the Health SDI and the data sources behind the Health Risk Index.

Chapter 3 provides additional details on the symbology and specific data sets used and published in this effort.

Chapter 4 provides discussion on key special topics that may be relevant and related to this effort going forward.

Chapter 5 details how the SDI works describing how actionable information is put into the hands of the two primary end users (emergency operations managers and emergency medical response personnel) in a way that they can use during disaster response scenarios.

Chapter 6 will present a Conclusion of this specific effort - which is Phase 1 of a longer effort to ensure emergency response personnel have the health and medical information they need, when they need it, and within their existing workflow to best serve the impacted population.

This chapter will also discuss viable future applications of the Health SDI. The utility of a SDI progresses is limited only by imagination. For example, the SDI can provide focus on the greatest health needs in the world and streamline medical and pharmaceutical research to address those needs.

The Health Risk Index leverages both data on social determinants of health as well as health outcomes and costs from a broad variety of sources. These include global non-governmental organizations, such as the UN and World Health Organization, national government organizations, such as the Centers for Disease Control and Prevention (CDC) and the Homeland Infrastructrue Foundation-Level Data (HIFLD) produced by the Department of Homeland Security (DHS) within the U.S., as well as U.S. State, and Local government agencies. Health Registries, disease-specific healthcare organizations, university and non-profit research organizations and other sources of data are also considered.

HSR continues to seek data from new sources, including US-sources at the National, State, and Local levels, non-governmental organizations, educational organizations, as well as international sources that make health and health-related data available for public consumption.

HSR is working towards building an SDI that can serve as a National SDI as well as an international SDI for health issues and therefore are working to integrate and collaborate with existing SDIs and catalog providers, including GeoPlatform.gov, AmeriGEOSS, and NextGEOSS.

As the near-term output of the Health Risk Index relates to disaster response efforts, the primary data consumers, or end users, include:

Emergency Operations Managers are those individuals who plan and oversee the overall disaster response effort. Emergency Medical Responders are the “boots-on-the-ground” who go into disaster-impacted areas and provide disaster relief services including needed medical care. The needs of these users are certainly related and both users can have visibility into the data provided to each other.

The Health Risk Index is published to the Health SDI. Effort to publish the Health Risk Index to GeoPlatform.Gov are underway.

At present, the Health Risk Index uses the most currently available data for the identified datasets that are in the public domain. Later phases of this effort will include discussion of licensing these data sets from providers to have access to either (or both where available) more regularly updated data and current data as well as access to data at the Census tract level. These discussions will involve addressing and documenting policy and procedure issues, security and privacy issues, as well as the republication of data.

To produce the Health Risk Index, HSR sought health and social data that, when considered together, stratify the health risks facing an impacted population due to a natural disaster.

Additionally, boundary layers were discovered through searching the U.S. Census Bureau (Census), guidance on symbology was taken from the National Alliance for Public Safety GIS Foundation, and address data was discovered from the U.S. Department of Transportation’s National Address Database (NAD).

The following data tables are used in the construction of the Health Risk Index:

The health and social data inputs are joined to crosswalk files which allows for the information to be displayed at multiple geographic levels. Data from the health and social datasets are then used to calculate the Health Risk Index at each geographic level. Additionally, the health needs information relevant to the on-the-ground Emergency Medical Responders is displayed and exported. After the Health Risk Index is calculated for each geographic level the table is joined to the geographic boundary files from the Census, where symbology is then established and the layer is published to the Health SDI. The layers available on the Health SDI can be retrieved using OGC standards. This allows our partners, such as Esri and Compusult to import and utilize the data in their platforms, allowing for wide use of the information by the First Responder community overall.

Finally, predicted storm paths from the U.S. National Oceanic and Atmospheric Administration (NOAA) or other sources of disaster impact information can be overlaid onto the Health Risk Index to retrieve the index at all geographic levels as shown in Figure 2 for the National Level.

Emergency Operations Managers can isolate the index for the impact region as shown in Figure 3.

All of the health data is joined with crosswalk files from the U.S. Department of Housing and Urban Development (HUD) and republished as the HSR Medical Dataset on the Health SDI. Similarly, the social data is joined with HUD crosswalk files (https://www.huduser.gov/portal/datasets/usps_crosswalk.html) and republished as the HSR Social Determinants Dataset on the Health SDI. Joining the health and social data inputs to the crosswalk files allows for the information to be displayed at multiple geographic levels. Additionally, the Health SDI will republish other data and resources related to disaster response efforts, including but not limited to a link to NOAA’s National Hurricane Center data archive and the symbology guidelines of the National Alliance for Public Safety GIS Foundation.

The symbology used in the Health Risk Index is based on the symbology guidelines produced by the National Alliance for Public Safety GIS Foundation. This includes the symbols for hospitals, roads, the thematic coloring, as well as the legend.

The legend is a 7-point scale ranging from very low risk, or a white color on the map, to a extremely high risk, or purple. The greater the risk, the more urgent the need to provide appropriate medical care and intervention, up to and including potential evacuation to a specific medical facility, to the population within the geographic region.

In emergency situations, time is of the essence. It is critical that decision makers have the right data - accurate & reliable data - at the right time so they can make immediate decision and take action. In that light, great care must be given to ensure that only accurate and relevant data is brought together to create a Health Risk Index that Operations Managers can use to construct effective disaster response plans.

As in the case of Hurricane Maria, Emergency Operations Managers did not know the need for medical personnel until the response was already underway and it was too late to include medical personnel in the earliest stages of response. The Health SDI will highlight the medical and health needs of the impacted population so they can be addressed - and at the right stages of the response to have the greatest impact on saving lives.

Through field tests in later phases of this overall effort, we will ensure the data included in the Health Risk Index is comprehensive. Field tests will also ensure the Health Risk Index is coordinated with the First Responder communities' existing workflow and processes.

Within the Healthcare space, there currently is limited semantic mediation overall. In other words, across the globe - and often also at local levels within individual health systems, the same medical conditions can be referred to differently. This does impact diagnosis and treatment to a certain extent. It also complicated medical research. There are efforts underway that indirectly aim to address or mediate such variation, one example being the current Gravity Project. A great value of a Health SDI is that it can be an effective and powerful tool for achieving Semantic Mediation within the space. The Health SDI can provide a platform for other industry efforts, including the Gravity Project, to achieve mediation.

The Health Risk Index incorporates population-level health data that determines the population’s risk of adverse health event specifically under the stresses of a natural disaster. This is related to, and a sub-case of general health risk. Included are data on the presence of health conditions and diseases that may be flair up in a disaster. Also included are data on the presence of those health conditions and diseases that are susceptible to health degradation after just one or two missed treatments or medication dosages. Disruption in treatment and medication dosage are common in disaster scenarios.

Also considered is data on those who are heavily reliant on infrastructure that may be unavailable in a disaster, such as transportation, power, community support, etc. The ambulatory status of the underlying population is also considered. For instance, some individuals may simply need the evacuation route identified (and can drive themselves out of harm’s way), others may need a vehicle for evacuation, others may need medical transport (e.g., ambulance), and still others may need assistance or stretchers from their home to the medical transport vehicle.

The health data gathered to produce the Health Risk Index has been merged into a single, new dataset called the HSR Medical Dataset that is published on the Health Geo Platform, or Health SDI.

In order to determine a disaster-impacted population’s health risks, data on social determinants of health (SDoH or social data) has been included in the Health Risk Index as social determinants are widely acknowledged as predictors of health outcomes. Care is given to include those SDoH factors that suggest how health posture may change in a disaster scenario - rather than the SDoH that inform health outcomes overall. In the disaster response scenario, it’s important to know who the at risk patients are within the environment of that disaster event, and not just who may be at risk. One is a subset of the other, no doubt. In a disaster, where time and speed of response is of the essence, it is essential to address needs in the order of criticality.

The social data gathered to produce the Health Risk Index has been merged into a single, new dataset called the HSR Social Determinants Dataset that is published on the Health Geo Platform, or Health SDI.

Smart Cities, and their large-scale application of distributed sensors, provide an exciting opportunity to continue and in fact drive further advancements in global health. Distributed sensors that are associated with smart cities can allow for new means of gathering data on the health risks present within communities at the population as well as at the individual level.

This constitutes both a larger and more real-time health data set that can help predict adverse health events, diagnose and treat patients, as well as monitor the success of prescribed treatments. And treatment plans that prove ineffective can be adjusted more quickly.

Examples include the use of sensors to:

This information can address current challenges in assessing hospital and medical facility status during emergencies. Overall, the analytics that a smart city can provide will improve the ability to assess the overall medical care needs of the population.

Assisting first responders address the needs of populations during a disaster is critical for saving lives.

It is also important to ensure post-disaster recovery leads to full restoration of the necessary healthcare service delivery capacity. The health risk index can be used to track overall recovery and adapting the index for such a purpose can certainly be an important and beneficial follow-on activity. For example, a storm that enables a source of pollutant to contaminate ground water can have an adverse impact on crop yields, area plant life and wildlife, as well as human health. Efforts to purify the water supply and rebuild safeguards can be tracked through the Health SDI to restore, or improve, pre-disaster population health risk levels.

The OGC’s Disaster Resilience Pilot is a response to a recognized gap in overall disaster response planning and execution - that is the lack of health information of disaster-impacted population. The response to Hurricane Maria demonstrated the consequences of that gap: If geospatial health and medical needs are not known in advance, efforts to address those needs may not be executed in the most timely fashion to have the greatest impact. This can lead to increased economic costs in addition to costs in human lives and suffering.

HSR is able to establish a Health SDI hosting a Health Risk Index to fill that gap by leveraging publicly available health and social data from numerous governments, research organizations, education institutions, and other sources. OGC standards can then be leveraged to provide that data to Emergency Operations Managers and Emergency Medical Responders directly, as well as through partner organizations. Further efforts (discussed below) will address licensing issues, field tests to refine the Health Risk Index, as well as ensuring the information is actionable by First Responder community by integrating with their current workflow and process flow. The end result will be more timely delivery of essential medical care and intervention resulting is fewer lives lost to natural disasters.

In Phase II of this effort, HSR will work to formalize the data sharing Policy and Procedure agreements with all data providers. Currently, publicly available data sources are sufficient for assessing the initial health risk of an impacted population to a specific disaster event - however, closer to real-time data will be more helpful for both response planning as well as on-the-ground delivery of emergency medical care. This Phase II effort will discuss with data owners the licensing, policy, and procedures issues involved in accessing the most currently available data (hopefully closer to real-time), as well as data at the Census tract and address level. Data at the Census tract and address level will involve security and privacy considerations.

In addition, HSR will continue to search and determine optimal data sets and sources for inclusion in the Health Risk Index. This will include exploring data at the Census tract and the address levels.

In Phase III of this effort, the Health Risk Index will be refined through formal field tests with FEMA, their peers, and all current and potential new partners within and/or serving the First Responder community. This will also ensure the Index conforms to the First Responder community’s planning and execution workflows.

In addition, HSR will also work with the provider community on mechanisms to identify, track, and aid post-disaster recovery efforts. In other words, depending on the size and scope of the disaster, the underlying health posture of a community may change - demanding specific and targeted efforts to reverse that impact. For example, a storm that enables a source of pollutant to contaminate ground water can have an adverse impact on crop yields, area plant life and wildlife, as well as human health. Efforts to purify the water supply and rebuild safeguards can be tracked through the Health SDI to restore, or improve, pre-disaster population health risk levels.