At a recent conference, Henry, an epidemiologist with the Austin/Travis County Health and Human Services Department, had heard about Florida’s use of keywords in hospital emergency department (ED) reports. They were used to estimate the impacts of the 2010 earthquake in Haiti through Florida’s syndromic surveillance system, which collects data from various sources to detect disease trends by condition occurrence not just confirmed cases.
For visits to emergency departments by people in Haiti on or after the earthquake on Jan. 12, 2010, Florida added the word “Haiti” to the patient’s list of chief complaints. “I thought that was simple, elegant and on target,” Henry said.
For epidemiologists, it has always been difficult to get real-time data during an emergency. Hospitals can be low on personnel, who are often so busy that they have difficulty reporting data during an event, she noted. “But this was pretty resource-neutral. All we did was ask the two major hospital systems’ ED personnel to type ‘weather’ into the chief complaint data that we received.”
Within four or five hours of making that request, the term started to show up in syndromic surveillance reports, and the department could better attribute accidents to weather and the ice storm. “It is imprecise, but it gives us trends,” Henry said. “We were able to identify ZIP codes and send short narrative reports and tables to the EOC. If we have a longer event with more injuries, we will share that data with public safety.”
Providing Situational Awareness
The projects in Florida and Texas are examples of the way that public health agencies’ syndromic surveillance efforts are becoming more sophisticated and starting to realize their potential to provide situational awareness to emergency management personnel who must make timely decisions about resource allocation.
After 9/11 and the anthrax attacks that followed, public health agencies received federal funding to invest in software systems that pull together and analyze disease reporting in their regions, with the hope that they would help with early detection of bioterror attacks.
However, the jury is still out on how valuable these systems will be in that type of event detection. In more recent anthrax events involving imported animal skins, astute physicians and lab tests were key to diagnosis, but the syndromic surveillance systems were used to search for additional cases.
But syndromic surveillance has evolved as state and local public health emergency response efforts support epidemiologists in their work and start pulling data from multiple sources, said Charles Ishikawa, associate director of public health programs for the International Society for Disease Surveillance in Brighton, Mass. “The more evolved the system is, the more sophisticated the number and type of data sources they are able to draw on,” he said, “ranging from ED visits to 911 call centers, poison control to school absenteeism.”
In an emergency response setting, it can be just as valuable for incident commanders to know that they don’t have to redeploy resources as to know that they do, Ishikawa said. “You may have anecdotal information about gastrointestinal outbreaks during a hurricane, but if the syndromic surveillance data is not showing it, then you may not have to apply more resources. We are seeing it being used to track injuries in winter storms, hurricanes, infectious disease outbreaks, water quality alerts and during mass gatherings such as the Super Bowl.”
Pooling Data With EMS
Following up on the pilot project during the ice storm in Austin, Henry partnered with Pat Murphy, manager of the Business, Analysis and Research Team for Austin-Travis County EMS, on ways they could combine data. During a record drought and heat wave last summer, they pooled their information. “It was remarkable how well the two data sources tracked,” Henry said. The combined data from EMS and public health was used to determine whether the county needed to go to a second-phase alert, which includes opening cooling stations, she said. “We could triangulate with other data sources, including nonprofits that work with the homeless population.” They later partnered again when wildfires in a neighboring county impacted air quality in Travis County.
After an event, public health agencies traditionally review the event to measure its impacts, Henry said. “But this is giving us the opportunity to have an impact in the middle of an emergency when you have to start making decisions about deploying resources,” she said. “We worked so well with EMS that we had to ask ourselves why we haven’t always been doing this.”
NC Detect
Another statewide syndromic surveillance system that has grown more valuable over time is the North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT). It started in 1999 with a statewide mandate that hospital EDs electronically report chief complaint data. After 9/11 it took on more of a bioterrorism focus from an all-hazards point of view, said Amy Ising, the NC DETECT program director in the Carolina Center for Health Informatics at the University of North Carolina Department of Emergency Medicine. “We are constantly adding features to handle a variety of scenarios,” she said. “User feedback from local public health officials drives what we work on.”
NC DETECT now provides near-real-time statewide surveillance capacity to local, regional and state-level users across the state, with twice daily data feeds from 117 emergency departments, hourly updates from the statewide poison center, and daily feeds from statewide EMS runs and select urgent care centers. A Web application allows public health officials to access aggregate data as well as information customized to their jurisdictions. But the system does not automatically push out alerts. “We made a conscious decision not to do that,” Ising said. “Other states have tried that and the epidemiologist users get alert fatigue. When our users log in, they can see a dashboard view of what is important to them. We have made data easy to export so that in a situation like Hurricane Irene, the people in the emergency operations center can get situation reports and answer questions for the governor.”
NC DETECT is involved in all of North Carolina’s emergency tabletop exercises, said Dr. Lana Deyneka, director of the Statewide Enhanced Surveillance and Hospital-Based Public Health Epidemiologist programs in the North Carolina Division of Public Health. “We don’t have to create something after an emergency crops up. We should already have a module that fits, although we can add new queries in just a few minutes if we need to.” For instance, she said, the system already has a disaster module that can monitor categories like emergency department utilization.
Hurricane-related health outcomes might include acute illnesses, injuries, poisoning and behavioral health tracking. The system also has modules to track food poisoning, chemical exposures and radiological exposures.
Surveillance in the Great Lakes State
Michigan’s syndromic surveillance system was built around an open source software project originally developed at the University of Pittsburgh called the Real-Time Outbreak Detection Surveillance system, said Jim Collins, acting director of the Bureau of Epidemiology in the Michigan Department of Community Health.
The goal, he said, is to recognize and respond to unusual outbreaks of illness that may be the result of bioterrorism, outbreaks of infectious disease, or other public health threats and emergencies. “We now have 84 hospitals using the system and a parallel feed for all poison control centers,” Collins said.
Real-time detection of a notable increase in patients presenting similar symptoms could allow early and appropriate public health intervention and minimize negative impact, he added. The Michigan system provides tools that include automatic data collection and automatic aberration detection algorithms, as well as tools that support temporal and spatial data analysis and visualization.
“Initially it was designed with standards based on bioterrorism funding, but it is set up so we can do ad hoc queries on the fly,” Collins said. It is based on chief complaints that are entered in free text by emergency department clinicians, but that data also is run through a coder to group entries that might be heat related or to track an influenza outbreak, for instance. The syndromic surveillance system also feeds the Michigan Health Alert Network, a secure Web-based notification system, to alert key personnel.
At the Border
One project that focuses on information sharing is the Border Infectious Disease Surveillance (BIDS) program, a public health collaboration with Mexican officials to monitor infectious diseases at the border. With a network of clinic and hospital sites along the United States/Mexico border, the BIDS team created a shared syndromic case definition, and standardized data collection instruments allow for the exchange of surveillance data.
“One of the important lessons learned in the binational effort is that with local, state and federal participation, we can overcome political, cultural and infrastructural barriers,” said Orion McCotter, an epidemiologist with the Office of Border Health at the Arizona Department of Health Services, in an email interview. It is important to have binational communication plans for sharing of important public health information with the local agency, emergency response officials on both sides of the border and other federal agencies working at border points of entry, he added.
The BIDS program has been involved in numerous outbreak identifications and/or investigations, including severe acute respiratory infections (SARI). That project enhances the existing surveillance because data on influenza and other respiratory pathogens as causes of SARI and SARI-related mortality in hospital patients did not previously exist, McCotter said. “Integrating hospital-based surveillance for SARI with existing influenza surveillance was done to strengthen surveillance activities, build on preparedness strategies, control measures to combat the influenza and detect new virus subtypes or other etiologic agents of SARI.”
In another example of international monitoring, the National Poison Data System was used to track potential incident-related exposures in the U.S. resulting from the Japan earthquake and radiological incident of 2011 and disseminate public health alerts. Shortly after international media coverage began, many of the 57 U.S. regional poison centers started receiving calls regarding the Japan radiological incident, according to researchers from the Centers for Disease Control and Prevention (CDC). State and federal health officials used that input to help epidemiologists and toxicologists determine if a true exposure had occurred and to target risk communication messages to and information about the use of potassium iodide and other iodide-containing products intended for thyroid protection. The system also supplied data to the CDC Emergency Operations Center, which was activated from March 11 to April 18, 2011.
BioSense 2.0
CDC provides funding for state-level syndromic surveillance efforts and coordinates interoperability and data sharing between states.
Mandated in the Public Health Security and Bioterrorism Preparedness and Response Act of 2002, the CDC’s BioSense program was launched in 2003 to establish an integrated national public health surveillance system for early detection and rapid assessment of potential bioterrorism-related illness. Now the CDC is revamping BioSense with a cloud computing approach. The goal of BioSense 2.0 is to provide an easy-to-use, Web-based tool that health officials can use to monitor or assess syndromic activity within and beyond their jurisdictions.
“The redesign is so that we will have more sharing among jurisdictions,” said Dr. Taha Kass-Hout, deputy director for information science and the BioSense program and manager of the CDC’s Division of Notifiable Diseases and Healthcare Information. “One thing we found with BioSense 1.0 is that the data often bypassed state and local public health and went straight from hospitals to the CDC,” he said. “The states never saw the data. We have corrected that in the states where it was happening.”
While planning the redesign, Kass-Hout said the CDC looked at the needs of epidemiologists and emergency response and planning teams so it could provide as much flexibility and interoperability as possible. For instance, during the Gulf of Mexico oil spill, the U.S. departments of Defense and Veterans Affairs had a lot of the data and played a critical role.
In a case like a major hurricane, BioSense will let public health and emergency responders set up shared space to see data in real time. In fact, BioSense 2.0 already has shared spaces that involve state and local governments within a state, Kass-Hout added, as well as multiple states working together. “They can see more than trends,” he said. “This gives data to people on the ground who understand the context and can interpret it at their level.”
---
Public Health Surveillance
- Categorical surveillance: an active or passive system that focuses on one or more diseases or behaviors of interest to an intervention program.
- Integrated surveillance: a combination of active and passive systems using a single infrastructure that gathers information about multiple diseases or behaviors of interest to several intervention programs (for example, a health facility-based system may gather information on multiple infectious diseases and injuries).
- Syndromic surveillance: an active or passive system that uses case definitions that are based entirely on clinical features without any clinical or laboratory diagnosis (for example, collecting the number of cases of diarrhea rather than cases of cholera, or “rash illness” rather than measles). Because syndromic surveillance is inexpensive and is faster than systems that require laboratory confirmation, it is often the first kind of surveillance begun in a developing country.
Source: National Center for Biotechnology Information
---
Autor(en)/Author(s): David Raths
Quelle/Source: Emergency Management, 04.06.2012
Bitte besuchen Sie/Please visit:
