Journal of Applied Communication Research Vol. 35, No. 1, February 2007, pp. 9 25
Emergency Communication Challenges in Response to Hurricane Katrina: Lessons from the Centers for Disease Control and Prevention Marsha L. Vanderford, Teresa Nastoff, Jana L. Telfer & Sandra E. Bonzo
In response to Hurricane Katrina’s extensive destruction and related public health threats, the Centers for Disease Control and Prevention (CDC) deployed more than 1,000 staff to its emergency operations center and to affected areas. Among them were members of CDC’s Emergency Communication System. This paper describes the strategies and tactics used by health communication specialists during the pre-event, response, and post-event stages to address a range of emergency communication exigencies. It highlights three difficult challenges for CDC communication specialists during Hurricane Katrina: rapid dissemination of health messages; adaptation of health messages for diverse audiences, locations, and circumstances; and phasing of key risk messages during the emergency response. Keywords: Health Communication; Emergency Communication; Hurricane Response; Chaos Theory Hurricane Katrina made landfall three times between August 23, 2005, when it originated in the Southeastern Bahamas, and August 29, 2005, when it reached Marsha L. Vanderford (Ph.D., University of Minnesota) is Associate Director for Communication Science at the CDC’s National Center for Injury Prevention and Control. Teresa Nastoff (R.N., B.S.N., Ohio State University) is Lead of the Pandemic Influenza Communication System at the CDC’s National Center for Health Marketing. Jana L. Telfer (M.A., Syracuse University) is Associate Director for Communication Science at the CDC’s National Center for Environmental Health/Agency for Toxic Substances and Disease Registry. Sandra E. Bonzo (M.L.I.S., University of South Carolina) is Enterprise Communication Officer at the CDC’s Coordinating Center for Environmental Health and Injury Prevention at the CDC’s National Center for Injury Prevention and Control. Correspondence to: Marsha L. Vanderford, Centers for Disease Control and Prevention, National Center for Injury Prevention and Control, 4770 Buford Highway, MS K-65, Atlanta, GA 30341, USA. Email:
[email protected]. The authors acknowledge and appreciate the editorial and research assistance of Jamila Howard, D’Angela T. Green, Christy L. Cechman, Rebecca Myers, Sue Swensen, and Tracey Foster-Butler. ISSN 0090-9882 (print)/ISSN 1479-5752 (online) # The United States Government has retained certain worldwide rights in this work and it is published with permission. DOI: 10.1080/00909880601065649
10
M. L. Vanderford et al.
Louisiana. The hurricane’s winds produced storm surges greater than any previously recorded (Federal Emergency Management Agency, 2006). About 80% of New Orleans’s 485,000 residents were evacuated to escape flooding, and an estimated 1,220 deaths were blamed on the storm (Daley, 2006). Along the Gulf Coast, about 1.7 million households in Alabama, Florida, Georgia, and Mississippi were without power (U.S. Government, 2006). In response to the extensive destruction and related public health threats, the Centers for Disease Control and Prevention (CDC) deployed more than 600 staff members to provide technical assistance in affected areas. Another 500 were deployed to CDC’s emergency operations center to execute response plans, develop needed resources, provide off-site leadership about potential health risks, and mitigate adverse health effects.1 Along with providing science and medical professionals, CDC activated its Emergency Communication System (ECS). ECS is an all-hazards communication response unit comprising health communication, education, and public affairs specialists that provides a coherent communications framework for coordinating surge capacity in an emergency and ensures that critical health protection messages2 can be delivered to diverse audiences (e.g., clinicians, affected communities, and state and local public health officials) through multiple channels (e.g., Web, mass media outlets, hotlines, and CDC’s Health Alert Network). Previously, CDC had activated ECS in response to such health threats as SARS and the West Nile Virus, as well as Hurricanes Charley, Ivan, and Jeanne. As Hurricane Katrina approached, CDC was better prepared to deliver public health information than for prior disasters. CDC had responded to previous hurricanes and knew how to develop and disseminate multiple health protection messages concerning health hazards related to typical hurricanes. Katrina, however, was not typical. Soon after the hurricane struck, CDC realized Katrina’s catastrophic devastation challenged even the fundamental tenets of emergency communication practice proven to be effective in past emergency responses. The purpose of this report is to describe the strategies and tactics used by ECS during the pre-event, response, and post-event stages of the Hurricane Katrina to meet the challenges posed by the extended nature of the disaster and by the failure of sophisticated electronic communication channels. These include: (1) rapid dissemination of health messages; (2) adaptation of health messages for diverse audiences, locations, and circumstances; and (3) phasing of key risk messages during the emergency response. The conclusions address how lessons derived in the aftermath can be applied to future events, as well as how CDC’s experience is relevant to health communication research. This progression of communication challenges is presented as a case study that is tested against an application of chaos theory (Sellnow, Seeger, & Ulmer, 2002). We address two major questions. First, does chaos theory provide insight that can help health communication practitioners in responder agencies to understand their challenges and better prepare for the future? Second, does the CDC case study shed insight on the usefulness of chaos theory as a heuristic for describing decision-making for crisis communication among responder agencies?
Emergency Communication and Katrina
11
Pre-event Communication Activities Message Development, Review, and Adaptation At the onset of the 2005 hurricane season, CDC reviewed its health protection guidance for hurricanes and flooding and confirmed that it was up to date and prominently displayed on CDC’s Emergency Preparedness and Response Web site (http://www.bt.cdc.gov). Topics included hurricane readiness and recovery, drowning prevention, food and water safety, sheltering in place, prevention of carbon monoxide (CO) poisoning, wound care, and prevention of mosquito-borne illnesses. Over several hurricane seasons, CDC had conducted rudimentary audience analysis and adapted its health messages to reflect the literacy levels and cultural contexts of diverse audiences. Easy-to-read hurricane-related documents include Key Facts about Hurricane Readiness (eighth-grade level), Protect Yourself from Carbon Monoxide Poisoning after an Emergency (sixth-grade level), and Preventing Chain-Saw Injuries during Tree Removal after a Hurricane (sixth-grade level). In addition to reducing literacy levels, CDC translated messages concerning common hurricane-related health risks into Spanish, Haitian Creole, Vietnamese, French, Portuguese, and German. These easy-to-read translations have been used to develop announcer-read public service announcements (PSAs) and prerecorded audio messages. Message Dissemination In anticipation of the storm, CDC prepared health messages and information for state public health information officers and local news media in areas at risk of hurricane-related health threats. CDC delivered these messages in advance so that short-term power outages during or after the storm would not prohibit access to health information. The strategy of disseminating pre-developed and pre-positioned messages at the onset of an emergency is one that many public health and protection agencies recommend. This strategy is consistent with risk communication principles for rapidly meeting public need for information during emergencies (Covello, 2003; Rudd, Comings, & Hyde, 2003; U.S. Department of Homeland Security, 2004; World Health Organization, 2005). The impact of four successive hurricanes on Florida in 2004 emphasized the need to identify additional partners to channel health information into the hands of consumers. From formative research, and as one might expect, CDC identified meteorologists and weathercasters as primary and trusted sources of information about storms and other emergencies. As a result, CDC partnered with The Weather Channel (TWC). Early in the 2005 hurricane season, TWC posted a health section on its Web site with links to CDC health information. This TWC Web feature, along with ‘‘crawlers’’ at the bottom of TWC television broadcasts, was in place throughout the 2005 hurricane season. CDC also initiated partnerships with major home-improvement chains to explore the use of prerecorded PSAs and simple educational materials to describe how CO poisoning can be prevented, especially when using heaters, power washers, or
12
M. L. Vanderford et al.
gas-powered generators. Educational materials were placed in stores near these appliances. At the beginning of the 2005 hurricane season, CDC established a partnership with Lowe’s Companies, Inc., and Lowe’s adapted CDC’s PSAs for use in its stores in Florida. During the 2005 tax-free holiday that opens hurricane season in Florida, Lowe’s played the announcements hourly. Stores in hurricane-prone areas continued to play the announcements every two hours throughout the season, increasing the play to hourly if a hurricane was forecast. Communication Activities and Challenges during Emergency Response As the hurricane made landfall, flooding followed, and massive power outages occurred, three primary challenges faced CDC’s communication specialists. The storm and its aftermath seriously compromised CDC’s ability to disseminate health and safety messages rapidly, adapt messages to diverse communities, and phase messages to address evolving needs for health information. Rapid Dissemination Despite extensive preparation and ready-to-go health messages, CDC experienced challenges in providing information to affected areas immediately after the storm. Power outages were more extensive and sustained than in previous response efforts, which prohibited rapid dissemination of health messages to the public and customary public health intermediaries. CDC’s reliance on electronic channels (e.g., Web sites, radio, and television) severely hampered its ability to deliver needed health information. Awareness that this response had shifted from emergency to crisis came in the first week of September when CDC tried to provide requested health protection information for emergency responders to a U.S. Coast Guard officer in Alabama. The officer had no ability to connect to the Internet to download existing information. For several days, CDC communication specialists tried alternative ways of delivering printed copies, but found that overnight delivery services could not reach the area. No operational fax machines or copiers were nearby to receive the documents. Further, a CDC truck chartered to deliver thousands of printed copies was turned back to Atlanta because roads were impassable. Localizing communication efforts. Faced with no electronic means to reach affected areas, CDC communication specialists turned to local, face-to-face channels to deliver health protection messages. Consequently, as transportation into affected areas gradually became available, CDC deployed about 30 health communication, health education, and public affairs specialists to local and state health departments in Louisiana, Mississippi, and Texas. The deployed staff served as liaisons between the Atlanta-based ECS and remaining local populations needing health protection information. Although CDC had responded to previous disasters by sending staff, such as scientists and epidemiologists, to the field, this was the largest deployment of health communication and education specialists in agency history.
Emergency Communication and Katrina
13
The health communication, education, and public affairs specialists went to evacuation centers and emergency response centers along the Gulf Coast to identify and fill information needs. They helped Atlanta-based staff to develop and adapt needed health messages. As needs for information emerged, the specialists identified and used whatever local communication channels were operational. Often, they assisted the local agencies by hand-delivering printed copies of health guidance to workers and affected communities. Developing new channels for communication. Reaching populations was complicated not only by the lack of electricity, but also by the dispersion of many people who no longer had access to routine communication channels. Deployed CDC communication staff worked with state and local agencies to identify likely places where evacuees and returning residents could access such resources as ice and water. These sites also became distribution points for health information involving what individuals trying to survive in communities without basic infrastructure needed to know: preventing CO poisoning, avoiding electrical hazards, avoiding injuries during clean-up, and managing stress to prevent violence. CDC also delivered these key messages and related products at American Red Cross (ARC) feeding tents, hurricane Disaster Recovery Centers, and Salvation Army service centers in Alabama, Louisiana, and Mississippi. In addition to this distribution, the field communicators provided valuable feedback to CDC concerning localized issues. For example, staff in Disaster Recovery Centers requested information about preventing and treating dog bites. In urban and suburban areas, many people had evacuated without their pets. As a result, dogs roamed affected neighborhoods. CDC developed an easy-to-read fact sheet and PSAs to help residents avoid and treat dog bites. These PSAs were broadcast on radio because it was the first electronic medium to reactivate. Disseminating consumer information through partnerships. Along with Disaster Recovery Centers, the American Red Cross, and the Salvation Army, two other partnerships played important roles in the Katrina response: faith-based organizations (FBOs) and major home-improvement retailers. Before Hurricane Katrina, CDC had initiated outreach to these groups, but it was the response to Katrina that extended and deepened the partnerships. Katrina’s devastation led many FBOs to supplement the work of recognized relief agencies, such as the American Red Cross and Salvation Army. Local churches delivered food and other household items to evacuees and to people whose belongings had been damaged in the storm. Some FBOs opened their doors as non-governmental evacuation centers or provided volunteers to assist other facilities. The relationship established between local FBOs and those seeking assistance made these organizations trusted sources of health information. FBOs were already integrated into community networks and represented a potential critical path for the flow of health messages during emergencies. The ability of FBOs to deliver nonemergency health promotion information is well documented (Brooks & Koenig, 2002; DeHaven, Hunter, Wilder, Walton, & Berry, 2004). Existing links between FBOs, the U.S. Department of Health and Human Services (HHS), and CDC were
14
M. L. Vanderford et al.
used to offer pertinent health information. CDC sent information to approximately 300 FBOs during the response. In turn, the FBOs made CDC aware of key communication resources and distributed health messages to underserved populations. Prior to Hurricane Katrina, CDC’s partnership with Lowe’s retail stores was limited to Florida. After Katrina, many Lowe’s outlets throughout the Gulf Coast were damaged and/or inaccessible. As power was restored and water receded, some Lowe’s stores reopened, and CDC worked with the chain to expand the partnership for communicating health protection information to stores outside Florida. Despite CDC’s strong working relationship with Lowe’s corporate office, extending this partnership took longer than anticipated. Communication in retail chains often occurs through district offices which, in this instance, had to be oriented to the health information partnership while they responded to their own disaster recovery and tried to service their customers. By October, when many Gulf Coast communities began rebuilding, Lowe’s operations in Texas were broadcasting CO poisoning prevention messages and had requested safety messages about chain-saw use for distribution in their stores. At the federal level, CDC formed a partnership with the U.S. Consumer Product Safety Commission (CPSC) in the immediate aftermath of the storm. This helped in distributing materials directly to homes in the hurricane zone. CPSC adopted CDC’s health messages and graphics*rebranding, reprinting, and distributing thousands of CDC’s CO poisoning prevention materials among CPSC agencies and industry partner representatives in the field. These representatives gave the information directly to consumers at home or in communities they were visiting. In addition to reaching residents in communities with limited communication channels and infrastructure, CDC was also challenged to reach displaced residents in evacuation shelters. Typically in disasters, shelters or evacuation centers are established and operated for only short periods. Following Katrina, the large number of long-term shelters and evacuation centers introduced new communication challenges. The extended congregate-living arrangements created the potential for new health problems, as residents had less access to communication channels routinely used to disseminate health messages. In Texas, where electrical power was available in evacuation centers, CDC communication specialists used the broadcast capacities of large stadiums housing captive audiences of evacuees. At the communication specialists’ request, CDC in Atlanta produced about 30 videotaped PSAs on a range of public health issues specific to occupants of evacuation centers (e.g., hygiene, stress management, infection control, and immunizations). The initial video PSAs were delivered to specific evacuation centers that could be reached by express delivery. As more evacuation centers recovered electricity and connectivity, the videos were incorporated into the media content for evacuation centers and delivered to dozens of centers in affected areas via the Katrina Information Network (KIN). This 24-hour broadcast video communications channel, launched September 14, 2005, provided critical response and recovery information relating to hurricanerelated issues. Initially available on the DISH network, KIN evolved to become the
Emergency Communication and Katrina
15
Emergency Response Information Network (ERIN), accessible via DISH, DirecTV, and online at http://www.erintv.org. Message Adaptation for Local Use Over the course of several hurricane seasons, CDC had conducted rudimentary audience analysis and adapted its health messages to reflect literacy levels and the cultural contexts of diverse audiences. CDC routinely partners with local and state public health departments to disseminate health messages through those agencies to increase CDC’s credibility with local populations. Fortunately, CDC has been consistently acknowledged as a source of reliable health information by a majority of the lay public (Pollard, 2003). Despite these preparations, audience analysis conducted by communication specialists during the Hurricane Katrina response indicated that substantially more adaptation was necessary. Most work performed in response to Katrina by Atlantabased and deployed CDC communication specialists involved adapting health messages developed before the storm for use with local populations affected by Katrina. Having staff in affected areas enabled CDC to assess and respond in real time to diverse communication needs at evacuation centers, service centers, reentry points, and local health departments. During the Katrina response, CDC modified formats and restructured its health messages to meet different needs. For example, CDC created door hangers for doorto-door delivery of CO poisoning prevention materials, stickers for children in evacuation centers to remind them about hand hygiene, and one-line messages for high-frequency radio broadcasts. Fact sheets were compressed and combined to create one-pagers customized to address multiple health concerns for residents returning to storm-damaged communities. Almost every health message on CDC’s hurricane Web site was reformatted, translated, recombined with other messages, or altered in some way to serve audiences along the Gulf Coast better. Because so many CDC staff had been deployed in multiple localities, the volume of requests for customization was overwhelming. Typically, local and state agencies adapt CDC information for targeted distribution. In this response, however, because the health communication and education capacity in local and state agencies varied widely and communities lacked electronic infrastructure, local responders were severely limited in their ability to adapt CDC messages to their needs. Easy-to-read messages. A primary request for adaptation included rewriting messages for lower literacy levels. In Louisiana, for example, 28% of the population reads at level I on the five-level reading classification system developed by the U.S. Department of Education for its National Assessment of Adult Literacy (NALS; National Center for Educational Statistics, 1993). Within the City of New Orleans, 3344% of the population read at level I (Literacy Initiative, 2002). Level I includes people who are capable of reading short passages of text to locate a single piece of easily identifiable information.
16
M. L. Vanderford et al.
Although CDC has worked for several years to transition its consumer-destined documents to appropriate reading levels for lay audiences, most of CDC’s public health recommendations and guidance are above an eighth-grade reading level. Typical government documents were unsuitable for the large population displaced by Katrina. Storm-related health threats called for swift retooling of CDC’s key emergency messages. Staff working in Texas evacuation centers further challenged CDC by requesting health messages that were primarily visual, or pictograms, instead of using pictures to enhance a text-based message. Pictograms are stand-alone pictures that convey health information with little or no text. To meet the need of low-literacy audiences, CDC followed the guidance in resources such as the National Cancer Institute’s (2003) Clear and Simple: Developing Effective Print Materials for Low-Literate Readers. Text was translated into pictures, and some pictures were simplified to line drawings or pictograms that incorporated international symbols. Between September 29, 2005, and October 7, 2005, ECS created easy-to-read versions and pictograms for topics ranging from mold clean-up to hand sanitation. Pictograms created a special challenge to reconciling scientific accuracy with the circumstances of individuals who would be using the information. The development of CDC’s chain-saw injury prevention messages illustrates this tension. As a result of input from occupational scientists, the first drafts of easy-to-read chain-saw safety messages included images and text recommending safety attire, such as pants, boots, and long-sleeved shirts. However, these scientific recommendations were not acceptable along the hot, humid Gulf Coast where people had no air-conditioning and wore T-shirts, shorts, and sandals. In addition, deployed CDC communication specialists observed that residents largely ignored the chain-saw safety messages because they included recommendations involving costly personal protective equipment including ear and eye protection. As negotiation to adapt the science behind the safety recommendations for Hurricane Katrina audiences continued, graphic artists working with ECS experimented with many versions of the images. It required 10 days to work from concept through revisions and approval to develop pictograms that included accurate, credible guidance that was also easy to understand and culturally appropriate. Adaptation to increase credibility. CDC also modified health messages to increase their credibility with affected audiences in affected areas. On the basis of feedback from health communication staff assessing the attitudes of affected communities, CDC realized that its health guidance should be attributed to local sources to increase acceptability and achieve greatest impact. Many Gulf Coast residents blamed the federal government for slow and inadequate response to the storm. Opinion polls conducted during the aftermath revealed that up to 70% of Americans believed the federal government’s response to Katrina was too slow; for 56%, the response had resulted in a loss of confidence in the government’s ability to respond to emergencies (Purdum & Connelly, 2005). In the wake of what many perceived as a failed federal response, some state and local agencies thought local residents would not find health
Emergency Communication and Katrina
17
information from federal agencies credible. To overcome this perception and relay CDC’s health protection messages, CDC helped state and local agencies replace the CDC logo with attribution to local agencies. Adaptation for specific circumstances. Some health messages were modified to suit particular living circumstances. As CDC’s staff conducted on-site needs assessments in evacuation centers, members anticipated the potential for stress and related injuries among families living for extended periods in close quarters with little privacy or productive activity. Although CDC had already developed prevention messages concerning intentional injuries, these messages were not in forms that addressed the circumstances within the evacuation centers. To meet the needs of evacuees, CDC developed simple, easy-to-read materials for handling stress that would decrease the potential for injuries from bullying, shaken baby syndrome, suicide, or sexual abuse. CDC also created a set of ‘‘palm cards’’ similar to playing cards. These contained simple prevention messages with colorful and illustrative graphics addressing stressful situations relevant to evacuation. The topics included ‘‘Stress and Relationships,’’ ‘‘Parenting under Stress,’’ ‘‘When Your Baby Cries,’’ ‘‘Preventing Violence,’’ ‘‘Rape Prevention,’’ and ‘‘Suicide Prevention.’’ Upon completion, CDC staff evaluated the cards and found them to be acceptable for use with intended audiences. The cards were later adapted for Hispanic and Vietnamese evacuees. CDC developed two versions of each card, one with the HHS and CDC logos and one without logos, adaptable for local rebranding. Message Phasing Most international, national, state, and local government emergency response plans include recommended communication activities for the pre-event, response, and recovery stages of an emergency (DeKalb County Board of Health, 2006; State of California Department of Health Services, 2002; U.S. Department of Health and Human Services, 2005; U.S. Department of Homeland Security, 2004; World Health Organization, 2005). CDC’s Emergency Communication System operations plan includes three pre-event planning stages, but only one for response, and it includes a single set of communication activities. ECS is typical of most emergency communication plans in treating the response stage as a unified whole. In an extended emergency, such as Hurricane Katrina, however, overlooking the multiple, distinct phases that constitute the response stage is a major oversight. The circumstances surrounding Katrina prompted CDC to question the adequacy of emergency plans that fail to recognize and plan for a phased communication response stage. The emergency response stage of Hurricane Katrina lasted for more than a month. Katrina, which struck on August 29, 2005, was accompanied by severe flooding in New Orleans and was followed by Hurricane Rita on September 24, 2005. Before recovery from one disaster could begin, two others followed. Communication tasks emerged and evolved as the incident focus changed. For example, health information was initially needed to protect people from storm winds and falling debris. Later, guidance was needed to prevent drowning, avoid driving in flood waters, protect
18
M. L. Vanderford et al.
against electrical hazards, and prevent CO poisoning. As displaced populations packed evacuation centers, other health information needs emerged: for instance, ways to promote hygiene, control infection, and store food safely. Extended stays for victims in evacuation centers created a need for information about management of chronic disease and stress reduction. Hurricane Katrina demonstrated the inadequacy of emergency communication response plans that fail to recognize the possibility of second (and third) scenarios and the certainty of subsequent health threats in extended disasters. CDC’s initial strategy of pre-event dissemination of comprehensive hurricanerelated health messages covering the anticipated needs of the entire response period was ineffective to meet local needs. As the response to Hurricane Katrina lengthened, CDC received requests for health protection messages that had been disseminated at the onset of the emergency. In their urgency to meet immediate needs after the storm, agencies in affected areas focused on specific needs of the moment and could not recall the pre-positioned information received before the storm. The need for a comprehensive body of messages phased in over time had not been anticipated at any level of government. Post-event Emergency Communication Activities An analysis of CDC’s emergency communication performance revealed three major challenges that require attention. Since fall 2005, CDC’s communication staff has worked to meet the challenges of low-technology delivery systems for health information, to develop a system for easier and faster adaptation of hurricanerelated messages, and to release disaster-related health information in a phased approach*particularly in an extended emergency. Improving Low-tech Information Delivery Health information can be disseminated rapidly to targeted audiences when messages have been prepared in advance, distribution channels of varying levels of technological sophistication have been developed, and partnerships with responder agencies are in place. CDC, like other federal agencies, is increasingly dependent on electronic formats for delivering information to the public. Hurricane Katrina demonstrated the error of exclusive dependency on high-technology delivery systems. Post-Katrina, CDC has focused on developing channels for delivering print information directly to primary audiences in disaster-affected communities. In view of lessons learned in the absence of electronic channels during the Katrina response, CDC now acknowledges that local organizations are essential links between residents at disaster/recovery sites and federal agencies trying to disseminate health protection information as effectively as possible. CDC is still in the process of solidifying and formalizing its partnerships with local intermediary groups encountered in the later stages of the Katrina, including the American Red Cross, Salvation Army, home improvement retailers, CPSC, and FBOs.
Emergency Communication and Katrina
19
CDC has learned that partnerships can be cultivated only by investing considerable time and resources, both human and fiscal. Clearly, CDC needs to increase its understanding of partner needs and interests to motivate their cooperation. For example, in a pre-Katrina meeting with representatives of various FBOs, it became apparent that CDC’s urgent agenda to produce and distribute health information was not the FBOs’ highest priority. To develop effective and reciprocal relationships, government agencies must identify mutual goals and acknowledge the priorities, criteria, and resource needs of partner organizations. Such relationships should be developed before an emergency occurs and assistance is expected. Since Katrina, CDC has contacted FBOs and other organizations with the capacity to assist CDC’s emergency response to identify common ground and goals. The process of building linkages between CDC and future disaster sites through partner organizations is ongoing. Developing a System for Faster Adaptation Hurricane Katrina also demonstrated the need to facilitate local adaptation of general health information developed from a national perspective. In fall 2005, the agency relied heavily on input from deployed personnel for local information needs; audience perceptions, attitudes, and knowledge; and other adaptation requirements. Immediately after Hurricane Katrina, health departments lacked the technological capacity to produce or adapt CDC’s Web-based health messages. Later, even when electricity was restored to some communities, health departments lacked personnel to modify health messages and meet local demands for information. In Katrina’s aftermath, CDC has searched for a system to help adapt its messages for affected communities. It has explored whether an automated system could tailor communication materials for staff and partners. Such a system could have key documents that allow images, including logos, to be inserted and deleted easily, an image library with visuals appealing to different audience segments, editions reflecting different literacy levels, and versions with and without CDC and HHS logos. This system might also enable new versions of critical messages to be created on the fly, while maintaining consistent messaging. Although power outages would prohibit health departments from accessing the system, CDC staff members in Atlanta would be able to meet adaptation needs of local communities rapidly, as delivery channels became available. In the absence of an automated system, CDC’s Emergency Preparedness Web site has been organized to include multiple versions of the same messages segmented by literacy level, by audience (clinicians, public, public health workforce, news media, and emergency responders), context (schools, clinics, and evacuation centers), and source (with and without CDC logos). Completing the matrix of messages is an ongoing activity, but as it is completed, CDC staff and partner organizations should find an enhanced library of health messages that will undoubtedly meet local needs better than was the case at the onset of Katrina.
20
M. L. Vanderford et al.
Rethinking Phasing of Message Dissemination Katrina response and recovery demonstrated that CDC’s standard practice of preevent notification to news media and partner organizations was insufficient in an extended emergency. A comprehensive pre-event dissemination strategy works well when local agencies have time to select relevant messages from CDC’s array of available materials and, thereby, harness the power of CDC’s science. This dissemination strategy is also effective when the duration of a crisis is short, and sustained media attention is unnecessary. On the other hand, this strategy is not effective when local agencies and responders are confronted with a series of immediate, focused, and urgent health threats; face numerous competing non-health issues; and have priorities that preclude reviewing, selecting, and adapting comprehensive CDC information to suit local needs. Before an event, mass media outlets may be most effective for communicating information; however, the aftermath of an event may leave communities with limited access to technology-based communication channels. If displaced populations move from disaster sites to evacuation shelters, the context in which information is received will require different channels and tactics. Over the course of extended crises, affected populations, and the agencies that serve them, can only address immediate needs. Early in the response, for example, primary needs include accessing clean water, shelter, and security. Until those primary needs are met, providing information on such topics as chain-saw injury prevention, stress management, and mold clean-up is overwhelming and counterproductive. Since fall 2005, CDC has reviewed information needs identified by staff, hotline calls, media analysis, and production/dissemination timetables during Katrina. On the basis of that information, the agency has developed a multiple-phased approach to emergency response for extended disasters relating to hurricanes and flooding: (1) the period immediately preceding the storm through the first 24 hours after the storm; (2) one to three days after the storm; (3) three to seven days after the storm; (4) two to four weeks after a disaster; and (5) one month and after. Table 1 shows selected topics addressed during the 2005 response to Katrina arrayed by recommended staging within the response period. Of course, the specific phases are not entirely predictable; future disasters would require variations of these phases. In addition, media analysis and epidemiologic surveillance during a disaster should reveal whether messages disseminated in early phases need to be reissued later to address persistent health threats or previously ineffective or unevenly distributed health information. However, many informational needs can be predicted and phased on the basis of prior experiences with hurricanes and flooding. The phased approach to disaster planning can help communication responders simultaneously delivery health information to meet immediate needs and forecast long-term needs.
Emergency Communication and Katrina
21
Table 1 Phased Message Dissemination for Hurricanes and Floods Period of dissemination Immediately preceding landfall through first 24 hours after the storm
1 3 days after the storm
3 7 days after the storm
2 4 weeks after the storm
1 month and after the storm (emphasis is on long-term health consequences)
Topics Hurricane readiness, preparation for power outages, preparation related to prescription medications, evacuating the area of a hurricane, staying safe in your home during a hurricane, worker safety in a power outage, CO poisoning prevention, flood readiness, electrical safety, prevention of heat-related illnesses, hand hygiene in emergency situations, coping with a traumatic event, emergency wound care, protecting your pets, animals in public evacuation centers Re-entering your flooded home, how to clean a flooded home safely, worker safety after a flood, preventing chain-saw injuries during tree removal, preventing injuries from falls (ladders/roofs), personal protective equipment and clothing for flood response, managing acute diarrhea after a natural disaster, cleaning and sanitation after an emergency, keeping food and water safe after a natural disaster or power outage Protection from animal- and insect-related hazards, electrical safety and generators, infection control and prevention in evacuation centers, impact of power outages on vaccine storage and other medicines, preventing violence after a natural disaster, animal disposal after a disaster Rodent control after hurricanes and floods, trench foot or immersion foot, environmental health needs and habitability assessments, protection from chemicals released during a natural disaster, respiratory protection for residents re-entering previously flooded areas and homes Suicide prevention, issues surrounding school-age hurricane evacuees attending new schools, mold removal from flooded homes, mold allergies related to flood clean-up
Implications for Crisis Communication and Chaos Theory CDC’s communication response has implications for communication scholars and researchers who study emergency communication or chaos theory (the idea that systems, no matter how complex and seemingly unpredictable, rely on an underlying order and that even small, simple changes in a system can bring about vastly different outcomes; Comfort, Sungu, Johnson, & Dunn, 2001; Matthews, White, & Long, 1999). Using a case study approach, Sellnow et al. (2002) demonstrated how chaos theory increased understanding of crisis communication during the 1997 Red River Valley flood in the Midwest. They noted how flood-related communication amid uncertainties and ambiguities affected relationships between responder agencies and forged new organizational links. This interplay between public and interorganizational communication provides a broad view of crisis communication in a single incident. CDC’s communication activities before, during, and after Hurricane Katrina present another opportunity to test the usefulness of chaos theory as an analytic framework in which to view crisis communication. However, our description of CDC’s response concerns the activities of a specific communication team within a
22
M. L. Vanderford et al.
larger response agency, as well as the interplay between that team’s decision-making and challenges as the public health communication infrastructure oscillated between chaos and order. The focus of the Red River Valley case study was external, whereas that of the present case study is largely internal. Does chaos theory provide insight that can help health communication practitioners in responder agencies to understand their challenges and better prepare for the future? And does the CDC case study offer insight into the theory’s utility for describing decision-making for crisis communication among responder agencies? The preceding discussion suggests that chaos theory does, in fact, help explain how CDC’s Emergency Communication System staff experienced and met the challenges of fall 2005. The concepts of bifurcation, sensemaking and cosmology episodes, and self-organization have particular utility. Bifurcation refers to ‘‘flashpoints’’ of change that fundamentally disrupt a system’s direction, character, or structure (Matthews et al., 1999; Sellnow et al., 2002). Bifurcation describes CDC’s experience when disrupted power systems, downed Internet capacity, disabled transportation systems, and extensive displaced populations blocked the agency’s ability to deliver health information*especially in the week after Katrina’s landfall. CDC communication staff were unable to share health information archived on CDC’s Web site, could not contact state public information officers in affected areas, and could not disseminate health messages via prepared radio or television PSAs. It was at this point that CDC communicators’ response changed in fundamental ways. They realized that having information to protect people’s health is meaningless in the absence of contemporary, flexible, delivery systems. In the days before and after Katrina’s landfall, communication staff members responded as they had to previous hurricanes, expecting quick repair to disruptions in communication infrastructure and intact delivery channels. As the extent of damage became apparent and flooding ensued, CDC’s communication response was hampered by unforeseen challenges. The concepts of sensemaking and cosmology episode help explain why these challenges emerged (Sellnow et al., 2002). Sensemaking refers to making sense of one’s circumstances on the basis of past experiences and personal interpretation. Cosmology episodes occur when one’s experiences, or the established rules of order, have no relevance to current scenarios, which, in turn, leads to confusion, uncertainty, and irrational behavior. A cosmology episode is an extreme collapse in sensemaking (Weick, 1993). In the case of Hurricane Katrina, CDC communicators responded according to their experience with previous hurricanes, but Katrina was not like other hurricanes. Failure to account for the exceptional nature of the event led to an extended, unfounded, and non-productive reliance on routine delivery of health information. As a result, adaptive responses to find alternate delivery channels were delayed. In the following weeks, CDC’s communication response demonstrated elements of self-organization, ‘‘whereby order re-emerges out of a random and chaotic state’’ (Sellnow et al., 2002, p. 272). During Hurricane Katrina, the leadership between field staff and health communication specialists at Atlanta headquarters reversed, and as
Emergency Communication and Katrina
23
the Emergency Communication System members redefined how they accomplished their mission, a new order arose. As CDC staff members established themselves in evacuation centers, disaster dissemination centers, and public health departments, they became the central authorities regarding what health information was necessary and how to disseminate it locally, even when power was eventually restored. Communication specialists at CDC headquarters in Atlanta responded to requests and directions from the field rather than taking the lead. Reliance on national networks of Internet, radio, and television to deliver health information was replaced with hand-delivery and local dissemination of health messages. PSAs, videotaped in Atlanta, were hand-delivered and then broadcast in local hardware stores and evacuation centers. Although CDC had anticipated that other federal agencies and public health departments would serve as dissemination channels, these expectations were not met; instead, the Salvation Army and local suppliers became more important and reliable partners. As flood waters receded and roads reopened, CDC delivered truckloads of printed health information to its new partners for hand-delivery to affected populations. Conventional ‘‘hard’’ communication systems and formal hierarchical relationships gave way to ‘‘soft’’ systems that emerged to meet contextual demands. Sellnow and colleagues (2002) suggest that chaos theory, despite its focus on collapsed sensemaking, bifurcations, and cosmological episodes, ultimately ‘‘seeks order and predictability’’ (p. 270). When viewed cyclically over extended periods, patterns emerge. CDC’s case study of Hurricane Katrina supports the notion that ‘‘predictable unpredictabilities’’ exist within novel, apparently disorganized communication systems. Although Hurricane Katrina had extensive novel characteristics that challenged CDC’s ability to disseminate and adapt health information rapidly, the backbone of the agency’s communication response had been appropriately informed by previous disasters, the dynamics of which were also chaotic at the time (i.e., Hurricanes Charley, Ivan, and Jeanne). Health guidance developed in response to those events formed a comprehensive core of relevant protection messages. Thus, CDC spent most of its time modifying, not creating, health messages for Katrina, supporting the notion that some elements of a chaotic system (a hurricane) are predictable. Other patterns may emerge over time. Will the challenges posed by Katrina better prepare responder agencies for future disasters? Probably so. Hurricanes, earthquakes, tsunamis, and floods are natural disasters that can evolve into large-scale crises affecting thousands of people and spawning second and third disaster scenarios. Certainly, terrorism could involve multiple or long-term events, such as multiple bombings, extended consequences of a nuclear event, or sustained disruption of electronic or computer systems. Phased dissemination of messages greatly helps organize and sustain communications in an extended disaster, which may be Katrina’s most enduring lesson for communication professionals. Hurricane Katrina also serves as a reminder of the impossibility of preparing completely for disasters, rendering the most extensive preparations inadequate. Katrina, through the lens of chaos theory, illustrates the dynamic nature of
24
M. L. Vanderford et al.
communication during crisis and the need to adapt systems, procedures, channels, and messages. Communication in this adaptive form is both a product and a facilitating factor in self-organization. Notes [1]
[2]
For the purposes of this paper, CDC’s emergency communication response to Hurricane Katrina, subsequent flooding, and Hurricane Rita have been combined because the staff and strategies employed were integrated into a single unit and system. In this case study, references to Hurricane Katrina should be interpreted as applying to the related disasters as a whole. Health protection messages referred to in this paper include disease prevention, safety and health promotion, and injury prevention and control information and guidance.
Author Note The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention/ Agency for Toxic Substances and Disease Registry. CDC’s ECS, and the health communication and health education staff in the National Center for Environmental Health and National Center for Injury Prevention and Control, must be recognized for their dedicated work during CDC’s response to Hurricane Katrina. Their work is described in this article. References Brooks, R. G., & Koenig, H. G. (2002). Crossing the secular divide: Government and faith-based organizations as partners in health. International Journal of Psychiatry in Medicine , 32 , 223 224. Comfort, L. K., Sungu, Y., Johnson, D., & Dunn, M. (2001). Complex systems in crisis: Anticipation and resilience in dynamic environments. Journal of Contingencies and Crisis Management , 9 (3), 144 158. Covello, V. T. (2003). Best practices in public health risk and crisis communication. Journal of Health Communication , 8 , 5 8. Daley, W. R. (2006). Public health responses to Hurricanes Katrina and Rita *Louisiana, 2005. Morbidity and Mortality Weekly Report , 55 (2), 29 30. DeHaven, M., Hunter, I. B., Wilder, L., Walton, J. W., & Berry, J. (2004). Health programs in faithbased organizations: Are they effective? American Journal of Public Health , 94 , 1030 1036. DeKalb County (GA) Board of Health. (2006). Center for public health preparedness . Retrieved January 17, 2006, from http://www.dekalbhealth.net/cphp/about-cphp.htm Federal Emergency Management Agency (FEMA). (2006). Hazards: Floods . Retrieved January 17, 2006, from http://www.fema.gov/hazards/floods/ Literacy Initiative. (2002). Final report of the Literacy Alliance of Greater New Orleans . Retrieved January 17, 2006, from http://www.literacyalliancegno.org/ Matthews, M. K., White, M. C., & Long, R. G. (1999). Why study the complexity sciences in the social sciences? Human Relations , 25 , 439 461. National Cancer Institute, National Institutes of Health (NIH). (2003). Clear and simple: Developing effective print materials for low-literate readers (NIH Publication No. 95-3594). Retrieved January 17, 2006, from http://www.cancer.gov/aboutnci/oc/clear-and-simple
Emergency Communication and Katrina
25
National Center for Educational Statistics. (1993). Adult literacy in America: A first look at the findings of the National Adult Literacy Survey. Retrieved January 17, 2006, from http://nces. ed.gov/pubsearch/pubsinfo.asp?pubid /93275 Pollard, W. (2003). Public perceptions of information sources concerning bioterrorism before and after anthrax attacks: An analysis of national survey data. Journal of Health Communication , 8 , 93 103. Purdum, T. S., & Connelly, M. (2005, September 15). Support for Bush continues to drop as more question his leadership skills. New York Times , p. A18. Rudd, R. E., Comings, J. P., & Hyde, J. N. (2003). Leave no one behind: Improving health and risk communication through attention to literacy. Journal of Health Communication , 8 , 104 115. Sellnow, T. L., Seeger, M. W., & Ulmer, R. R. (2002). Chaos theory, informational needs, and natural disasters. Journal of Applied Communication Research , 30 , 269 292. State of California Department of Health Services. (2002). Bioterrorism surveillance and epidemiologic response plan . Retrieved January 17, 2006, from http:/www.dhs.ca.gov/ps/ dcdc/bt/pdf/ca_bt_surv_epi_plan-2002b.pdf U.S. Department of Health and Human Services. (2005). HHS pandemic influenza plan, supplement 10: Public health communications . Retrieved January 17, 2006, from http://www.hhs.gov/ pandemicflu/plan U.S. Department of Homeland Security. (2004). National response plan: Public affairs support annex . Retrieved January 17, 2006, from http://www.dhs.gov/interweb/assetlibrary/ nrp_fulltext.pdf U.S. Government. (2006). FirstGov.gov. Retrieved January 17, 2006, from http://answers.firstgov.g ov/cgi-bin/gsa_ict.cfg/php/enduser/std_alp.php?p_sid /ktooisbh Weick, K. E. (1993). Enacted sensemaking in a crisis situation. Journal of Management Studies , 25 , 305 317. World Health Organization (WHO). (2005). WHO global influenza preparedness plan . Retrieved January 17, 2006, from http://www.who.int/csr/resources/publications/influenza/who_cds_ csr_gip_2005_5/en/
