Coordination of Multiple States and Federal Agencies
Timothy Jones and Craig Hedberg
- Rapidly Determining when an Outbreak is Multijurisdictional
- Identifying Potentially Involved Jurisdictions and Agencies
- Establishing Effective and Efficient Coordination Among Jurisdictions
- Clearly Identifying a Lead Outbreak Investigation Coordinator
- Monitoring Roles and Activities in Different Jurisdictions
- Coordinating Public Communications Across Multiple Jurisdictions
- Coordinating Evaluation and Reporting of the Outbreak Investigation
Multijurisdictional outbreaks are being increasingly identified. As laboratory technologies for disease surveillance advance and data-sharing tools for interagency communication improve, potential links among widely dispersed cases are more effectively identified. Among foodborne disease outbreaks alone, during 2010–2015, a total of 157 multistate outbreaks were reported, accounting for approximately 10,000 illnesses and 77 deaths (1). However, multistate outbreaks are not limited to foodborne diseases and have been associated with an array of etiologies and sources. Examples include fungal infections associated with contaminated methylprednisolone injections, initially recognized in Tennessee and later reported in 20 states (2); a measles outbreak associated with exposure at an amusement park in California which spread to 17 states (3); a widely dispersed outbreak of tuberculosis among travelers to major metropolitan areas (4); and a combined legionellosis/Pontiac fever outbreak among returning guests living in 26 states and exposed at a hotel in Atlanta (5). Such investigations require cooperation among epidemiologists, laboratory personnel, environmentalists, and regulatory agencies at local, state, and federal levels. Preparing for efficient and effective multijurisdictional coordination is critical for successfully responding to such outbreaks. The success of these efforts should be measured by their ability to identify the etiologic agent, mount a prompt public health intervention, and effectively control the outbreak (6).
In addition to steps involved in all outbreak investigations to identify the source of the outbreak and monitor its progression, particular considerations during multijurisdictional outbreaks include the following (7):
- Rapidly determining when an outbreak is multijurisdictional;
- Identifying potentially involved jurisdictions and agencies;
- Establishing effective and efficient coordination among jurisdictions;
- Clearly identifying a lead outbreak investigation coordinator;
- Monitoring roles and activities in different jurisdictions;
- Managing transitions of the investigation leader or involved parties as the outbreak evolves;
- Coordinating public communications across jurisdictions; and
- Coordinating evaluation of the outbreak investigation and production of a final report, which includes input and data from all involved jurisdictions and agencies.
A large majority of outbreaks are identified and investigated at the local and state levels and do not involve multiple jurisdictions or outside agencies. Local or state health departments detect approximately 75% of foodborne disease outbreaks through complaints of illness directly by consumers or by healthcare providers aware of clusters of illnesses associated with events or establishments (8). This is the primary detection system for outbreaks caused by norovirus, Clostridium perfringens, and other agents for which no pathogen-specific surveillance exists. Relatively few of these outbreaks result in large, complex investigations. For example, an outbreak of norovirus at a restaurant caused by an ill food worker might be an isolated event that can be effectively investigated and controlled by a local health department. However, an outbreak of salmonellosis at a restaurant might herald the interstate distribution of a contaminated fresh produce item that will require the efforts of multiple public health and food regulatory agencies to trace to its source. The exposure to a source of infection and/or the location of ill persons can span jurisdictions (Table 14.1). The variety of food-and water-related vehicles and the complexity of their distribution mean that even apparently simple outbreaks can result in complex investigations. Thus, multijurisdictional outbreaks are frequently detected at the local level or by one agency, and subsequent investigation leads to involvement of other parties.
At the other end of the surveillance spectrum, most multistate outbreaks of Salmonella, Shiga toxin–producing Escherichia coli, and Listeria monocytogenes infections are detected through pathogen-specific surveillance, increasingly coordinated through PulseNet (https://www.cdc.gov/pulsenet), created by the Centers for Disease Control and Prevention (CDC). The multistate distribution of cases implies the widespread distribution of contaminated food products or ingredients. Investigation of these clusters is multijurisdictional from the onset.
aBased on the ill person’s area of residence or location where the contaminated food was eaten (“exposure”).
“Jurisdictions” can be defined by a variety of criteria. These include geography (e.g., different states), regulatory responsibilities (e.g., the Food and Drug Administration [FDA] and US Department of Agriculture [USDA] have nationwide authorities but over different foods), administrative authority (e.g., FDA Office of Regulatory Affairs Regional Field Offices), and outbreak characteristics (e.g., Federal Bureau of Investigation involvement if a crime is suspected).
Many agencies have overlapping jurisdictions, and even within one agency multiple offices or groups can have different responsibilities or authorities, making coordination of communication among multiple parties exceedingly complex. Likewise, different agencies might have “jurisdiction” over different aspects of an outbreak (Table 14.2). For example, at a wedding party, persons from multiple states might have become ill from eating a food in one state from food produced in yet a different state. In such an instance, each individual state (or county) with ill persons is responsible for conducting the epidemiologic interview of affected persons in its jurisdiction. The local health department could have jurisdiction over the caterer that prepared and served the food. The local department of agriculture might regulate the retail store from which the implicated ingredient was purchased. The FDA might have jurisdiction over the facility that produced the product (as well as the farm at which it was grown, unless it came from another country, in which case the FDA would be responsible for coordinating international communications with the foreign agency regulating that facility).
aExamples of general responsibilities for government agencies frequently involved in investigating outbreaks. All states have unique laws, policies, and organizational structures that will affect investigations, and many other agencies and organizations might play important roles in certain situations.
CDC, Centers for Disease Control and Prevention; FBI, Federal Bureau of Investigation; FDA, Food and Drug Administration; USDA-FSIS, US Department of Agriculture, Food Safety and Inspection Service; WHO, World Health Organization.
Rapidly determining which jurisdictions and agencies need to be involved in any outbreak investigation is critical for prompt notification and coordination of investigations (Table 14.3). The agency that detected the outbreak might assume initial responsibility for coordination among jurisdictions. Investigations should be coordinated at the level where relevant investigation steps can be most effectively implemented (6), which requires that the agency have sufficient resources, expertise, and legal authority to collect, organize, analyze, and disseminate data from the investigation.
Coordinating outbreak investigations across jurisdictions requires effective interagency communication. Important tasks include making the initial notifications, establishing roles and responsibilities for each jurisdiction, providing updates on the progress of the investigations, revising priorities for the investigation, and establishing the next steps. Conference calls among collaborating agencies have become a common feature of multistate outbreak investigations. Although these can be an effective way to share information among all parties, the purpose of the call is to facilitate the rapid investigation of the outbreak and focus on issues of import to multiple agencies. In large and prolonged outbreak responses, the number and length of conference calls can impede the investigation’s efficiency, and the outbreak Incident Commander (who usually is a formally designated official in the lead coordinating agency) should appropriately manage the frequency and agenda of multijurisdictional discussions.
Important considerations in multistate outbreak investigations include legal barriers to sharing data, which can limit what can be communicated between agencies. Personally identifiable health information is protected under state disease reporting laws and often cannot be shared across jurisdictions (9,10). However, most demographic data, important clinical details, and exposure information can be extracted from case-patient interviews and shared in aggregate form or as a de-identified dataset. In many outbreak investigations, epidemiologists share exposure information with environmental health specialists and regulatory agencies for the purpose of conducting investigational trace-backs to improve the specificity of exposure assessments. In some instances, regulatory agencies are precluded from sharing the results of their investigations with the epidemiologists because of specific restraints within state laws or federal protections for industry trade secrets. FDA has a provision for sharing information (e.g., proprietary or personally identifiable data) with officially designated persons at the state level, but those persons are prevented from sharing specific information with other colleagues. Barriers to sharing results of investigation activities among jurisdictions are a primary challenge to an effective and efficient response. In addition, commercial entities are also protected, and their identity often cannot be disclosed to protect confidentiality of commercial interests. Some of these barriers may be addressed by involving agency attorneys early in an investigation to identify perceived versus actual legal barriers and help develop waivers or de-identification methods to achieve the desired purposes without violating confidentiality laws.
Source: Adapted from Reference 6.
CDC, Centers for Disease Control and Prevention; FBI, Federal Bureau of Investigation; FDA, Food and Drug Administration; USDA-FSIS, US Department of Agriculture, Food Safety and Inspection Service.
Establishing what agency (and what representative of that agency) is ultimately responsible for coordinating communication and activities of all involved parties is critical. This frequently involves clearly identifying an overall Incident Commander in the Incident Command System (ICS) or an equivalent management structure (see Chapter 16). The Incident Commander is a formally designated position, usually within the agency leading the overall investigation, who is responsible for coordinating overall response activities. Even in multistate outbreaks, the Incident Commander or central outbreak coordinator does not necessarily have to be from a federal agency. For example, in some cases, even in relatively small outbreaks, a local or state agency might invite CDC to lead an investigation. In other situations, for example when one state has a substantial proportion of cases and a robust capacity to investigate, that state might lead the coordination of activities, even across multiple other states and federal agencies. Frequently, the central outbreak coordinator changes as an outbreak response evolves, but any such transitions must be formal and clearly communicated to all participating agencies.
In establishing effective coordination among agencies during a multijurisdictional outbreak response, it is also important to identify a single (or limited number of) point of contact for communication with each agency. Although large numbers of persons might participate in conference calls, meetings, and email communications, it is impossible for a central Incident Commander or outbreak coordinator to ensure that all necessary information reaches the appropriate people in multiple agencies. One point of contact within each agency should be responsible for the “communication tree” within that agency. In all but the simplest outbreaks, the ICS structure should be implemented rapidly. Under such a system, communication can be further coordinated by function (e.g., the Public Information Officer within each agency communicates through the ICS Public Information Officer).
Compiling descriptive epidemiology and preliminary exposure assessments are important to generating hypotheses about the source and transmission routes of the outbreak. Early coordination of epidemiologic studies among affected states is important for efficiently gathering data. Most state health departments have developed their own standard questionnaires for routine interviews of persons identified as having reportable diseases, but clearly defining a consistent case-definition across jurisdictions is a critical step. When an outbreak is suspected, more complex hypothesis-generating questionnaires or analytic study questionnaires are needed. If the outbreak involves multiple states, early coordination to use the same questions in all jurisdictions can ensure that data can be combined for overall analyses and minimize the need for repeated interviews of case-patients as additional details are required. Typically, most states prefer (or are required) to investigate case-patients who reside in their jurisdictions. As noted earlier, legal privacy requirements might prevent sharing of personally identifiable data with other jurisdictions, which precludes cross-jurisdictional interviews and requires sharing of only de-identified data for overall analyses. This requires rapid development of a uniform data-collection instrument and a (preferably electronic) mechanism for capturing data quickly from all jurisdictions into one database that the coordinating agency can analyze centrally. Because conducting case-patient interviews is frequently a rate-limiting step in outbreak investigations, the progress of completing interviews should be routinely monitored by the Incident Commander. If agencies or individuals are unable to complete interviews in a timely manner, alternative arrangements should be made.
In addition to compiling the results of interviews, feedback from the interviews might be useful to (1) identify specific exposures that need to be assessed but were not included on questionnaires, (2) guide environmental health specialists investigating establishments, and (3) identify contamination sources and mechanisms of transmission that need to be traced to identify common distribution pathways that can be incorporated into subsequent analyses. As the outbreak investigation evolves from descriptive epidemiology and hypothesis generation to analytical studies and interventions, the leadership of the investigation might change to reflect its focus. As noted earlier, any such transitions must be formal and clearly defined to all participating parties. In general, investigation of reported cases should be coordinated within the appropriate public health agencies (e.g., state and local health departments, CDC). Similarly, investigation of suspected food items or sources of contamination should be coordinated within the relevant regulatory agencies (e.g., FDA, USDA, local food safety regulators), which will help ensure administrative support for determining the root cause of the outbreak and implementing effective control measures. However, because frequently epidemiologic and regulatory activities considerably overlap, sharing of information between public health and regulatory agencies is critical to the effectiveness of these multijurisdictional investigations.
Public messaging can be challenging during multijurisdictional outbreaks. It is imperative that media messages or other public communication be consistent and coordinated to prevent widespread confusion. Varying local and agency-specific policies and legal restrictions can complicate such messaging. For example, some states do not routinely release the names of restaurants or implicated establishments in an outbreak unless there is a compelling public safety reason to do so; in other states, such information is not legally confidential. The legal environment and focus of specific agencies also can affect messaging. Public health epidemiologists, for example, often consider early public notification about an outbreak, even if scientific analyses are preliminary, in the hope of preventing ongoing exposure. Regulatory agencies, in contrast, frequently have policy and legal constraints that require additional investigation and evidence before the release of specific information. In such situations, rapid coordinated negotiations among agencies and jurisdictions are necessary to determine the timing and detail of public messages and the audiences to which they are released.
All outbreak investigations should be documented with a written report that summarizes the principal findings. These documents can be used for training and as source material for more extensive evaluations of outbreak response procedures. In addition, multijurisdictional investigations should be reported to CDC, indicating that the investigation was multijurisdictional. Individual state reports should be consolidated as part of a multistate outbreak report and linked to summary reports written by regulatory agencies. In cases in which the investigation leads to new information that is generalizable and of potential interest to a wider audience, consideration should be given to publishing results in a peer-reviewed professional journal.
As a wrap up for the investigation, collaborating agencies should meet or hold a conference call to review findings and actions taken. In particular, lessons learned about the outbreak investigation methods should be discussed and disseminated. After the meeting or conference call, the lead agency should summarize findings in an after-action report (AAR). All collaborators should review the AAR to ensure consensus and common understanding of the findings. AARs, like other outbreak investigation reports, should be available for evaluation to identify common problems that arise during multijurisdictional investigations.
The following case studies of past events highlight many issues that need to be addressed during coordination of investigative activities by multiple states, local health agencies within those states, and the federal public health and food regulatory agencies.
Case Study 1
Large, multistate outbreaks associated with contaminated ingredients incorporated into many different products have emerged as a substantial challenge to investigations of foodborne disease outbreaks. Such outbreaks are increasingly detected through rapidly evolving molecular subtyping laboratory techniques. A well-documented example of this type of outbreak was the 2008– 2009 outbreak of Salmonella enteriditis serotype Typhimurium infections associated with peanut butter and peanut butter–containing products (11,12).
On November 10, 2008, CDC PulseNet (a food laboratory network for molecular subtyping of foodborne pathogens) identified a cluster of indistinguishable Salmonella Typhimurium isolates in 12 states. Two weeks later, a cluster of 27 Salmonella Typhimurium cases from 14 states that had a second DNA fingerprint was identified. Epidemiologic assessment of the clusters began on November 25 for the first cluster and on December 2 for the second cluster. Involved states were notified and asked by CDC to identify potentially common sources from routine interviews conducted by state and local health departments. On December 4, the descriptive epidemiology of both clusters was recognized to be similar, and the clusters were effectively joined for investigation purposes.
Hypothesis-generating interviews using a standardized questionnaire were initiated on November 25. During the next 6 weeks, state and local health department personnel interviewed 86 case-patients in 26 states. Local environmental health specialists, state departments of agriculture, the FDA, and the USDA’s Food Safety and Inspection Service all participated in trace-back and trace-forward activities to explore hypotheses for suspected food products.
A turning point in the investigation occurred on December 28, when the Minnesota Department of Health recognized several clusters of cases associated with institutional settings, including nursing homes and schools. By January 4, 2009, Minnesota Department of Agriculture investigators identified a common supplier of peanut butter to the institutions. On January 9, the Minnesota Department of Agriculture microbiology laboratory isolated Salmonella from the peanut butter, and FDA began an investigation of the facility at which it was produced. On January 10, the distribution company recalled the product.
Concurrently, staff in CDC’s Emergency Operations Center conducted a multistate case–control study implicating peanut butter–containing products in the outbreak. FDA investigators at the peanut butter production facility learned that the company also produced peanut paste that was an ingredient used in crackers. On January 18, the Canadian Food Inspection Agency reported isolating Salmonella from intact packages of peanut butter crackers.
Multiple recalls and follow-up investigations were conducted during the ensuing few weeks. Ultimately, almost 4,000 products containing peanut butter produced at a single facility were recalled. A total of 714 cases and nine deaths from 46 states and two countries were associated with the outbreak. An AAR was issued May 18, 2009 (13). At least 190 persons within CDC (the lead agency) participated in the response over the course of 127 days, along with innumerable staff from other federal and state agencies. The outbreak substantially affected the political process that led to passage of the national Food Safety Modernization Act in 2011 (14).
Case Study 2
In 2008, a local health department in Florida identified a cluster of patients with symptoms of acute selenium poisoning (including hair loss and nail changes) (15). As the health department began a local investigation, FDA simultaneously initiated an independent investigation as a result of complaints received through its Safety Information and Adverse Events Reporting System (MedWatch). The local health department investigation rapidly identified a specific vitamin supplement marketed by company A as the source of the outbreak. Case finding identified Internet and phone order sales to customers nationwide, with retail distribution primarily in southeastern states. State health departments, CDC, FDA, and poison control centers were notified. CDC coordinated daily conference calls with all involved agencies.
Ultimately, epidemiologic questionnaires were administered to 227 affected persons in nine states, using a questionnaire developed jointly by the state health departments in Florida and Tennessee. Contact information collected from company A, public calls to health departments, and persons calling regional poison control centers was provided to state health departments, and case-patients were interviewed directly. Because identifying information from persons calling MedWatch could not be released to states, the FDA had to contact patients and ask them to call their health departments directly. The FDA conducted trace-back investigation of the implicated product. The product was distributed by company A in Georgia, which received finished product from a manufacturer in Arkansas, which in turn received ingredients from suppliers in Louisiana. No ill persons were identified in either of the latter states. A nationwide recall of the product was issued. At the conclusion of the investigation, 201 poisoning cases resulting from exposure to selenium were identified in 10 states.
These outbreak investigations involved epidemiology, laboratory, environmental, and regulatory investigators from multiple local, state, and federal agencies. They highlight the importance of multijurisdictional coordination of complex and often overlapping activities to ensure rapid and effective response.
Multistate outbreaks are becoming increasingly recognized because of improvements to public health surveillance. Whether detected through pathogen-specific surveillance or through consumer complaint systems, the need for ongoing multijurisdictional communication and coordination should be anticipated at the start of every outbreak investigation. Plans for efficient and effective multijurisdictional coordination need to be developed and exercised to successfully respond to these outbreaks.
- CDC. Foodborne Outbreak Online Database (FOOD). https://wwwn.cdc.gov/foodborneoutbreaks/
- Kainer MA, Reagan DR, Nguyen DB, et al. Fungal infections associated with contaminated methylprednisolone in Tennessee. N Engl J Med. 2012;367:2194–203.
- Zipprich J, Winter K, Hacker J, Xia D, Watt J, Harriman K. Measles outbreak—California, December 2014–February 2015. MMWR. 2015; 64:153–4.
- Sterling TM, Thompson D, Stanley RL, et al. A multistate outbreak of tuberculosis among members of a highly mobile social network: implications for tuberculosis elimination. Int J Tuberc Lung Dis. 2000;4:1066–73.
- Benin AL, Benson RF, Arnold KE, et al. An outbreak of travel-associated Legionnaires disease and Pontiac fever: the need for enhanced surveillance of travel-associated legionellosis in the United States. J Infect Dis. 2002;185:237–43.
- Council to Improve Foodborne Outbreak Response. Investigation of clusters and outbreaks. In: CIFOR guidelines for foodborne disease outbreak response. 2nd ed. Atlanta: Council of State and Territorial Epidemiologists; 2014:139–65.
- Council to Improve Foodborne Outbreak Response. Special considerations for multijurisdictional outbreaks. In: CIFOR guidelines for foodborne disease outbreak response. 2nd ed. Atlanta: Council of State and Territorial Epidemiologists; 2014:191–204.
- Li J, Shah GH, Hedberg C. Complaint-based surveillance for foodborne illness in the United States: a survey of local health departments. J Food Prot. 2011;74:432–7.
- Association of State and Territorial Health Officials. Improving your access to electronic health records during outbreaks of healthcare-associated infections. A toolkit for health departments. http://www.astho.org/Toolkit/Improving-Access-to-EHRs-During-Outbreaks/
- Association of State and Territorial Health Officials. Address patient privacy, authority and security concerns. http://astho.org/Toolkit/Improving-Access-to-EHRs-During-Outbreaks/Address-Patient-Privacy-Concerns/
- Cavallaro E, Date K, Medus C, et al. Salmonella Typhimurium infections associated with peanut products. N Engl J Med. 2011; 65:601–10.
- CDC. Multistate outbreak of Salmonella infections associated with peanut butter and peanut butter–containing products—United States, 2008–2009. MMWR. 2009;58:85–90.
- CDC. 2008–2009 Salmonella Typhimurium outbreak response, Nov 2008–March 2009: After action report. May 18, 2009. https://www.cdc.gov/salmonella/typhimurium/SalmonellaTyphimuriumAAR.pdf
- US Food & Drug Administration. US Department of Health and Human Services. https://www.fda.gov/food/guidanceregulation/fsma/ucm247546.htm
- MacFarquhar JK, Melstrom P, Hutchison R, et al. Acute selenium toxicity associated with a dietary supplement. Arch Int Med. 2010;170:256–261.