Public Health Emergencies and Responses: What Are They,How Long Do They Last,and How Many Staff Does Your Agency Need?

Methods

Seven outbreaks or events were included in this analysis: the intentional dissemination of B. anthracis through the US Postal Service (2001), SARS (2003), monkeypox (2003), Hurricane Katrina (2005), Salmonella Saintpaul (2008), the 2009 H1N1 pandemic (2009-10), and the Deepwater Horizon oil spill (2010). We reviewed each outbreak or event to determine (1) why CDC considered the outbreak or event to be a public health emergency and what criteria were used to determine that the emergency was over, (2) how and why CDC's response strategy required additional resources, (3) the duration of each public health emergency response, and (4) how many staff (in absolute terms as well as the proportion of CDC's total workforce) were needed for each response.

Although we had resource management data on several international public health emergency responses (eg, the 2004 Southeast Asia tsunami, the 2005 Marburg outbreak in Angola, and the 2010 earthquake in Haiti), we had reliable resource management data for only the 7 domestic events included in this article. We limited the scope of this article to domestic events since very few public health agencies have both a domestic and an international mission as CDC has. We also limited the scope to domestic outbreaks or events that occurred after September 11, 2001, for which an emergency operations center was used and for which documentation of resource use was available. We limited the perspective to the response efforts of CDC, acknowledging that affected state, local, tribal, territorial, and international health agencies were concurrently expending resources on these same events.

Limiting analyses to outbreaks or events that occurred since September 2001 reflects the fact there have been profound changes in both public health policy and budgets associated with public health preparedness and emergency response since that time.^4,5 For example, since 2001 billions of dollars have been the appropriated and available to federal, state, local, territorial, and tribal public health agencies. This has allowed CDC and state and local health departments to improve the public health infrastructure, develop preparedness and response plans, conduct exercises, form new partnerships—for example, with the Department of Homeland Security (DHS) and the Federal Bureau of Investigation (FBI)—and incorporate preparedness and response strategies into daily activities. New policies, such as the National Incident Management System (NIMS) ⁶ and the National Response Framework (NRF), ⁷ and new organization response structures and platforms, such as the Incident Command System (ICS), ⁶ have been incorporated into public health emergency response guidelines and reflect the reality that “public health” is now expected to be part of almost any response alongside first responder agencies. Therefore, since all future public health emergency responses will be tempered and guided by these changes in expectations, partnerships, policies, platforms, and procedures, it makes sense to extrapolate future actions based on events that have occurred since September 11, 2001, as responses after this date are likely to be the most relevant.

In order to describe the operational and resource management requirements of public health emergency responses, we first had to define “public health emergencies” and distinguish them from routine outbreaks that public health agencies face every day. To define “public health emergency,” we started with the assumption that it was different (ie, more severe) from routine outbreaks that health departments investigate every day. We used the conventional definition of an outbreak as being the occurrence of cases of disease in excess of what would normally be expected in a defined community, geographical area, or season. ⁸ We then conducted a literature review of peer-reviewed journal articles using the keyword “public health emergency” in searches of PubMed and CDC's Morbidity and Mortality Weekly Report. We also reviewed internal notes and after-action reports to identify other factors (eg, presidential declarations, requests for technical assistance from state health departments) that served as a trigger indicating that a threshold had been crossed from a routine outbreak to a public health emergency. Our intent was to provide tangible criteria to our state, local, tribal, territorial, and international partners as they discuss various public health scenarios that they might face and determine if an event would be considered a public health emergency in their area.

Quantitative data (eg, duration of events, use of personnel resources) were abstracted from internal documents, situation reports, leadership meetings, and internally produced response management tools that were developed and used during the public health emergency responses.

Results

Table 1 summarizes the results of our analyses. “Declaring” that a public health emergency exists occurred in 4 of 7 events. The declarations were not based on defined clinical, laboratory, or epidemiologic criteria. Rather, declaring a public health emergency was a financial act that allowed congressionally appropriated funds to be made available for the public health emergency response. From a public health perspective, CDC leadership used professional judgment to determine that a threshold from a routine outbreak or event to a public health emergency had been crossed. The literature^9,10 suggested slightly more defined characteristics that distinguish an outbreak response from a public health emergency response; these included (1) the appearance of a rare or previously unknown disease (eg, SARS); (2) the appearance of a disease in an area where it is not endemic; (3) the occurrence of a seasonal disease during an atypical time of the year; (4) the attribution of an outbreak caused by a known pathogen to a strain with an unusual antimicrobial pattern, suggesting strain drift or intentional engineering of the pathogen; (5) an unusual age distribution of people involved in the outbreak; (6) other unusual epidemiologic features of an outbreak due to a known pathogen (eg, a typical foodborne pathogen found to be transmitted from person to person); and (7) an unusual clinical presentation or progression associated with infection with a known pathogen.

Five of the 7 outbreaks or events included in this analysis (intentional dissemination of B. anthracis, SARS, monkeypox, Salmonella Saintpaul, 2009 H1N1) were purely infectious disease responses, whereas 2 events (Hurricane Katrina and Deepwater Horizon) were either natural or man-made events with the potential for infectious disease implications. Column 2 of Table 1 shows that 4 of the 5 infectious disease outbreaks had criteria that were consistent with the literature cited above, and all 5 were multijurisdictional outbreaks.

In determining why CDC needed to redirect staff from core activities to a response effort, column 3 shows that all responses required staff with unique skillsets that did not exist in the organization that was leading the response effort. The other most common reason for needing more resources was that response operations were conducted 24 hours per day, 7 days a week. In all responses, either multiple agencies were involved or thresholds were crossed that were in keeping with preexisting agreements that required activation of an emergency operations center.

Column 4 of the table shows that in 5 of the 7 outbreaks or events, the primary factor that led to ending the emergency response was the lead organizational unit's ability to manage the remaining response activities on its own. For the infectious disease emergencies, the decision to end the emergency response was predicated on a reduction in new cases. With the exception of SARS and anthrax, ending the emergency response was based on the return to endemic incidence rates or expected number of cases of disease. With SARS, the decision was made to sustain the emergency response until 2 complete incubation periods passed without a new infection. This was universally accepted since SARS was a new disease and there was no reason to believe there had been any background or endemic level of disease. In the case of anthrax in 2001, because it was an intentional and malevolent dissemination of a disease-producing pathogen that brought with it great uncertainty, emergency operations continued far beyond 2 incubation periods of no new cases. The decision was made to sustain enhanced vigilance and increased staffing levels through the 2002 Winter Olympics in Salt Lake City, Utah, and the completion of a supplemental medical countermeasures program.

With regard to duration, each emergency response had a start date, which was defined as either the date when personnel resources from outside the lead organizational unit began to work on the emergency response or when the emergency operations center was activated. The end date was defined as when routine outbreak operations resumed (including the return of personnel to their routine jobs), when the level of morbidity and activity warranted the issuance of the final situation report, or when the incident command structure was deactivated, whichever came first. The duration of the emergency response is the number of days between these 2 dates. As seen in column 7, CDC emergency responses lasted an average of 118 days.

Columns 8 and 9 present the staffing burden caused by redirecting personnel from their day-to-day responsibilities to the emergency response. Although the absolute number of CDC staff used for each emergency response is presented in column 8, we present the surge staff burden as a percent of CDC's total workforce in column 9 since the absolute number of staff available to any state, local, tribal, territorial, or international public health agency may not compare to the number of staff available to CDC. The percentage of total workforce may be a more equivalent indicator of burden to both CDC and its state, local, tribal, territorial, and international partners.

For each emergency response, a mean of 934 staff, or approximately 9.5% of CDC's full-time equivalent (or noncontract) workforce, were redirected from their core day-to-day responsibilities to support the public health emergency response. On some days, the percentage was greater than 9.5% (usually at the peak incidence of illness), while near the end of the emergency response, the percentage of staff working on the event or outbreak was far below 9.5%. The staffing database was designed to assist us in locating staff with the skill sets necessary to fill roles in the emergency response. It was not designed to be robust enough to fully explore the impact of staff time diverted from core public health programs, but based on the data we have, it may be reasonable to suggest that approximately 110,212 total staff days were diverted from core public health activities for the average public health emergency response included in this analysis. This is estimated by multiplying the mean number of staff (934) per response by the mean duration of each emergency response (118 days). Of course, staffing burden as a metric varies by response, but an estimate of staff time diverted from core public health programs can be made by either multiplying the number of staff redirected from normal activities (assuming each staff worked a full day each day they were deployed to a team) by the number of days of the response, or by summing the number of hours or days that every staff member worked on the response.

Discussion

On one hand, not having a preestablished definition or trigger to indicate that an outbreak has crossed a line to become a public health emergency is troubling because it accepts location-to-location variability in what constitutes a public health emergency and relies on the ad hoc professional judgment of federal, state, local and international public health leaders to determine that extraordinary resources need to be redirected from core activities to respond to a particular event. However, the benefit of not having a preestablished definition or trigger to signal the start of a public health emergency is that it allows senior-level public health leaders to use their (and their staff's) best judgment to weigh many variables that can be assessed only as an event unfolds. In other words, much as we view the concept of “public health preparedness” as a continuous outcome variable, so should we view the concept of public health emergency as a continuous outcome variable. Preparedness is not a dichotomous concept, and an agency or a geographic area is neither prepared nor unprepared to respond to a public health emergency. Rather, they are more or less prepared to respond to a public health emergency than at another time, and the level of preparedness today may be different from the level of preparedness tomorrow. Equally, a particular outbreak may be an emergency in one location today, but if that location receives more resources, training, experience, and infrastructure, the same event may not be considered a public health emergency in the future. Conversely, if resources, training, and experienced staff are reduced, an outbreak today may be a public health emergency in the future.

The scope of this article was limited to exploring several, but not all, of the resource management issues associated with preparing for and responding to public health emergencies. For example, although we did not plot the day-to-day distribution of the number and type/discipline of staff assigned to each emergency response, we presented such a distribution for the SARS response in an earlier manuscript. ¹¹ We also could not fully explore and describe the fact that, although there are some differences in skill sets needed to respond to infectious disease emergencies compared with natural disaster emergencies, the primary public health disciplines (eg, physicians, epidemiologists, microbiologists, health educators, program managers) and functions (eg, surveillance, epidemiology, laboratory testing, health education and communication, developing patient management guidelines) were common to all public health emergency responses. Equally, this article does not describe CDC's incident command response structure, but this subject has been described elsewhere. ¹²

Figure 1 shows the general model under which CDC operates that describes the association between the need for staff and incident disease or illness for all public health emergencies. The solid-lined bell curve represents incident disease or affected population, and the dotted-line bell curve represents staffing levels that we would typically assign to this type of response. As the number of cases or public health need increases, so will the number of staff that the organizational unit normally responsible for this disease will allocate to the response until it runs out of people. The horizontal dotted line represents a threshold when the available resources in an organization are inadequate to control the outbreak and as the scope of the response expands. At this point, the determination is made that more resources are needed, the public health emergency response begins, personnel are redirected from their routine work, and the impact of their reassignment on daily public health activities may be felt on existing programs. As the number of redirected staff increases and the emergency response extends, the impact may become more pronounced. The shaded area of the curve represents the surge staff recruited from outside the normally responsible organizational unit. The space between the 2 curves will vary depending on the duration from when the outbreak was recognized to when the response begins.

OPEN IN VIEWER

Figure 1.

Model for Surge Staffing

Figure 2 represents our experience that there may be multiple surges in any single emergency response. Although the figure is presented showing the second surge as smaller than the first, this is for demonstration purposes only. The second surge can be larger, as large as, or smaller than the initial surge. Multiple surges can be caused by fluctuating morbidity and mortality as infections spread geographically or through different population groups. Another reason for a second surge might be the implementation of a new or different intervention or countermeasure, such as the distribution of a new vaccine. These surges can also indicate different phases of a response and indicate that different staff skills are needed during the course of the response. For example, many microbiologists and epidemiologists may be needed at the beginning of a response to describe the scope of the problem, followed by a surge of health educators and public health advisors to manage intervention activities and messaging.

OPEN IN VIEWER

Figure 2.

Model for Dual Phase Surge Staffing

Two distinct surges occurred during the 2001 dissemination of B. anthracis when supplemental postexposure prophylaxis became available to potentially exposed populations after the initial investigation of the morbidity and mortality that occurred from the contaminated envelopes. A second surge also occurred during the 2009-10 H1N1 pandemic when the initial temporal appearance slowed during the summer only to increase again in the autumn.

Since determining when an outbreak becomes a public health emergency is primarily a function of professional judgment and is therefore subjective, the operational and resource management requirements for responding to public health emergencies is also subjective and is geographically dependent, predicated on the resources that are available to conduct the response from within the responsible organizational unit (Figure 3).

OPEN IN VIEWER

Figure 3.

From Outbreak to Public Health Emergency

Figure 3 suggests that the primary difference between the decision to conduct an outbreak response compared with a public health emergency response is the severity (Y-axis) of the outbreak or event. The degree or extent of the severity (ie, the threshold between an outbreak response and a public health emergency response) depends partly on the morbidity and mortality associated with the outbreak or event, but it also depends on the resources and capabilities that each jurisdiction needs compared with what it has available to respond (X-axis). The threshold separating an outbreak response from a public health emergency response is represented as the overlap in the Venn diagram. It suggests that because of any number of variables (eg, existing capabilities and resources, preexisting memoranda/agreements with other agencies), the decision to initiate a public health emergency response in one geographic area may not be made at the same point in another geographic location, even with the same illness and identical morbidity and mortality. In some locations, a public health emergency response may begin at point 1 on the diagram (eg, an attack rate of X, or a reported number of Y new cases in a defined time period), yet in other locations, the severity would have to get to point 4 before they felt that they were at a point where they needed to implement a public health emergency response. Equally, the need for additional resources (between points 1 and 3) would be dependent on the difference between the resources that were available in the particular organizational unit that normally responds to this type of outbreak or event and the resources that were needed to conduct the emergency response. The need to use staff most efficiently becomes more complex if a percentage of the workforce could become ill during the response or would need to care for an ill family member. This was a major concern during the H1N1 emergency response.

We have seen that, although public health emergencies are certainly less common than standard outbreaks, they are not rare. As is the case with other public health agencies, CDC is constantly refining its planning for and execution of public health emergency responses.

As our knowledge base increases and as available resources decrease, there are many reasons to consider the impact of conducting public health emergency responses from a resource management perspective. We realize that there are differences in scope, mission, legal authority, resources, and expertise between CDC and its state, local, territorial, tribal, and international partners, but we believe that each partner can adapt what we have learned to its own situation and consider the following factors in developing a resource management component of its preparedness strategy.

First, preparing for public health emergencies that occur infrequently but inevitably allow public health agencies to plan for the fact that, assuming there are no legal or administrative prohibitions, their staff will be redirected from their day-to-day responsibilities to respond to public health emergencies when they occur, and that they will be redirected for weeks at a time. Preparedness planning forces organizations to describe the conditions under which their staff will be redirected from their normal jobs (ie, when an outbreak has become a public health emergency).

Second, anticipating the inevitability of responding to public health emergencies allows agencies to conduct some activities before the emergency begins. This may include identifying and training staff who will play key roles during the emergency response; identifying the skill sets needed during different phases and stages of the response; training staff in the basics of the incident management system; purchasing, storing, and preparing to ship medical countermeasures; creating guidelines, response plans, and other communications messages; and establishing or enhancing liaisons with affiliated organizations (eg, hospitals, poison control centers) and with organizations with which there is less frequent interaction (eg, law enforcement, civil defense). State and local public health agencies may need to anticipate that they would have to develop or implement processes to redirect federal funding to conduct the response or take the steps necessary to request and receive supplemental funding.

Third, while the duration of a public health emergency response cannot be predicted with certainty, improving the ability to predict the duration of these responses will help public health authorities to decide which of their ongoing or existing programs and projects should be curtailed, delayed, or canceled while resources are redirected to work on the public health emergency responses.

Fourth, knowing that surge staffing will be available for only a few weeks to a few months, the leaders and managers of the response effort may be better prepared to ask strategic questions earlier in the response—for example, We know that we will only have increased staffing for X days or weeks. What do we want to accomplish by day 7, 14, or 30 before they return to their regular jobs?

Fifth, realizing that resources will need to be redirected from core public health programs should motivate response leaders to allow the redirected or surge staff to return to their normal responsibilities as soon as possible. The need to resume normal activities should encourage leaders to devise and articulate an exit strategy (ie, establish criteria to indicate when the public health emergency has ended) as the strategy is developed and not while the response is in progress. This thoughtful inductive and deductive process should lead to the establishment of clear goals, objectives, and metrics for the response effort, thus making the redirection of staff from critical day-to-day activities more understandable and justifiable to the entities that are giving up staff and other resources for the response.

Finally, although we are continuously learning how to use resources more efficiently during each public health emergency response, we should use the times between these events to try to enhance the efficiency with which we anticipate, request, and use resources. Specifically, although current modeling efforts regarding surges do not focus on staffing needs,^13-15 it may be reasonable for modelers to explore and infer the impact of increased staffing on future outbreaks. The Venn diagram in Figure 3 suggests a linear relationship between the severity of morbidity and mortality and the need for additional staff, but the relationship may not be linear. For example, it is worth asking if surge staffing doubles or triples the number of staff working on the emergency response, whether the duration or severity of the outbreak or event can be halved or reduced by two-thirds simply by adding more staff, or if there is a point of diminishing returns when additional staff has no impact on incident disease.

Conclusions

The decision that an outbreak or event has become a public health emergency will likely continue to be a matter of the professional judgment made by public health leaders. Given that a public health emergency in one location would not be considered a public health emergency in another, we encourage state, local, tribal, territorial, and international public health agencies to picture themselves in Figure 3 to determine at what point for any given outbreak or event they would need additional resources, the potential sources for those resources, the criteria they would use to end emergency operations, and the impact these decisions have on core public health programs. In this article we have presented some criteria that public health agencies may use or adapt to better understand the time and resource parameters in planning for and responding to public health emergencies.

The applicable lessons here may not be fully transferable to state, local, tribal, territorial, and international public health agencies because of CDC's unique mission and global geographic scope, and although it is unlikely that any single public health agency will have to respond to all types of public health emergencies (eg, bioterrorism, natural disasters, foodborne outbreaks, part of a multijurisdictional respiratory outbreak of unknown etiology), they should be prepared for the fact that they may be called on to respond to or contribute staff to respond to any of them. Although there will continue to be state-to-state variability in risk, skills, and resources, the basic concept of defining a public health emergency and establishing its starting and ending points applies to all public health leaders charged with conducting core daily public health functions.