Investigations of Selected Historically Important Syndromic Outbreaks: Impact and Lessons Learned for Public Health Preparedness and Response

Abstract

Public health readiness has increased at all jurisdictional levels because of increased sensitivity to threats. Since 2001, with billions of dollars invested to bolster the public health system’s capacity, the public expects that public health will identify the etiology of and respond to events more rapidly. However, when etiologies are unknown at the onset of the investigation but interventions must be implemented, public health practitioners must benefit from past investigations’ lessons to strengthen preparedness for emerging threats.

We have identified such potentially actionable lessons learned from historically important public health events that occurred primarily as syndromes for which the etiological agent initially was unknown.

Ongoing analysis of investigations can advance our capability to recognize and investigate syndromes and other problems and implement the most appropriate interventions.

OVER THE PAST DECADE, PUBlic health vigilance and readiness have increased at the international, national, state, and local levels because of increased sensitivity to both naturally occurring and intentional emerging threats, the former including problems such as the 2009 influenza A (H1N1) virus pandemic and the latter including severe illnesses resulting from the deliberate introduction of biological and chemical agents.^1–5 Only 18 months following the intentional anthrax attack of 2001, the world was confronted by the natural emergence of severe acute respiratory syndrome (SARS), a paradigm for newly recognized syndromes for which the specific etiological agents initially are unknown.^6–9 However, unlike the case for SARS, for which well-coordinated global efforts resulted in rapid identification of the causative agent, there are and will continue to be infectious and other disease problems with causative agents that remain undiagnosed or with pathophysiology that remains unknown despite intensive investigative efforts.^10–12

Since the anthrax attacks of 2001, the expectations placed on the public health community by the public, the media, and elected officials have steadily increased. Billions of dollars have been invested to bolster the capacity, capability, and proficiency of the public health workforce and other public health system components.^13,14 The public’s expectation is that given this large investment in public health preparedness and response in recent years, “public health” will respond to and identify the infectious or noninfectious etiology of events more rapidly than in the past. However, precisely when the etiologies are unknown at the onset of public health investigation and response, but control and intervention measures must be implemented to protect the public’s health, public health and medical practitioners must mine lessons from past public health investigations to strengthen preparedness for emerging threats.

We have identified such potentially actionable lessons that can help inform approaches to identify and control newly emergent infectious and noninfectious threats more rapidly. We focus on those events that occurred primarily as syndromes for which the etiological agent was unknown at the time of recognition and initiation of investigation. For some of these, the etiology remains unknown. Some of those events came to the attention of the Centers for Disease Control and Prevention (CDC) and the broader public health arena as emerging threats (e.g., SARS) or were brought to light by astute clinicians who recognized a potential problem and reported it to the public health authorities (e.g., immune-mediated polyradiculoneuropathy). For others (e.g., Guillain-Barré syndrome and Reye’s syndrome), the response emphasizes the importance of surveillance for recognizing unanticipated problems and bringing them in focus for investigations (L. Schonberger, MD, personal communication, December 8, 2009). We also indicate how, for some problems, public health professionals need to initiate a rapid, comprehensive investigation that may include selecting and implementing control measures even if the etiology remains unknown.

INVESTIGATIONS OF SYNDROMIC PUBLIC HEALTH EVENTS

We included investigations of syndromic events if (1) the initial public health event had an apparent acute or novel onset (i.e., was not readily linked to a previously recognized problem) and was detected in the period from 1963, when Reye’s syndrome was first reported,¹⁵ to 2008; or (2) an etiology was identified for what appeared to have been a new problem but subsequently was confirmed to have been an earlier incident or prevalent problem. These public health events were identified by using the keyword “syndrome” in searches of PubMed and CDC’s Morbidity and Mortality Weekly Report and through our and consultant colleagues’ familiarity with major public investigations that CDC and other public health partner agencies had conducted.

Our review identified investigations of 14 syndromic events during the period from 1963 (Reye’s syndrome) to 2008 (immune-mediated polyradiculoneuropathy [previously progressive inflammatory neuropathy]^15–83; Table 1). The other investigations included the syndromic events of Kawasaki disease; Legionnaires’ disease; Guillain-Barré syndrome associated with receipt of swine influenza vaccine; Lyme disease; toxic shock syndrome (TSS); HIV/AIDS; toxic oil syndrome; Brainerd diarrhea syndrome; chronic fatigue syndrome; eosinophilia-myalgia syndrome (EMS); hantavirus pulmonary syndrome; and severe acute respiratory syndrome (SARS).

TABLE 1—

Selected Public Health Syndromes, by Year of Recognition, Epidemiological Characteristics, Duration From Recognition to Etiological Identification, and Implications: 1963–2008

Syndrome and Year First Recognized	Etiology, Risk Factors, and Mode of Spread or Exposure	Time From Recognition of Problem to Identification of Etiology	Public Health and Medical Care Implications and Effect
Reye’s syndrome, 196315–21	First described in 1963 in Australia by Reye et al.15 and in the United States.16	Not applicable: specific etiology remains unknown.	Precautions were issued in 1980 and 1982 regarding use of salicylates to treat chickenpox and influenza in children19; in 1982, the US Surgeon General made similar recommendation.20 In 1986, the FDA required over-the-counter salicylate products to contain warning labels.19
	Subsequent investigations implicated risk factors, including salicylate use in setting of certain antecedent viral infections (i.e., influenza and chickenpox), particularly in children.17,18,20		Incidence of Reye’s syndrome began declining dramatically in the United States in the early 1980s following the studies reporting an association between salicylate use in children with chickenpox and influenza and coinciding with a decline in aspirin use in children following publicity about studies.19
	Specific etiology remains unknown.		Reported cases peaked at 555 in 1980; during 1987 through 1993, no more than 36 cases were reported annually, and during 1994 through 1997, no more than 2 cases were reported annually.21
Kawasaki disease, 196722–27	An acute febrile illness and inflammatory vasculitis primarily affecting young children and characterized by winter to spring seasonality, first described in Japan in 1967.22,23	Not applicable: specific etiology remains unknown.	Etiology suspected to be infectious with involvement of superantigenic mechanism.
	Specific etiology remains undetermined.24,25		Effective treatment documented in early 1980s27 approximately 15 y after syndrome first reported. Age-specific risk, seasonality, and clinical presentation allow for rapid clinical management.
			Treatment with intravenous immunoglobulin and aspirin decreases risk of coronary artery complications (dilatations and aneurysms).26
Guillain-Barré syndrome associated with swine influenza vaccination, 197628–30	Investigation of the epidemic of Guillain-Barré syndrome cases in late 1976 determined that increased risk of Guillain-Barré syndrome was associated with antecedent receipt of killed swine influenza vaccine (A/New Jersey/1976/H1N1) administered in the National Influenza Immunization Program.28	The specific etiology and pathogenesis of Guillain-Barré syndrome following receipt of the 1976 swine influenza vaccine remain unknown.	Guillain-Barré syndrome was characterized as an abnormal immunological response to specific, usually common antigens associated with common infectious agents; although the swine influenza vaccines did not contain live pathogenic agents, they triggered the Guillain-Barré syndrome response (L. Schonberger, MD, Centers for Disease Control and Prevention, personal communication, July 6, 2009). Although multiple studies examined associations between receipt of other influenza vaccines and subsequent Guillain-Barré syndrome, no substantial increase in Guillain-Barré syndrome is associated with influenza vaccines other than the 1976 swine influenza vaccine.29 In addition, although Guillain-Barré syndrome was associated with receipt of the swine influenza vaccines, the specific antigens within the vaccines that triggered Guillain-Barré syndrome remain unknown (L. Schonberger, personal communication, July 6, 2009). One investigation hypothesized that the 1976 swine influenza vaccine contained contaminating moieties (e.g., Campylobacter jejuni antigens mimicking human gangliosides of other vaccine components) that elicited antiganglioside antibodies that are associated with development of Guillain-Barré syndrome.30
	Specific etiology and pathogenesis remain undetermined.		The National Influenza Immunization Program was instituted October 1, 1976; by early December, when > 35 million doses of vaccine had been given, clusters of Guillain-Barré syndrome in vaccine recipients were reported to the CDC by 2 states. Because of the increase in reported Guillain-Barré syndrome cases following vaccination, the National Influenza Immunization Program was suspended on December 16, 1976. The period of increased risk primarily was within the 5-wk period following vaccination, although it extended through approximately 9 or 10 wk.28
Pontiac fever (1968) and Legionnaires’ disease, 197631–38	Legionella pneumophila, causing isolated cases and outbreaks of acute, febrile respiratory disease among persons exposed to infectious aerosols (e.g., showers, hot tubs, air-conditioning cooling towers, contaminated soil); organism characterized following (and genus named in relation to) major outbreak among American Legion convention attendees and others in July 1976.31–33	The syndrome of Pontiac fever, characterized during a 1968 outbreak investigation,34 was attributed to L. pneumophila only after the investigation of the 1976 Legionnaires’ disease outbreak and report of isolation of L pneumophila in January 1977.35	Excellent example of an emerging infectious pathogen identified as result of investigation of large outbreak, but then determined to have existed as previous cause of disease, including the 1968 Pontiac fever outbreak centered in a health department, other outbreaks investigated in the 1950s and 1960s, and a pneumonia patient in 1947.36 Interventions include disinfection of potential aerosol-producing devices and other sources of Legionella bacteria and other environmental risk reduction measures.
Lyme disease, 197739–43	Borrelia burgdorferi, gram-negative spirochete, transmitted to humans by bite of infected deer ticks (Ixodes scapularis). US risk localized to specific geographic areas.	Lyme disease first recognized in 1975 through cluster of pediatric cases in Connecticut.39 The causative agent (B burgdorferi) was identified in 1982.41,42	Twelve pediatric patients initially received diagnoses of juvenile rheumatoid arthritis. In summer of 1977, ticks were evaluated for secondarily acquired organisms, and after a tick-borne spirochete was identified and implicated, interventions used for similar tick-borne diseases were implemented.
			In 1998, FDA licensed the first vaccine for use for prevention of Lyme disease. However, the manufacturer pulled the vaccine from the market in early 2002.
TSS, 197844–50	For TSS associated with menstruation and tampon use (TSS toxin-1), an exotoxin produced by Staphylococcus aureus isolates in affected patients.44,45	The time period varies as function of which recognition report is used: 1978 report by Todd et al.46 of small cluster of cases primarily in children or first report in May 1980 of major multistate outbreak among younger women indicating emergence of that distinct outbreak in October 197947; using the latter as reference, then 11 mo elapsed from recognition to identification of putative agent.	A June 1980 report on follow-up investigations described an association between TSS risk and use of tampons,48 and subsequent studies detected increased risk of TSS in women using a certain tampon brand. In September 1980, the manufacturer voluntarily withdrew tampons of that brand from the market. Withdrawal of that brand and subsequent decrease in use of high-absorbency tampons correlated with marked decrease in occurrence of menstrual TSS.Investigation of this problem led not only to recognition of the emergence of menstrual TSS but also to enhanced legal protection at the federal level of the public health research process and health information privacy.49 In the face of lawsuits filed by women with menstrual TSS, a tampon manufacturer filed suit to compel CDC to release personal identifying information about the women who participated in related studies. In the federal court’s ruling denying the manufacturer’s request, the court commented on privacy expectations and noted that such disclosure could be detrimental to the public health mission.50
HIV/AIDS, 198151–57	HIV-1 and HIV-2 transmitted from person to person through sexual contact, sharing of HIV-contaminated needles and syringes, receipt of transfused contaminated blood or components, and receipt of transplanted infected tissues or organs.	Although the first epidemic-associated cases (Pneumocystis carinii pneumonia in 5 homosexual males) were reported in June 1981,51 confirmation of the epidemic was not reported until early 1982.52 Identification of putative causative agent reported in 1983–1984.54,55	The initial 1981 descriptive report of the cases that ultimately were recognized as part of the HIV/AIDS epidemic suggested a role for lifestyle (sexual contact) risk factors but proposed no interventions. Risk factors were more completely identified by late 1982, and soon after, the US Public Health Service recommended interventions involving sexual contact and donation of plasma or blood.57
			The initial recommendations to curb this epidemic, published 21 mo after the first reported cases, were based primarily on findings of epidemiological studies and made available even before publication about the determination of the etiological agent for AIDS.54,55,57
Toxic oil syndrome, 198158–61	Epidemic of severe, progressive multisystem disease had onset in Spain in May 1981.58,59	Not applicable; specific etiology remains unknown.	Following implication of risk in association with consumption of oil sold in certain plastic containers, Spain’s Ministry of Health and Consumer Affairs initiated an official recall program in late June 1981.
	Risk of illness was associated with consumption of rapeseed oil that was denatured with 2% aniline for industrial use, subsequently refined, then sold illicitly as pure olive oil.60		This epidemic included an estimated 20 000 cases of illness and 300 associated deaths. Nearly 3 decades later, although laboratory findings suggest an autoimmune mechanism, the etiology remains unknown despite prolonged and extensive laboratory and toxicoepidemiological investigations.60,61
	Specific etiology remains undetermined.
Brainerd diarrhea syndrome, 198362–64	Syndrome of acute onset of watery diarrhea lasting ≥ 4 wk.	Not applicable; specific etiology remains unknown.	Risk reduction by avoiding consumption of raw milk and water improperly chlorinated or boiled.63
	Risk factors include consumption of unpasteurized (raw) milk or unboiled or inadequately chlorinated water.62		Despite knowledge of this syndrome for more than a quarter-century, its etiology remains unknown, and a key unanswered question is whether Brainerd diarrhea occurs only in outbreaks or also causes sporadic cases of chronic diarrhea.64
	Specific etiology remains undetermined.
CFS, 198865,66	The term chronic fatigue syndrome was proposed in 1988 to replace what, since 1985, had been referred to as chronic Epstein-Barr virus syndrome, a symptom complex of unknown cause characterized primarily by chronic fatigue.65	Not applicable; specific etiology remains unknown.	With etiology and risk factors still unknown, there is no specific diagnostic test for CFS, and diagnosis remains one of exclusion; there is no intervention for CFS.
	Specific etiology remains undetermined.66		CDC has suggested that a single cause for CFS may yet be identified, but CFS could represent a common endpoint of disease resulting from multiple precipitating causes, including, for example, viral infection, trauma, stress, toxins, and endocrine disorders.66
Eosinophilia-myalgia syndrome, 198967–72	Epidemic of severe multisystem illness detected and investigated in 1989.	Not applicable; specific etiology remains unknown.	State public health officials notified of first outbreak cases on October 30, 1989, and initiated investigation that rapidly implicated oral use of l-tryptophan as risk factor for syndrome. On November 11, 1989, FDA advised consumers to discontinue use of l-tryptophan-containing tablets and capsules; on November 17, FDA recalled all dietary supplements that provided a daily dose of l-tryptophan ≥ 100 mg. FDA subsequently expanded that recall.69
	Implication of use of specific brand of dietary supplement, l-tryptophan, as risk factor for illness.67,69		Clinical features of this syndrome are similar to those seen in patients during the toxic oil syndrome epidemic of 1981. This similarity helped the investigators in focusing their epidemiological and clinical research.71 Despite considerable agreement that risk of illness was associated with ingestion of l-tryptophan from a single lot produced by 1 manufacturer, the precise etiological agent(s) and mechanism(s) remain unknown.68 The magnitude of this outbreak (> 1500 cases through May 1991) and prolonged period of case detection highlighted limitations in public health surveillance systems for detecting and monitoring such problems.70,72
	Specific etiology remains undetermined68 (R. Philen, personal communication, March 16, 2009).
Hantavirus pulmonary syndrome, 199373–76	Sin Nombre virus, a type of hantavirus; exposure primarily through inhalation of aerosolized virus excreted into environment in the urine and feces of infected mice.74	Approximately 1 mo: first indication of 1993 outbreak in US Four Corners region was on May 1473; by time of report on June 11, 1993, preliminary serological evidence suggested hantavirus infection as the outbreak’s cause, and by June 18, polymerase chain reaction had documented presence of hantavirus genome in autopsy specimens of 2 fatal cases.73,74	Avoidance of activities that result in contact with rodents or aerosolization of rodent excreta was recommended on June 11, 1993, ≤ 1 mo from initial detection of the outbreak.73
			Sin Nombre virus’ role as the causative agent was not immediately recognized in the US outbreak in 1993, even though a related virus previously was shown to cause a different disease in humans (e.g., before World War II, disease was described in persons in Manchuria). That hantavirus was isolated from a field rodent, Apodemus agrarius, in 1977 near the Hantaan River in Korea.76 However, the 1993 US outbreak presented different clinical features (e.g., abrupt onset of fever, myalgias, headache, and cough, followed by rapid development of respiratory failure).
SARS, 200377–81	Worldwide outbreak of SARS was caused by SARS-CoV with isolation of previously unrecognized, novel SARS-CoV reported in late March 200377,78; spread by person-to-person transmission through respiratory secretions and droplet exposure.	Some uncertainty because of limitations in reporting from initially affected countries but generally believed to be less than 6 months—in early February 2003, China notified World Health Organization that 305 cases of acute respiratory syndrome of unknown etiology had occurred in southern China during November 16, 2002, through February 9, 2003,79 but with the first case likely occurring in October 200280; isolation of SARS-CoV was reported in April 2003.	At least as early as March 21, 2003, pending clearer definition of mode of transmission, CDC recommended the use of standard infection control measures, airborne and contact precautions, and eye protection for all patient contact.79
			Following activation of CDC’s Emergency Operations Center on March 14, 2003, CDC’s emergency response phase lasted 133 d and used more than 850 people who provided quarantine and other technical assistance to countries reporting large numbers of cases, met passengers and crew from these locations on arrival in the United States, and ensured case reporting and investigation.81
Immune-mediated polyradiculoneuropathy (previously progressive inflammatory neuropathy), 200882,83	A subacute neurological syndrome among pork plant workers with onset during November 2006 through November 2007, with report received by state health department (Minnesota) in late October 2007; characterized by sensory-predominant polyradiculoneuropathy.82 Thought to be an autoimmune phenomenon associated with presumed mucosal exposure to aerosolized swine brain83 (J. Sejvar, personal communication, May 6, 2009).	Not applicable; specific etiology remains unknown.	This syndrome occurred in swine slaughterhouse workers exposed to aerosolized pig brain through inhalation or mucous membrane contact. Measures instituted within 1 mo of public health notification and investigation included mandated use of additional personal protective equipment to reduce exposure and plant operators’ voluntary suspension of harvesting of brains.
	Specific etiology remains undetermined.		Exposed, affected workers experienced nerve conduction abnormalities involving very distal and proximal nerves; most patients improved following cessation of exposure, both with and without immunotherapy.83

Note. CDC = Centers for Disease Control and Prevention; CFS = chronic fatigue syndrome; FDA = US Food and Drug Administration; SARS = severe acute respiratory syndrome; SARS-CoV =  SARS-associated coronavirus; TSS = toxic shock syndrome.

Through the collection and analysis of epidemiological and laboratory evidence, a specific etiology was identified in 8 of 14 events. Six of those 8 events were caused by (or associated with) viral and bacterial pathogens (Legionella pneumophila; Borrelia burgdorferi; HIV; Sin Nombre virus; exotoxin-elaborating Staphylococcus aureus; and a novel SARS-associated coronavirus, SARS-CoV). Spread or acquisition occurred by a variety of routes, including airborne, food-borne, person-to-person, and vector-borne transmission. In the instances of identification of a new etiological agent, the duration from detection to identification of the putative cause ranged from 4 to 5 weeks (hantavirus pulmonary syndrome) to 7 years (Lyme disease). Two syndromic outbreaks resulted from novel exposures to noninfectious agents: adulterated rapeseed oil, the vehicle associated with the toxic oil syndrome epidemic that began in Spain in 1981; and l-tryptophan, the contaminated dietary supplement responsible for the EMS outbreak that was recognized in 1989.

Although the cause and epidemiology were determined for more than half of the syndromic events reviewed in this article, for 6 of them—Reye’s syndrome, Kawasaki disease, Guillain-Barré syndrome associated with receipt of swine influenza vaccine, chronic fatigue syndrome, Brainerd diarrhea syndrome, and immune-mediated polyradiculoneuropathy—the specific etiologies have remained undetermined for years, even decades, after their initial recognition despite extensive clinical, epidemiological, environmental, and laboratory studies.

Investigations were initiated in response to each of these syndromes to define the outbreak better or to alter or eliminate risk factors that sustained transmission, or both (last column of Table 1). However, the type, scope, and focus of the investigations varied depending on the amount of information that was available. For example, the 1993 hantavirus pulmonary syndrome outbreak led to rapid epidemiological determination of risk factors, laboratory confirmation of the etiological agent, and establishment of a combined epidemiological and laboratory foundation for targeted interventions, including environmental control measures and risk reduction behaviors.⁷⁵

The response to the HIV/AIDS epidemic following its initial recognition in 1981 illustrates a situation in which epidemiological investigation determined risk factors and a basis for recommended interventions that were implemented far in advance of the identification of a definitive causative agent.^53,56 Another example of control measures that could be implemented following the initiation of epidemiological investigations, even when the etiology remained unidentified, is the Brainerd diarrhea outbreak of 1983; risk factors were determined, and some interventions could be formulated.⁶²

ANALYSIS

Responding to outbreaks is an essential public health function that characterizes the epidemiology, identifies the causative agent or pathogen, and establishes a basis for implementing control measures that will reduce, alter, or eliminate risk factors for ongoing illness.⁸⁴ However, as illustrated by many of the public health events in this article, public health investigators may not possess all the information necessary to implement a totally defensible intervention in the same way as if conducting a prospective study in the field—at the same time, the public health imperative requires initiation of a public health response based on the limited information available even in the face of undetermined etiology. In such circumstances, professional judgment and incomplete information may need to be combined to reduce morbidity and mortality. Several such actionable lessons learned can be drawn from the 14 investigations analyzed in this study.

Identify and Adopt Actionable Lessons

Sustain and enhance the core public health system.

Sustained, predictable investments in the core public health system, including the ongoing investment in strengthening of public health surveillance for selected conditions (e.g., novel or exceptionally virulent strains of influenza virus and other emerging infectious diseases), are critical so that clusters of illness can be identified quickly and the likelihood of sustained transmission of disease or multiple waves of illness will be minimized. These investments support not only strengthening existing surveillance systems but also designing and implementing new systems more effectively with communications technologies such as the Internet to enhance detection and notification of problems (e.g., CDC’s PulseNet, CDC’s Epi-X, and Pro-MED^85–87) and applying standards specified in agreements such as the revised International Health Regulations that require multinational notification and information sharing following the recognition of certain types of public health threats.⁸⁸ Indeed, since 2001, the US Congress has placed great emphasis on strengthening the capacity, capability, and proficiency of public health professionals and systems at all jurisdictional levels, including upgrading epidemiological, laboratory, and communications core public health functions.^14,89

The need for fostering collaborative efforts beyond those of traditional public health is underscored by many of the events in this study, including the movement toward interdisciplinary engagement. As an example, given that a substantial proportion of emerging and reemerging infections are zoonoses, interdisciplinary engagement in all aspects of health care for humans, animals, and the environment is a key consideration driving The One Health Initiative.⁹⁰ An additional example, that of a government public health agency’s commitment to strengthening partnership with veterinary authorities, is CDC’s establishment of the National Center for Emerging and Zoonotic Infectious Diseases. Advances in formal, interdisciplinary cooperation have emerged or accelerated over the past decade, particularly in the settings of national and global public health emergencies (e.g., the anthrax attacks and avian influenza).

Enhance development, understanding, and use of traditional and nontraditional response options.

The historic model for public health response is to establish and sustain methods and programs (e.g., surveillance, epidemiology, laboratory screening and diagnostics, vaccinations, health education, and other interventions) that can rapidly identify potential public health anomalies, assess their effect on the population, and determine the causative agent to initiate steps to reduce, alter, or eliminate the risk for continued transmission. Although many response tools and other options have been available to public health practitioners for investigating problems such as those included in this article, enhancing diagnostic and technological tools and expanding policy tools and partnerships warrant additional discussion because maximizing the use of these strategies may be critical in responding to events when information is insufficient for initiating control measures.

Enhance diagnostic technologies.

The investigation of the epidemics of Legionnaires’ disease in 1976 and hantavirus pulmonary syndrome in 1993 clearly illustrates how changes in problem recognition and advances in diagnostic technologies can help to retrospectively diagnose previously unsolved problems. For example, after L pneumophila was identified as the cause of the major acute respiratory disease epidemic in Philadelphia, Pennsylvania, in 1976, this pathogen was confirmed as the cause of several previously unsolved outbreaks, including some as early as 1957 and 1965^37,38 and the 1968 Pontiac fever epidemic.³⁴ Similarly, following the 1993 hantavirus pulmonary syndrome outbreak, immunohistochemical evaluation of retrieved autopsy tissues showed widespread deposition of hantaviral antigens within endothelial cells in 12 patients who had died before 1993 from histologically unexplained noncardiogenic pulmonary edema; the earliest such case had been investigated in 1978.⁹¹ Continuous investment of a portion of limited resources into research and development is crucial, with its relevance clearly illustrated in the field of laboratory science. Enormous advances in microbiological, molecular, and other diagnostic technologies (e.g., polymerase chain reaction and enzyme-linked immunosorbent assay) allow for substantially more rapid and specific identification of a broader spectrum of biological agents than imaginable even 2 decades ago. Consider that following the wider notification of the 2003 SARS outbreak, the causative agent, SARS-CoV, was identified and confirmed within weeks,⁷⁸ a feat not readily possible only 22 years earlier at the onset of the HIV/AIDS epidemic.

Transform public health policy and legal authorities at local, state, federal, and global levels.

The experience with the then-newly emerging SARS-CoV in 2003 is a case study of the extent to which investigative findings can have both an immediate and a long-term effect on public health preparedness policy and practice. This multilateral investigative response reinforced the World Health Organization’s role in coordinating selected international epidemiological and laboratory investigations directed toward identifying an etiological agent⁹² and showed how the response to some problems can advance key policy initiatives, such as catalyzing the World Health Assembly’s completion of the revision of the International Health Regulations.⁸⁸ In the United States, SARS also prompted federal, state, and local public health agencies to review their legal authorities for isolation and quarantine, disease control tools that many jurisdictions had not used on a wide scale for many years, if not decades, but that increasingly could be used to contain the spread of virulent influenza strains and other communicable respiratory disease threats. Beyond the effects relevant to policy development and legal authorities, this event elucidated the paramount role of case definitions during investigations⁹³; underscored the importance of rapidly identifying causative infectious agents and their mode of transmission⁸⁰; and showed the potential role of the Internet as an adjunctive investigative tool.⁹⁴

Implement provisional control measures even as the etiology remains under investigation.

The following question is particularly challenging for public health practitioners: When, during the course of an investigative response, do public health decision makers have sufficiently defensible information to enable selection and implementation of control measures, particularly when that understanding is based predominantly on epidemiological findings in the absence of a specific etiological determination? This issue encompasses an overarching goal of epidemiological field investigations—the need for a scientifically rational basis for acting, including selecting and using specific interventions.⁹⁵ In addition to determining the disease-causing agent’s identity and mode of spread, other crucial factors to be weighed by public health officials include the extent to which an epidemiological investigation addresses the tenets of causation, the need for urgent action because of the magnitude of the public health effect (e.g., its morbidity, mortality, and speed of spread), and the potential for adverse consequences of taking action. Taking scientifically defensible action grounded in experience also may help investigators to assess the effectiveness of control measures, which could enable characterization of the problem’s root cause.

In the case of the HIV/AIDS pandemic and its recognition in 1981, epidemiological investigations identified risk factors and established the basis for some key interventions approximately 2 years before the causative agent was identified.⁵³ By comparison, for the epidemics of toxic oil syndrome and EMS, epidemiological studies implicating vehicles of exposure (oil and a dietary supplement, respectively) provided a relatively prompt basis for interventions without having determined the specific etiological agents, which have continued to elude identification despite investigative efforts extending from 2 (for EMS) to nearly 3 (for toxic oil syndrome) decades. However, care must be taken in prematurely selecting an intervention because such action might have unintended consequences. For example, steps taken to mitigate one public health threat (e.g., the focal occurrence of swine influenza in humans in 1976) could lead to an action (e.g., the nationwide vaccination program to prevent an epidemic) that results in an unanticipated outcome (e.g., the temporal cluster of cases of Guillain-Barré syndrome), thereby compounding an already complex public health response. A substantive, unanticipated outcome of an entirely different nature was the legal precedent that emerged following the investigation of the nationwide epidemic of menstrual TSS, which clarified legal protections for health information privacy considerations during public investigative responses.⁴⁹

The complexity of societal and other factors affecting the pace of implementation of public health interventions is further illustrated in the investigation of Reye’s syndrome and in the study of the neuropathy epidemic in Cuba during the early 1990s. The experience with Reye’s syndrome, in particular, underscores that even though scientific evidence is of central importance, it is only 1 of several factors that influence decisions about implementing public health interventions. Although Reye’s syndrome was initially described in 1963, not until 1980 were the first case–control studies associating aspirin use and increased risk of Reye’s syndrome in children reported, including an accompanying cautionary message about giving aspirin to children.¹⁷ In June 1982, after completion of multiple expert reviews of the relevant epidemiological evidence, the US Surgeon General issued an advisory against the use of salicylates for children with influenza and chickenpox.⁹⁶ However, not until 1986, after completion of a fifth case–control study supporting the link between aspirin and Reye’s syndrome, did the US Food and Drug Administration rule that aspirin-containing products were required to be labeled with a warning about Reye’s syndrome. Reasons for the substantial, prolonged lapse in time from recognition of this syndrome to promulgation of the various control measures (publicizing the association, issuing an advisory, requiring warning labels) were complex and reflected several challenges, including developing a good case definition that could be used for both fatal and nonfatal cases; dealing with the common use of the putative risk factor, aspirin (used in up to 70% of children with febrile respiratory illnesses), in the setting of a problem of rare occurrence; obtaining comparably reliable, unbiased histories of antipyretic use to treat viral-like illnesses in both case participants and their appropriate control participants; and confronting the efforts of a strong interest group that opposed the recommended science-based public health measure (L. Schonberger, MD, personal communication, July 19, 2011).^19,97

The epidemic of optic and peripheral neuropathies in Cuba during 1991 through 1993 occurred within a context of economic deterioration and associated limited food availability. Public health and other investigators from Cuba and the United States who studied this epidemic assessed multiple potential contributing factors (e.g., nutritional status, other dietary considerations, tobacco use, and other toxin exposure) but were unable to assign a definitive cause.^98,99 In addition, implementation of a population-level nutritional intervention prior to the investigation further complicated interpretation of findings (B. Bowman, PhD, personal communication, August 3, 2011). The overall experience with this epidemic documented the myriad challenges of detecting, investigating, and intervening in a syndromic problem in this complex setting of substantial social and economic transition.

Improve Response Capacities for “Intentional” Events

As we discussed earlier, the greater the time between recognition of a “naturally occurring” event and identification of the causative infectious etiology, the longer it may take to initiate measures to treat affected and exposed persons and, therefore, begin reducing the likelihood of future-generation transmission to others in the at-risk population. However, potential bioterrorist events pose special and complex considerations reflecting the seeming paradox of the controlled—or “nonrandom”—exposure. Most well-characterized infectious pathogens that could be used for biothreat purposes are fairly predictable with regard to the risks they pose. For example, in the case of many naturally occurring diseases (e.g., food-borne, vector-borne, and respiratory), certain commonalities likely will be found during an investigation, even when the source or the etiology, or both, are still under investigation. However, because of deliberate human manipulation of the pathogen and the ability to distribute that pathogen in locations that may be predictably linked with one another, the unfolding of a biothreat event may not conform to, and likely will depart substantially from, the expected epidemiological patterns characteristic of the pathogen that are associated with otherwise “natural” (or nondeliberate) spread.

The anomalous epidemiological patterns associated with deliberate actions causing a bioterrorist event, therefore, necessitate strengthening the means for detecting and responding to such events. In addition to improving specific technological and public health capabilities that contribute to diminishing the elapsed time between initial detection of an event to identification of the etiology, events of the 21st century have underscored the need for expanding public health partnerships to include ongoing collaborations with other sectors and disciplines (e.g., law enforcement, defense, security, agriculture, and transportation) that share in the mission of ensuring the public’s health, safety, and welfare. For instance, since 2001, the CDC has established new or substantially strengthened existing relationships with previously nontraditional partner agencies, such as the US Department of Justice, US Department of Defense, and US Department of Homeland Security, and formed strategic partnerships with agencies such as the US Food and Drug Administration, US Department of Agriculture, and US Environmental Protection Agency.

CONCLUSIONS

The investigations and responses we included not only reflect a broad spectrum of syndromes and affected populations but also present a range of intervention measures that have current, and potentially future, applications. Many of those interventions have been mediated through efforts to educate and modify the voluntary behavior of at-risk groups, including, for example, not using or consuming certain drugs and foods, such as abstaining from the use of aspirin in certain circumstances to reduce the risk of Reye’s syndrome, of specific cooking oils to minimize the risk of toxic oil syndrome, and of l-tryptophan to reduce the risk of EMS. Interventions for other problems have required that persons in at-risk groups and settings take affirmative steps to reduce the likelihood of disease, such as using barrier methods to prevent HIV infection and insect repellants to prevent tick bites and Lyme disease. Another intervention category comprises voluntary and, in some instances, mandatory law-based measures that restrict the movement of some persons, such as the use of isolation and quarantine to curb the spread of airborne communicable diseases (e.g., SARS). The investigation of the cluster of cases of progressive inflammatory neuropathy is an example of a focal, setting-specific (occupational) risk for a severe and noninfectious chronic syndrome for which the specific etiology has not yet been determined but for which it was possible to implement targeted control measures.

Public health agencies at all jurisdictional levels have prepared for, investigated, and responded to disease outbreaks for decades prior to and encompassing the occurrence of the problems summarized in this article. However, in recent years, the demands for and challenges confronting public health preparedness have increased substantially, in part as a consequence of the emergence of the global community in which people, as well as some goods and animals, can reach any point on the planet within a single day. In this environment, some emerging or reemerging disease threats may arrive at US ports of entry from countries in the form of foods or other vehicles not regulated in the country of origin. Although the global community is facing an increase in these and other types of challenges to preparedness, the armamentarium of tools and methods available to address them, including enhanced postresponse evaluation, is on the rise. Continuing analysis of investigations to identify practically relevant lessons can assist in advancing our capacity, capability, and proficiency to recognize and investigate syndromes and other problems; to identify specific etiological pathogens in some categories of disease-causing agents; and to implement the most appropriate disease control interventions and countermeasures.