Perspectives

Outbreak Investigations—A Perspective

Arthur L. Reingold University of California, Berkeley, California, USA

Outbreak investigations, an important and challenging component of and public health, can help identify the source of ongoing outbreaks and prevent additional cases. Even when an outbreak is over, a thorough epidemiologic and environmental investigation often can increase our knowledge of a given and prevent future outbreaks. Finally, outbreak investigations provide epidemiologic training and foster cooperation between the clinical and public health communities.

Investigations of acute infectious disease outbreaks of known , as in the case of outbreaks are very common, and the results of B among the patients of an oral such investigations are often published; surgeon in Connecticut and patients at a weight however, surprisingly little has been written reduction clinic (6,7). The former outbreak was about the actual procedures followed during first suspected when routinely submitted such investigations (1,2). Most epidemiologists communicable disease report forms for several and public health officials learn the procedures patients from one small town indicated that all by conducting investigations with the initial of the patients had recently had oral surgery. assistance of more experienced colleagues. This However, it is relatively uncommon for article outlines the general approach to outbreaks to be detected in this way and even conducting an outbreak investigation. The more uncommon for them to be detected in this approach applies not only to infectious disease way while they are still in progress. Finally, outbreaks but also to outbreaks due to sometimes public health officials learn about noninfectious causes (e.g., toxic exposure). outbreaks of disease from the local newspaper or television news. How Outbreaks Are Recognized Possible outbreaks of disease come to the Reasons for Investigating Outbreaks attention of public health officials in various The most compelling reason to investigate a ways. Often, an astute clinician, infection control recognized outbreak of disease is that exposure to nurse, or clinical laboratory worker first notices the source(s) of infection may be continuing; by an unusual disease or an unusual number of identifying and eliminating the source of cases of a disease and alerts public health infection, we can prevent additional cases. For officials. For example, staphylococcal toxic shock example, if cans of mushrooms containing syndrome and eosinophilia myalgia syndrome botulinum toxin are still on store shelves or in were first noted by clinicians (3,4). Frequently, it homes or restaurants, their recall and destruc- is the patient (or someone close to the patient) tion can prevent further cases of botulism. who first suspects a problem, as is often the case However, even if an outbreak is essentially in foodborne outbreaks after a shared meal and over by the time the epidemiologic investigation as was the case in the investigation of a cluster of begins—that is, if no one is being further exposed cases of apparent juvenile rheumatoid arthritis to the source of infection—investigating the near Lyme, Connecticut, which led to the outbreak may still be indicated for many reasons. discovery of Lyme disease (5). Review of routinely Foremost is that the results of the investigation collected surveillance data can also detect may lead to recommendations or strategies for preventing similar future outbreaks. For Address for correspondence: Arthur L. Reingold, Division of example, a Legionnaires’ disease outbreak Public Health Biology and Epidemiology, School of Public investigation may produce recommendations Health, University of California, Berkeley, 140 Warren Hall, for grocery store misting machine use that may Berkeley, CA 94720-7360, USA; fax: 510-643-5163; e-mail: [email protected]. prevent other outbreaks (8). Other reasons for

Vol. 4, No. 1, January–March 1998 21 Emerging Infectious Diseases Perspectives

investigating outbreaks are the opportunity to 1) Case Definition describe new diseases and learn more about known In some outbreaks, formulating the case diseases; 2) evaluate existing prevention strategies, definition(s) and exclusion criteria is straightfor- e.g., vaccines; 3) teach (and learn) epidemiology; ward; for example, in an outbreak of gastroenteri- and 4) address public concern about the outbreak. tis caused by Salmonella infection, a laboratory- Once a decision is made to investigate an confirmed case would be defined as a culture- outbreak, three types of activities are generally confirmed infection with Salmonella or perhaps involved—the epidemiologic investigation; the with Salmonella of the particular serotype environmental investigation; and the interaction causing the outbreak, while a clinical case with the public, the press, and, in many definition might be new onset of diarrhea. In instances, the legal system. While these activities other outbreaks, the case definition and exclusion often occur simultaneously throughout the criteria are complex, particularly if the disease is investigation, it is conceptually easier to consider new and the range of clinical manifestations is each of them separately. unknown (e.g., in a putative outbreak of chronic fatigue syndrome). In many outbreak investiga- Epidemiologic Investigation tions, multiple case definitions are used (e.g., Outbreak investigations are, in theory, laboratory-confirmed case vs. clinical case; definite indistinguishable from other epidemiologic in- vs. probable vs. possible case; outbreak-associated vestigations; however, outbreak investigations case vs. nonoutbreak-associated case, primary case encounter more constraints. 1) If the outbreak is vs. secondary case) and the resulting data are ongoing at the time of the investigation, there is analyzed by using different case definitions. great urgency to find the source and prevent When the number of cases available for study is additional cases. 2) Because outbreak investiga- not a limiting factor and a case-control study is tions frequently are public, there is substantial being used to examine risk factors for becoming pressure to conclude them rapidly, particularly if a case, a strict case definition is often the outbreak is ongoing. 3) In many outbreaks, preferable to increase specificity and reduce the number of cases available for study is limited; misclassification of disease status (i.e., reduce therefore, the statistical power of the investiga- the chance of including cases of unrelated tion is limited. 4) Early media reports concerning illness or no illness as outbreak-related cases). the outbreak may bias the responses of persons subsequently interviewed. 5) Because of legal Case Confirmation liability and the financial interests of persons and In certain outbreaks, clinical findings in institutions involved, there is pressure to reported cases should be reviewed closely, either conclude the investigation quickly, which may directly, by examining the patients, or indirectly, lead to hasty decisions regarding the source of the by detailed review of the medical records and outbreak. 6) If detection of the outbreak is delayed, discussion with the attending health-care useful clinical and environmental samples may be provider(s), especially when a new disease very difficult or impossible to obtain. appears to be emerging (e.g., in the early Outbreak investigations have essential com- investigations of Legionnaires’ disease, AIDS, ponents as follows: 1) establish case definition(s); eosinophilia myalgia syndrome, and hantavirus 2) confirm that cases are “real”; 3) establish the pulmonary syndrome) (4,9-11). Clinical findings background rate of disease; 4) find cases, decide if should also be examined closely when some or all the