Outbreaks Where Food Workers Have Been Implicated in the Spread of Foodborne Disease - Part 1: Description of the Problem, Methods, and Agents Involved

ABSTRACT

Food workers in many settings have been responsible for foodborne disease outbreaks for decades, and there is no indication that this is diminishing. The Committee on Control of Foodborne Illnesses of the International Association for Food Protection was tasked with collecting and evaluating any data on worker-associated outbreaks. A total of 816 reports with 80,682 cases were collected from events that occurred from 1927 until the first quarter of 2006. Most of the outbreaks reviewed were from the United States, Canada, Europe, and Australia, with relatively few from other parts of the world, indicating the skewed set of data because of availability in the literature or personal contact. Outbreaks were caused by 14 agents: norovirus or probable norovirus (338), Salmonella enterica (151), hepatitis A virus (84), Staphylococcus aureus (53), Shigella spp. (33), Streptococcus Lancefield groups A and G (17), and parasites Cyclospora, Giardia, and Cryptosporidium (23). Streptococcal, staphylococcal, and typhoid outbreaks seem to be diminishing over time; hepatitis A virus remains static, whereas norovirus and maybe nontyphoidal Salmonella are increasing. Multiple foods and multi-ingredient foods were identified most frequently with outbreaks, perhaps because of more frequent hand contact during preparation and serving.

The Centers for Disease Control and Prevention (CDC) estimates there are up to 76 million cases of foodborne illness each year in the United States (31), and some other countries report similarly large numbers. The contribution of the infected food worker (whether symptomatic or not) to these cases has been difficult to establish. Bryan (4) noted that in 18% of 766 outbreaks occurring between 1961 and 1982, a colonized food worker had touched the implicated food. However, the infected food worker was documented as responsible for only 7% of the salmonellosis outbreaks in England and Wales over a 10-year period (42). More recently, the CDC estimated that 20% of foodborne illnesses caused by bacterial agents are a result of transmission from the infected worker (51), which is similar to the earlier percentage of outbreaks determined by Bryan (4). In many outbreaks, it was unclear whether the workers were the cause or the victims of the infections (12, 14). This is partly because the outbreaks are not thoroughly enough investigated and partly because the disease transmission patterns are complex. More specifically, investigations are often hampered because (i) there is too long a delay between the outbreak event and the start of the investigation, with the likelihood that the persons involved in the outbreak are no longer available for further questioning or have forgotten the details; (ii) the information is limited because of language difficulties or poor employee communication skills; or (iii) there is ineffective questioning by the investigators (26, 39, 53).

Many of the outbreaks reported in the literature where the contribution of the food worker to the case numbers was investigated were decades old; there appears to be less interest today in reporting details. This could be because it is assumed that we know all there is to understand about worker involvement, and there are fewer resources assigned to make complete investigations, especially for small outbreaks. However, because outbreaks involving food workers still continue today, there is a need for a more comprehensive assessment of the role they play in disease transmission. For instance, in the Lansing, Michigan, area, there were three large outbreaks involving restaurants in the spring of 2006 in which food workers were known or suspected to have been the cause of approximately 800 norovirus infections (18, 43).

A review on the involvement of the ill or asymptomatic food worker in foodborne illness outbreaks was initiated as a project of the Committee on Control of Foodborne Illnesses of the International Association for Food Protection (IAFP). The Committee on Control of Foodborne Illnesses decided that the database should include outbreak data from homes, restaurants, institutions, processing plants, and farms from both the United States and other countries. However, it was recognized that the review was far from a complete analysis of all the available information, and this article should be considered an initial report. The goal of the study was to develop an understanding of the dynamics of transmission of infectious agents to and from the food worker in a variety of settings. The Committee on Control of Foodborne Illnesses approached the task with the premise that all foodborne illness is fundamentally preventable and that by influencing human behavior, there will be fewer opportunities for the spread of infectious disease agents and, thus, human infections.

This article is the first of a series of several that review the role of food workers in foodborne outbreaks. It contains the rationale for reviewing the data, the methodology used, and a summary of the general results. The remaining articles will categorize the outbreaks by worker involvement, risk factors, and means of prevention.

METHODS

Review of existing literature and criteria used. Outbreak data available from 1927 to the present were obtained in which food workers were reported to have been instrumental or at least contributory to an outbreak. An outbreak was defined as two or more persons infected or intoxicated after consuming a food that had been linked epidemiologically or microbiologically to the ill persons. Water and ice used in beverages are included as food. Secondary cases arising from contact with any of those who became ill because of contaminated food or contact with an infected food worker were noted and excluded in the listed case numbers. The term food worker is used in this context to describe individuals who harvest, process, prepare, and serve food. By definition, the task of food handlers is more limited to preparation and serving duties, but both worker and handler are often used interchangeably in investigative reports and in the literature. Thus, we use food worker to describe both worker and handler in this study.

Criteria for selection of outbreaks. The data used in assessing the role of the food worker in outbreaks were derived from a variety of published and unpublished sources. These articles were identified through searches of whole text abstracts and outbreak summaries documented by MEDLINE with key words or phrases pertaining to foodborne illness in various segments of the food industry, including restaurants, delicatessens, hospitals, catering establishments, cruise ships, airplanes, trains, camps, cafeterias, and homes, and were as follows: food preparer, food handler, food worker, ill worker, ill employee, asymptomatic carrier, infected employee, excreter, kitchen help, family transmission, household illness, household transmission, outbreak, hand contamination, and cross-contamination. In addition, searches were made by specific disease, e.g., salmonellosis, linked with worker, handler, staff, and food service. Food-associated key words were seafood, poultry, bakery goods, cheese and dairy, produce, salads, sandwiches, meat, hors d'oeuvres, and ready-to-eat (RTE) food. We also requested and obtained outbreak data over a multiyear period from the states of Michigan (2000 to 2003), Minnesota (1999 to 2004), New York (1985 to 2000), and Washington (1990 to 2003). Because one of us (C.A.B.) was employed by the Washington Department of Health and was involved with foodborne disease investigations during this time frame, much additional information came from personal communication. Data from individual states were received in the form of line listings through the respective state Departments of Health and, for Michigan, also from the Department of Agriculture. Line listings were also obtained from annual reports of foodborne and waterborne disease outbreaks published by Health Canada (1976 to 1996). Line listings are summaries of narrative reports of outbreaks in a tabular format, typically expressed in a few lines of text, with information, when available, on etiological agent, date of onset, location, food mishandling location, food vehicle, number of persons ill and number exposed, incubation period, duration, symptoms, laboratory data, factors contributing to the outbreak, and other relevant data.

Most of these outbreaks showed a factor such as handling by an infected person or carrier of a pathogen. However, a few selected outbreaks were included where strong epidemiological data suggested that food workers were the likely source of the pathogen, but all food workers denied illness, or else the patrons themselves, rather than the food workers, were identified as the likely source of the pathogen. These line listings may or may not have been accompanied by more detailed information through separate reports or appendices; however, these were rarely available to the reviewers. All these sources were reviewed by the authors, and selections were made on the basis of the completeness and relevance of the information. Even so, it was recognized that the role of the worker in some reports was much more clearly stated than in others. References and other comments were obtained from existing reviews (19, 32, 46).

Criteria for selection of factors contributing to the outbreaks. The authors searched the available information from the outbreaks selected for review, whether contained in a peer-reviewed publication, line listing, or narrative, for any pertinent factors that contributed to the occurrence of the outbreaks. These were written in English or translated from another language. The data could be evaluated only as presented. So, undoubtedly, some data were missed or not included in some of the reviewed accounts. One key aspect of outbreak investigation is the identification of factors contributing to outbreaks. CDC form 52.13 (Revised 11-2004), "Investigation of a Foodborne Outbreak," was the basis for the majority of the factors used in this study (11). The factors identified in the form are based on earlier research done by Bryan and others on factors related to foodborne outbreaks (5, 6, 52) and are coded C, P, and S. Factors identified with a "C" are contamination factors, while those with a "P" refer to factors that allow proliferation or amplification of bacterial pathogens. An "S" designates factors that allow pathogens to survive in the food. In addition, another factor was used from the Washington State Health Department coded as C-15, "failure to properly wash hands when necessary." C-15 included some of the following types of observations: food workers' hands were not washed after using the toilet; running water was not available for hand washing; no soap or towels were used; or food workers failed to wash their hands after contaminating events occurred (e.g., handling raw meat). Factors reported as linked to an outbreak should have occurred near the time of that outbreak. If the factors were observed either earlier or later (e.g., during routine inspections), they would not be reported, although sometimes it is not possible to tell exactly when observations were made regarding factors in the reports. Frequently, during the investigation of viral or parasitic outbreaks, P factors related to temperature abuse or conducive to bacterial growth were excluded, because viruses or parasites do not grow in foods, even though they were noted in the investigative reports. These same factors, however, were included in reviewed outbreaks of bacterial etiology.

Factors listed in the outbreak reports involving food workers were as follows:

. . .Continue to read rest of article (PDF).