Lydia C. Medeiros

Food safety education: what should we be teaching to consumers?

Food safety education is most effective when messages are targeted toward changing behaviors most likely to result in foodborne illness. The five major control factors for pathogens are personal hygiene, adequate cooking, avoiding cross-contamination, keeping food at safe temperatures, and avoiding foods from unsafe sources. Pathogens associated with poor personal hygiene have the highest incidence and costs. Inadequate cooking and cross-contamination have lower incidence. Keeping food at safe temperatures and unsafe food sources have the lowest incidence, although costs per case are sometimes very high. We recommend that consumer food safety educators primarily focus on hand washing, adequate cooking, and avoiding cross-contamination. Secondary messages should focus on keeping food at safe temperatures and avoiding food from an unsafe source. Evaluation tools are needed to evaluate self-reported behavior changes. The evaluation questions must focus on salient behaviors that are most likely to result in foodborne illnesses and must withstand rigorous standards of reliability and validity.

Evaluation of Food Safety Education for Consumers

Traditionally, nutrition educators have used a fairly global approach to teach food safety by teaching a broad range of safe food handling behaviors in the expectation that this will lead to the avoidance of foodborne illness. This approach can be confusing and lead to evaluation data that are difficult to interpret. This article suggests that food safety education and evaluation in the future be organized around five behavioral constructs: practice personal hygiene, cook foods adequately, avoid cross-contamination, keep foods at safe temperatures, and avoid food from unsafe sources. These five constructs are derived from data on actual outbreaks and estimated incidences of foodborne illness. Research is needed to establish reliable and valid evaluation measures for these five behavioral constructs. Evaluation instruments can be tailored to fit specific education programs. If evaluation instruments focus on these five behavior areas, the result will be meaningful evaluation data that can be more easily summarized across food safety education programs for consumers.

Observation versus self-report: validation of a consumer food behavior questionnaire

A reliable and validated set of food safety behavior questions that could be used with confidence when evaluating food safety education programs was identified in this study. A list of 29 food-handling and consumption behaviors rank-ordered within five pathogen control factors by nationally recognized food safety experts was the basis for the development of the behavior questions. Questions were evaluated for reliability and several forms of validity. During a kitchen activity session, 70 graduates of a nutrition education program completed four food preparation tasks while being observed and videotaped. The individuals also participated in an in-depth interview to validate behaviors that could not be observed during the food preparation activity, e.g., refraining from preparing food for others when experiencing diarrhea. Criterion validity was established by comparing questionnaire responses to observed behavior and interview responses. Twenty-eight questions met the validity criterion (> or = 70% agreement between observed and interviewed responses and self-reported responses), with three or more questions from each of five pathogen control factor areas. Observation assessments revealed that hand washing was more likely to be performed prior to beginning food preparation than between working with raw meats and fresh produce. Errors in methods of washing hands, utensils, and preparation surfaces between food preparation tasks were common. Most participants did not use thermometers to evaluate doneness but still cooked to safe internal temperatures. The results provide a tool that educators can use to evaluate food safety programs and will help guide the development of more effective food safety education programs targeting needed improvements in behavioral skills.

Consumer food-handling behaviors associated with prevention of 13 foodborne illnesses.

To be effective in reducing the incidence of foodborne illness, consumers and food safety educators need information about behaviors that will decrease exposure to foodborne pathogens. A four-round Delphi technique was used to survey nationally recognized experts in food microbiology, epidemiology, food safety education, and food safety policy with the aim of identifying and ranking food-handling and consumption behaviors associated with 13 major foodborne pathogens. The food safety experts ranked behaviors related to keeping foods at safe temperatures as of primary importance in preventing illness caused by Bacillus cereus and Clostridium perfringens and of secondary importance in preventing illness caused by Staphylococcus aureus. The use of a thermometer to cook foods adequately was ranked as of primary importance for the prevention of illness caused by Campylobacter jejuni, Salmonella species, Escherichia coli O157:H7, Toxoplasma gondii, and Yersinia enterocolitica, with the avoidance of cross-contamination being ranked as of secondary importance for most of these pathogens. Hand washing was the top-ranked behavior for the prevention of shigellosis. The avoidance of certain foods that are likely to be contaminated was the top-ranked behavior for the prevention of illnesses caused by Listeria monocytogenes, Noroviruses, and Vibrio species. The expert panels ranking of behaviors for the reduction of the risk of illness caused by major foodborne pathogens can enable consumers to make informed choices about food consumption and handling behaviors and can guide food safety educators in prioritizing their educational efforts.

Identification and Classification of Consumer Food-Handling Behaviors for Food Safety Education

OBJECTIVE To identify key behaviors associated with major food safety pathogen control factors. DESIGN World Wide Web-based descriptive study with 4 tasks: compile a list of food-handling behaviors from the literature; use Delphi process to build consensus, rank order, and edit lists of behaviors; assess content and construct validity; and review results of Delphi process at a summit meeting of selected experts to make final recommendations for a food safety behavior list. SUBJECTS/SETTING A convenience sample of 10 epidemiologists, 11 food microbiologists, 10 food safety educators, and 10 food safety policy makers was recruited from lists of nationally known experts. Twenty-four experts completed 4 rounds of surveys presented to them via a Web site. Six experts attended a face-to-face meeting to finalize behavior lists. Statistical analyses performed Descriptive statistics and sum of observation data were used to determine rank order and validity. RESULTS Twenty-nine key food-handling behaviors for maintaining the safety of food and reducing the number of cases and outbreaks of foodborne illness were identified. These were rank-ordered within 5 pathogen control factors: practice personal hygiene, cook foods adequately, avoid cross-contamination, keep foods at safe temperatures, and avoid foods from unsafe sources. APPLICATIONS/CONCLUSIONS This list of rank-ordered behaviors grouped by pathogen control factor should help dietitians and educators confidently teach food safety guidance that will be most effective in preventing illness.

Awareness and Acceptance of Current Food Safety Recommendations During Pregnancy

Objectives: To better understand pregnant womens food safety attitudes and beliefs that affect food selection, preparation and handling behaviors, sources of food safety information, motivators and barriers to adopting current recommendations, and preferences for receiving food safety materials. Methods: Eleven focus groups were conducted with 69 women (57 pregnant and 12 less than 6 months postpartum). The Health Belief Model guided development of the moderators guide. In the presence of each focus group, participants completed a food safety attitude/behavior questionnaire. Sessions were audiotaped, transcribed, and analyzed for common themes across and within groups. Results: Most participants indicated moderate concern about food safety and had made some food handling or consumption changes since becoming pregnant; however, many were not following 7 of the 12 specific recommendations discussed. Further, there was resistance to change habits, especially for less well-known recommendations. The women assumed their food is safe, and wanted strong evidence regarding why they should change current practices. Common barriers included lack of prior awareness of most recommendations, no prior illness from implicated foods and the convenience, perceived health benefits of, and personal preference for many risky foods discussed. Participants wanted food safety information that was quick and easy to read, sufficiently thorough, and specifically targeted to pregnant women. Conclusions: The women studied had not internalized the connection between risky food consumption during pregnancy and risk to the unborn child, but expressed interest in valid information that might cause them to change their behaviors. The information gained will be useful in developing food safety educational materials for pregnant women.

A Logic Model Framework for Community Nutrition Education

Logic models are a practical method for systematically collecting impact data for community nutrition efforts, such as the Food Stamp Nutrition Education program. This report describes the process used to develop and test the Community Nutrition Education Logic Model and the results of a pilot study to determine whether national evaluation data could be captured without losing flexibility of programming and evaluation at the state level. The objectives were to develop an evaluation framework based on the Logic Model to include dietary quality, food safety, food security, and shopping behavior/food resource management and to develop a training mechanism for use. The portability feature of the model should allow application to a variety of community education programs.

Inactivation of Listeria monocytogenes, Escherichia coli O157:H7, and Salmonella typhimurium with compounds available in households.

Solutions of selected household products were tested for their effectiveness against Listeria monocytogenes, Escherichia coli O157:H7, and Salmonella Typhimurium. Hydrogen peroxide (1.5 and 3%), vinegar (2.5 and 5% acetic acid), baking soda (11, 33, and 50% sodium bicarbonate), household bleach (0.0314, 0.0933, and 0.670% sodium hypochlorite), 5% acetic acid (prepared from glacial acetic acid), and 5% citric acid solutions were tested against the three pathogens individually (five-strain composites of each, 10(8) CFU/ml) by using a modified AOAC International suspension test at initial temperatures of 25 and 55degrees C for 1 and 10 min. All bleach solutions (pH 8.36 to 10.14) produced a >5-log reduction of all pathogens tested after 1 min at 25 degrees C, whereas all baking soda solutions (pH 7.32 to 7.55) were ineffective (<1-log reduction) even after 10 min at an initial temperature of 55 degrees C. After 1 min at 25 degrees C, 3% hydrogen peroxide (pH 2.75) achieved a >5-log reduction of both Salmonella Typhimurium and E. coli O157:H7, whereas undiluted vinegar (pH 2.58) had a similar effect only against Salmonella Typhimurium. Compared with 1 min at 25 degrees C, greater reductions of L. monocytogenes (P < 0.05) were obtained with all organic acid and hydrogen peroxide treatments after 10 min at an initial temperature of 55 degrees C. The efficacies of household compounds against all tested pathogens decreased in the following order: 0.0314% sodium hypochlorite > 3% hydrogen peroxide > undiluted vinegar and 5% acetic acid > 5% citric acid > baking soda (50% sodium bicarbonate). The sensitivity of the tested pathogens to all tested household compounds followed the sequence of Salmonella Typhimurium > E. coli O157: H7 > L. monocytogenes.

Generational differences in selenium status of women.

In this cross-sectional study of three generations of women, daughters (19–26 yr), mothers (40–58 yr) and maternal grandmothers (67–84 yr) from the same 10 families in central Ohio were studied to determine the effect of life-cycle differences, including estrogen status, on selenium status. Plasma and red blood cell (RBC) selenium and glutathione peroxidase (GPx) activities were determined and typical dietary selenium intakes were calculated from food-frequency questionnaires. Selenium status was lowest in the oldest generation. Plasma selenium of daughters and grandmothers were significantly lower than those of mothers, and plasma GPx and RBC selenium of grandmothers were also lower than those of the mothers. A positive correlation (r=0.42, p<0.04) was found between plasma estrogen and plasma selenium concentrations. Selenium intakes of all groups were adequate and no differences in selenium intakes were found among groups. The results of this study indicate that selenium status fluctuates during the female life cycle and is related to estrogen status.

Differences in Listeria monocytogenes contamination of rural Ohio residences with and without livestock.

To determine the contribution of on-site livestock to the environmental contamination of rural households with Listeria monocytogenes, a total of 1779 environmental and food samples were collected from 26 ruminant-farm households and 26 rural households in Ohio. L. monocytogenes isolates were identified and differentiated using sequence comparisons of the intragenic regions of inlB and inlC. L. monocytogenes was isolated from shoes, 9.6% (20/208); utility gloves, 5.4% (6/111); kitchen sinks, 1.5% (3/204); washing machines, 0.96% (2/204); food, 1.11% (7/631); and animal feces, 8.7% (9/104), over the course of four household visits at monthly intervals. Notably, L. monocytogenes-contaminated shoes were identified more frequently from ruminant farmhouses than from rural households that did not raise ruminants on site (odds ratio = 4.8). L. monocytogenes isolated from animal feces was indistinguishable from strains recovered from shoes and gloves stored in several homes. Our results highlight the potential of the rural household environment as source of L. monocytogenes exposure.