1939 Journal o f Food Protection, Vol. 76, No. 11, 2013, Pages 1939-1947 doi: 10.4315/0362-028X.JFP-12-269 C o p yrig h t © , International A sso ciation fo r Food Protection

Food Safety Practices among Norwegian Consumers ELIN HALBACH R0SSVOLL,1 RANDI LAVIK,2 0YDIS UELAND,1 EIVIND JACOBSEN,2 THERESE HAGTVEDT,1 a n d SOLVEIG LANGSRUD1* 1Nofima, Norwegian Institute o f Food, Fisheries and Aquaculture Research, P.O. Box 210, N-1431 As, Norway; and 2S1F0, National Institute fo r Consumer Research, P.O. Box 4682 Nvdalen, N-0405 Oslo, Nonvay MS 12-269: Received 20 June 2012/Accepted 19 April 2013

ABSTRACT An informed consumer can compensate for several potential food safety violations or contaminations that may occur earlier in the food production chain. However, a consumer can also destroy the work of others in the chain by poor food handling practices, e.g., by storing chilled ready-to-eat foods at abusive temperatures. To target risk-reducing strategies, consumer groups with high-risk behavior should be identified. The aim of this study was to identify demographic characteristics associated with high-risk food handling practices among Norwegian consumers. More than 2,000 randomly selected Norwegian consumers were surveyed, and the results were analyzed with a risk-based grading system, awarding demerit points for self-reported food safety violations. The violations were categorized into groups, and an ordinary multiple linear regression analysis was run on the summarized demerit score for each group and for the entire survey group as a whole. Young and elderly men were identified as the least informed consumer groups with the most unsafe practices regarding food safety. Single persons reported poorer practices than those in a relationship. People with higher education reported poorer practices than those with lower or no education, and those living in the capital of Norway (Oslo) reported following more unsafe food practices than people living elsewhere in Norway. Men reported poorer food safety practices than women in all categories with two exceptions: parboiling raw vegetables before consumption and knowledge of refrigerator temperature. These findings suggest that risk-reducing measures should target men, and a strategy is needed to change their behavior and attitudes.

Food contamination creates a significant social and economic burden on communities and their health systems, and food safety is an increasingly important public health issue. In industrialized countries, up to 30% of people annually suffer from foodbome illnesses (57). The true frequency of instances of foodbome illness is difficult to ascertain because of underreporting, and the proportion of foodbome illnesses associated with foods eaten in the home is difficult to determine because these cases are generally sporadic (18, 57). Krause et al. (38) analyzed 30,578 outbreaks of infectious diseases that were reported in Germany from 2001 through 2005 and found that 90% of these outbreaks were caused by gastrointestinal pathogens such as Salmonella, norovirus, rotavirus, hepatitis A vims, enteropathogenic Escherichia coli, and Campylobacter sp. Using data from 2004 and 2005, the first 2 years that the location setting was included in the outbreak reports, Krause et al. (38) identified the home environment as the origin of 53% of the 9,946 outbreaks reported. Redmond and Griffith (50) reviewed 10 years of consumer research on food safety that summarized reported foodbome outbreaks in the United Kingdom, Europe, Australia, New Zealand, the United States, and Canada and found that 9 to nearly 90% of these outbreaks were * Author for correspondence. Tel: +4764970100; Fax: +4764970333: E-mail: [email protected].

caused by food prepared in the home, with high variation among countries. Not all foodbome illnesses attributed to food eaten in the home are the fault of the preparer (26), e.g., contamination of peanut butter with Salmonella (9,10, 20). Therefore, some of these incidents could have been prevented by eliminating pathogens earlier in the food production chain. However, some incidents are probably directly caused by improper consumer handling of food or at least by a combination of faults at various stages of the production chain, including the consumer (43). The number of foodbome illness outbreaks that could be prevented by better consumer handling practices is probably high but is hard to determine because few of these cases are reported and investigated (30). Thus, more attention should be given to factors that can ultimately reduce the prevalence of foodbome diseases caused by food prepared in the home. Depending on the characteristics of the food risks, different strategies for disease prevention may be employed. Knowledge of what characterizes consumers or consumer segments at risk and identification of critical hazard points associated with consumer handling practices may help us target information and dissemination strategies more effectively. Improved knowledge of consumer food handling practices can be used by food safety authorities, the food industry, schools, and other educational facilities to target specific food safety information to the consumers who need it. For the food

1940

R0SSVOLL ET AL.

industry, this knowledge can also be used to develop products with lower risk for causing foodborne illness. One example is ready-cut meat that requires less handling by the food preparer, thus reducing the opportunity for cross­ contamination in the home. In Norway, new food habits have arisen from the increased availability of new and unfamiliar foods imported from all over the world, including countries where pathogens such as Salmonella and enterohemorrhagic E. coll are more commonly associated with food products (53). Thus, food risks and risk-reducing strategies known in other parts of the world may be unfamiliar to Norwegian consumers (47). At present, there is little information on Norwegian consumers’ knowledge of microbial risks and hygienic handling of food, their perception of the home as a location for foodborne illness, their attitudes toward food hygiene, and their actual food handling behaviors. In studies from other countries, young men with higher education have been identified as the group with the most unsafe practices with regard to food safety (1, 2, 8,17,33,39,45,48,52). In addition to young men and elderly single men, Brennan et al. (6) identified two additional high-risk groups with regard to microbiological food safety in Ireland: young women and older women (45 + years of age). The young women were not in paid employment and were without home economics training. The older women were also not in paid employment but had home economics training. The older women were not expected to be characterized as a high-risk group, and Brennan et al. (6) presented two possible explanations for this surprising result. Much has changed since these women received their formal food safety training 25 or more years ago, and these women may have been overconfident in their own judgment. The continuing globalization of the food market is introducing both new types of foods and new types of pathogens, which require new approaches to preparation and hygienic practices (53). The old rules for food handling that were taught 25 to 30 years ago may not pertain to new dishes and new pathogens. The findings of Brennan et al. (6) illustrate the complexity of the study of consumer food safety and how studies might fail to include important subgroups. Such studies can still provide a good starting point for identification of the critical steps in consumers’ food handling practices but will provide insufficient information to draw conclusions about the present conditions in Norway because of different risk scenarios concerning eating habits, culture, climate and so forth. The World Health Organization stated that “ education of consumers and training of food handlers in safe food­ handling is one of the most critical interventions in the prevention of foodborne illnesses” (57). A resourceful and well-informed consumer, as the last line of defense in food safety, can compensate for food safety errors that might occur earlier in the food production chain (5). However, consumers also are in a position, through their own unsafe food handling practices, to undo much of the good work done previously in the food production chain. In a previous study, we identified and ranked the most high-risk food handling practices among Norwegian consumers (53).

J. Food Prot., Vol. 76, No. 11

However, education and information should both focus on the most common food handling practices of the average consumer and be targeted toward vulnerable consumer groups (such as the elderly, children, and pregnant women) and groups that are most likely to handle food improperly. The aim of the present study was to identify demographic characteristics associated with high-risk behavior with regard to food handling and consumption practices among Norwegian consumers to provide a better foundation for planning risk reduction strategies. M A TER IA LS A N D M E TH O D S A W eb-enabled food safety survey was e-mailed to a random sample of Norwegian consumers listed in a consumer panel database maintained by Norstat (46), a professional market research company. The sample was selected proportionally to the population in different parts o f Norway, and quotas were allocated for gender and county. Respondents were required to be between the ages of 18 and 80 years. The data were weighted to adjust the sample to the current population distribution. In addition to Norstat’s own incentive program, all respondents who completed the Web survey had the opportunity of winning a gift voucher of NOK 5,000 (27). The questionnaire was part o f a broader interdisciplinary study with a total of 54 questions. The questionnaire took approximately 20 min to complete (27, 53).

Questionnaire development. The study method has been described in detail elsewhere (53). A hazard analysis and critical control point (HACCP) plan was made for each of four food preparation processes that normally take place in a Norwegian domestic kitchen. The HACCP flow diagrams were validated by interviewing and observing consumers in their own kitchens. A consumer food safety questionnaire was designed and developed from the HACCP analyses. The W eb-enabled food safety survey was conducted, and the results were analyzed with a risk-based grading system based on Worsfold and Griffith’s food operation risk score (21, 58, 59). The self-reported answers were given a score that was based on demerit points awarded for food safety and hygienic violations. A frequency coefficient of 0 to 1 was assigned according to how often respondents reported performing the specific violation and was multiplied by the violation’s demerit point to get a risk score (Table 1). Statistical analysis. The independent variables chosen were age and gender, marital status, children in the household, the household’s total pretax income, the highest level of education obtained, occupational status, and area o f residence. The residence variable comprised five options, with options 2 through 5 compared individually against option 1: option 1, Oslo (the capital and largest city in Norway); option 2, a large city o f more than 50,000 inhabitants; option 3, a small city of 5,000 to 50,000 inhabitants; option 4, a small town; and option 5, a rural area. These demographic variables were chosen because they are common socioeconomic background variables that can explain differences in the risk variables (11,16). MX demographic variables except age were categorized into dummy variables. Age was categorized as a continuous variable and was squared; we hypothesized that the relationship between age and the dependent variables was not linear but rather curvilinear, as Fein et al. (15) found in their analysis o f food safety surveys of consumers in the United States. Age and age squared must be interpreted together. A negative P value for age and a positive P value for age squared give a U-shaped curve, in which the middle age group is on the

J. Food Prot., Vol. 76, No. 11

FOOD SAFETY KNOWLEDGE IN NORWAY

1941

TABLE 1. Questions from the consumer food safety survey and demerit score distribution % respondents awarded demerit scores for each question*

Question

Demerit points"

Maximum score"

Middle scorerf

Minimum score"

Factor loading

Consumption of high-risk foods

Cronbach’s alpha 0.59

How often do you eat the following? Half-fermented fish (rakfisk) or cured salmon (gravlaks) Unpasteurized cheese Raw meat as steak tartare or carpaccio Sprouts (bean or alfalfa sprouts)

50 50 50 50

1.5 4.8 1.1 4.7

28.9 33.6 12.8 38.1

69.6 61.6 86.0 57.2

0.72 0.76 0.58 0.59

Consumption of undercooked meat

0.61

To what extent do you agree with the following? I prefer hamburger pink in the middle (not completely done) I prefer chicken fillet pink in the middle (not completely done)

50

12.1

19.8

68.1

0.81

50

4.0

11.3

84.7

0.84

Cross-contamination practices

0.58

After cutting raw meat, what do you do with the knife? Continuing using the knife as it is

10

3.6

21.0

75.4

0.63

10

2.0

28.2

69.8

0.70

15

0.4

15.8

83.7

0.56

15

4.3

13.0

82.7

0.68

How often do you do the following? Using soap when washing hands during food preparation Wash hands immediately after handling raw meat The last time you were handling raw meat, how did you clean your hands?^ Hygiene practices

0.53

How often do you do the following? Wash Wash Wash Wash

hands hands hands hands

before preparing food after blowing your nose after playing with a pet after toilet visits

10 10 10 15

Total risk (all questions)

0.2 1.7 1.7 0.1

5.4 25.4 19.3 2.2

94.5 72.9 79.0 97.7

0.57 0.67 0.70 0.61 0.60

a From Rrissvoll et al. (53). b n = 2,008. c Frequency coefficient = 1. d Frequency coefficient = 0.5. e Frequency coefficient = 0. f Possible responses for this question: did not clean them; wiped them on a towel, kitchen towel, or kitchen cloth; rinsed them in cold water (awarded maximum demerit score); rinsed them in warm water; wiped them on paper towels (middle demerit score); washed with soap and warm water; wiped on a disinfection wipe; cannot remember; none of these (minimum demerit score).

minimum point of the curve with fewest demerit points. A positive P value for age and a negative P value for age squared give an inverted U-shaped curve, where the maximum point of the curve is the middle age group with the highest demerit score. To detect underlying dimensions of the dependent variables, an exploratory factor analysis was conducted. The variables and questions are presented in Table 1. The individual risk score assigned for all questions in one group were added together, and a regression analysis was run on the summarized risk score for each group variable. All risk scores for the complete questionnaire were summed into a total group variable, i.e., the total risk score, and a regression analysis was performed on this total risk score. Some of the dependent variables did not give any reasonable factor loading and were excluded from the factor analysis. However, because we

found these variables interesting as one-indicator variables, we conducted two cross-tabulation analyses of these single variables by gender. All food safety questions are listed in their entirety elsewhere (53). The questionnaire responses were analyzed using SPSS Statistics version 19 (IBM, Armonk, NY). Table 2 shows the multiple linear regression analysis conducted to estimate values. To compare the unstandardized b coefficients between the latent variables, the dependent variables were standardized to range from 0 to 100. The values can be interpreted as percentages. The b coefficients will be read as a 1-unit change in the independent variable, i.e., as how many percentage points will be increased or decreased in the dependent variable, controlling for the other independent variable(s) in the equation. The unstandardized b

1942

R0SSVOLL ET AL.

J. Food Prot., Vol. 76, No. 11

TABLE 2. Multiple linear regression analysis on the demerit score o f self-reported consumer fo o d handling and consumption practicesa Consumption of high-risk foods

Consumption of undercooked meat

Cross-contamination practices

Hygiene practices

Total risk

Variable

b

P

b

P

b

P

b

P

b

P

Constant Women Age Age2 Relationship Children University, lower degree University, higher degree Education other Live in large city Live in small city Live in small town Live in rural area Income, NOK 301,000-600,000 Income, NOK 601,000-900,000 Income, NOK >900,000 Do not know income In paid employment Adjusted R2

35.90 -1 .6 8 - 0 .6 9 0.01 -2 .1 8 -0 .6 5 1.43 3.66 3.32 - 4 .5 0 - 6 .4 0 - 8 .1 0 -5 .9 9 - 0 .7 4 2.09 6.36 0.61 0.02 0.052

Food safety practices among Norwegian consumers.

An informed consumer can compensate for several potential food safety violations or contaminations that may occur earlier in the food production chain...
5MB Sizes 0 Downloads 0 Views