- Click here for explanation of classification scheme.

To determine the prevalence of unsafe food practices of individuals in a Canadian-based population, specifically, high-risk food consumption and home food safety.

• English-speaking
• Older than 18 months.

Traveling outside of Canada in the seven days before the interview.

Recruitment

Residents of the Waterloo Region, Ontario, Canada who were randomly selected based on residential telephone numbers.

Design

Cross-sectional study.

Statistical Analysis

• Individuals who refused to answer a question were excluded from the analysis of that question
• Prevalence estimates were calculated for the consumption of high-risk food items and for food handling and preparation habits
• Exact confidence intervals (CIs) were computed for all proportions at the 95% level
• Differences in proportions were tested with Pearson's chi square (x²) test and Fischer's exact test when the expected frequencies for a contingency table were less than five
• To assess the relationship between high-risk practices and demographic variables, odds ratios (ORs) were used to estimate the strength of association for statistically significant differences between binary variables such as gender and the Cochran-Armitage test for trend was used to test for linear trends among ordinal variables such as age, income and education
• Multiple x² tests were then used for any non-linear trends identified among age, income and education groups as well as for non-ordered variables such as residence
• Data on high-risk practices were sub-divided into single demographic variables
• Cross-tabulations and x² were used to determine differences between high-risk food consumption practices and food-handling behaviors
• ORs were used to estimate the strength of association for statistically significant differences between high-risk practices
• In all analyses, statistically significant differences were determined by a two-tailed probability of less than 0.05.

Timing of Measurements

Telephone survey occurred from November 2005 to March 2006.

Dependent Variables

• Consumption of high-risk foods
• Prevalence of unsafe food practices:
  • Hand-washing practices
  • Safe food-handling knowledge
  • Source of food safety education
  • Meat thawing and cooking practices
  • Cross-contamination after raw food preparation
  • Refrigeration temperatures.

Independent Variables

Demographics of the Canadian community:

• Age
• Income
• Education
• Type of living area.
  

• Initial N: 7,142 contacted and eligible respondents
• Attrition (final N): 2,332 completed interviews
• Age: Described elsewhere
• Ethnicity: Described elsewhere
• Other relevant demographics: Described elsewhere
• Location: Waterloo Region, Ontario, Canada.

Key Findings

High-Risk Food Items 

• The prevalence of drinking unpasteurized milk decreased with increasing education level (P=0.009). Drinking unpasteurized milk was significantly more prevalent among rural residents (9%) than among urban residents (0.4%, P<0.001). However, the prevalence of consumption of cheese made from unpasteurized milk was not significantly different between rural residents (3%) and urban residents (0.9%, P=0.075).
• The prevalence of undercooked egg consumption increased with increasing age (P<0.001), education (P<0.001) and income (P=0.0004). Eating undercooked eggs was significantly more common among suburban residents (49.6%) than among urban residents (40.5%, P=0.006). Overall, fewer (5.9%) respondents reported eating foods containing raw eggs, such as cookie dough, cake batter or steak tartar; the prevalence of this behavior increased with increasing income level (P<0.001).
• Males were significantly more likely than females to consume chicken nuggets (OR=1.37; 95% CI=1.09, 1.73; P=0.005). Chicken nugget consumption increased with increasing income level (P=0.006) and decreased with increasing education (P<0.001). Consumption was highest (59.9%) in children younger than 12 years and decreased with increasing age (P<0.001).
• Individuals aged 13 to 64 years were significantly more likely to consume spouts than were individuals older than 64 years of age (OR=1.6; 95% CI=1.05, 2.51; P=0.025). Males were significantly more likely to drink unpasteurized juice than were females (OR=1.48; 95% CI=1.03, 2.1; P=0.025). Raw shellfish consumption increased as education level increased (P=0.011).
• Males were significantly less likely to eat lettuce (OR=.65; 95% CI=0.51, 0.82; P<0.001) and pre-bagged mixed salad greens (OR=0.77; 95% CI=0.65, 0.93; P=0.004) than were females. Children younger than 12 years were significantly less likely to eat lettuce (OR=0.19; 95% CI=0.12, 0.29; P<0.001 and pre-bagged mixed salad greens (OR=0.52; 95% CI=0.34, 0.79; P=0.001) than were elderly individuals. Adults aged 25 to 54 years were significantly more likely to eat lettuce than were the elderly (OR=1.45; 95% CI=1.01, 2.07; P=0.033).
• Suburban residents (79.4%) were significantly less likely to eat lettuce than urban residents were (85.5%, P=0.006), whereas suburban residents (49.8%) were significantly more likely than were urban residents (37.6%, P<0.001) to eat pre-bagged mixed salad greens. Eating lettuce increased as education (P<0.001) and total annual household income level (P<0.001) increased.
• Males were significantly more likely to consume fresh basil than were females (OR=1.33; 95% CI=1.01, 1.73; P=0.032.) and adults aged 18 to 64 years were significantly more likely to eat fresh basil than were elderly respondents (OR=1.78; 95% CI=1.17, 2.8; P=0.005)
• Children younger than 12 years of age were significantly more likely to consume melon than elderly respondents were (OR=1.51; 95% CI=1, 2.28; P=0.04). As total annual household income level increased, the prevalence of eating melon increased (P=0.015). Residents in suburban areas (33.6%) were significantly more likely to consume melon than were residents of urban areas (22.1%, P<0.001).

Hand-washing Practices

• Males were significantly more likely to not always wash their hands before eating or handling food (OR=1.61; 95% CI=1.34, 1.94; P<0.001) and to not wash their hands with soap and water after handling raw meal (OR=2.55; 95% CI=1.78, 3.66; P<0.001) than were females. Also, males were significantly more likely to think that hand washing is not important in the prevention of disease (OR=2.11; 95% CI=1.31, 3.41; P=0.001) compared with females. As age increased, both the prevalence of people who always washed their hands before eating or handling food (P<0.001) and those who thought hand washing was important in the prevention of disease (P<0.001) increased.
• As education level increased, the prevalence of people who thought hand washing was important in the prevention of disease (P<0.001) increased. Respondents without a high school diploma were significantly less likely to always wash their hands before eating or handling food compared with respondents with an education beyond high school (P<0.001) .  Respondents with a high school or college diploma were significantly more likely to always wash their hands with soap and water after handling food compared with respondents without a high school diploma (P=0.045 and P=0.001, respectively).  
• As total household income level increased, always washing hands before eating or handling food (P=0.002) and reporting that hand washing was important in the prevention of disease (P<0.001) decreased
• Suburban residents were significantly more likely to always wash their hands before eating or handling food than were urban residents (P<0.001). Rural residents were significantly more likely than were urban residents to always wash hands with soap and water after handling raw meat. (P=0.042).

Individual Food Safety Knowledge

• Respondents who had been given information about foods to avoid eating were significantly less likely  to eat undercooked eggs (OR=0.78; 95% CI=0.65, 0.95; P=0.009) and chicken nuggets (OR=0.52; 95% CI=0.41, 0.66; P<0.001) than were those who had not been given information. However, respondents who had been given information about foods to avoid eating were significantly more likely to eat raw nuts (OR=1.32; 95% CI=1.03, 1.67; P=0.022), lettuce (OR=1.37; 95% CI=1.08, 1.73; P=0.007) and fresh basil (OR=1.56; 95% CI=1.19, 2.04; P=0.001).
• Respondents who reported electronic media as a source of information were significantly less likely to consume:

Less Likely to Consume	OR	95% CI	P
Sprouts	0.68	0.49, 0.93	0.012
Chicken Nuggets	0.35	0.26, 0.47	<0.001
Unpasteurized Juice	0.45	0.27, 0.071	<0.001
More Likely to Consume
Lettuce	1.53	1.17, 2.02	0.001

• Raw oyster was the only high-risk food item whose consumption was significantly associated with knowledge of associated disease risks. Respondents who had heard of the risks associated with raw oysters were significantly more likely (OR=1.96; 95% CI=1.36, 2.84; P<0.001) to eat raw shellfish than were those that had not heard of risks.
•  Males were significantly less likely than females were to have heard of food-borne disease risks associated with:

Food	OR	95% CI	P
Hamburger	0.47	0.38, 0.59	<0.001
Chicken	0.61	0.48, 0.76	<0.001
Unpasteurized juice	0.72	0.59, 0.88	<0.001
Sprouts	0.57	0.48, 0.69	<0.001

• As age increased, knowledge of disease risks associated with consumption of each food item increased, then decreased among individuals over the age of 64 years. Knowledge of disease risks associated with consumption of each food item increased with increasing education level (P<0.001). Individuals who reported a total annual household income between Can$40,000 and Can$80,000 were significantly more likely to have heard of risks associated with raw oysters (47.6%) than were individuals who reported a total annual household income of less than Can$20,000 (30.2%, P<0.001).
• Knowledge of risks associated with drinking unpasteurized juice was significantly higher among individuals with a total annual household income between Can$60,000 and Can$80,000 (40.7%) than among individuals with a total annual household income of less than Can$20,000 (28.6%, P=0.026). Rural residents (68%) were significantly more likely to have heard of disease risks associated with drinking unpasteurized milk than were urban residents (53.4%, P=0.004), while suburban residents were significantly less likely to have heard of risks associated with drinking unpasteurized milk (38.3%) and unpasteurized juice (21.8%) then were urban residents (53.4%, P<0.001 and 30.4%, P=0.002, respectively).

Household Food-Handling and Cooking Practices

• As age increased, the prevalence of noticing a safe handling label on meat packages increased (P=0.011). Urban residents (53.6%) were significantly more likely to notice safe handling labels on meat packages than were all other residents (41.1%, P<0.001). Males were significantly more likely to thaw frozen meat at room temperature than were females (OR=1.22; 95% CI=1, 1.48; P=0.046).
• Rural residents (40%) were significantly more likely to thaw frozen meat at room temperature than urban residents were (26.5%, P=0.003), while suburban residents (12.8%, P<0.001) were significantly less likely to thaw frozen meat at room temperature. As total annual household income level increased, the use of thermometers to determine the doneness of cooked meat increased (P<0.001). Suburban residents (21.8%) were significantly more likely to use a thermometer to determine when meat was cooked well enough to eat compared with urban residents (12.2%, P<0.001).

Cross-contamination

• As age increased, there was a decrease in the prevalence of using soap and water to clean the kitchen sink or cutting board after preparing raw meat (P<0.001). Urban residents (62.3%) were significantly less likely to use soap and water to clean the kitchen sink and cutting board after preparing raw meat than were other residents (71.2%, P<0.001). 
• Males were significantly more likely to not always wash raw fruits before eating them than were females (OR=1.34; 95% CI=1.1, 1.64; P=0.003). Children younger than 12 years (OR=1.74; 95% CI=1.12, 2.71; P=0.009) and respondents between 18 and 54 years of age (OR=1.57; 95% CI=1.16, 2.13; P=0.002) were significantly more likely to not always wash fruits before eating them than were respondents older than 64 years.
• As age increased, the prevalence of washing vegetables before consumption increased (P<0.001). As total annual household income level increased, the prevalence of washing fruits and vegetables before consumption decreased (P=0.001 and P=0.03, respectively). Suburban residents were significantly more likely to always wash fruits and vegetables before consumption then were urban residents (P<0.001 and P<0.001, respectively).

Refrigerator Temperatures

• Males were significantly more likely to know the correct refrigerator temperature (OR=1.85; 95% CI=1.53, 2.23; P<0.001) as well as personal refrigerator temperature (OR=1.6; 95% CI=1.21, 2.1; P<0.001 than were females. Adults aged 25 to 64 years were significantly more knowledgeable of the correct recommended refrigerator temperature than elderly individuals were (OR=1.43; 95% CI=1.08, 1.89; P=0.009). As age increased, knowledge of personal refrigerator temperature also increased (P=0.012).
• Respondents with a college diploma (25.5%) were significantly less likely to know the correct recommended refrigerator temperature than were respondents without a high school diploma (37.3%, P=0.001). However, both respondents with a high school diploma (22%) and a university degree (21.1%) were significantly more likely to know the temperature of their own refrigerator than were respondents without a high school diploma (14.1%; P=0.011 and P=0.008, respectively).
• Individuals who reported a total annual household income between Can$40,000 and Can$60,000 were significantly more likely to know the correct recommended refrigerator (51.5%) and personal refrigerator temperature (25.1%) than were individuals who reported a total annual household income of less than Can$20,000 (35.7%; P=0.003 and 12.1%; P=.012, respectively)
• Individuals who reported a total annual household income between Can$20,000 and Can$40,000 (22.7%) were also significantly more likely to know their own refrigerator temperatures than were individuals who reported a total annual household income of less then Can$20,000 (12.1%, P=0.041)
• Suburban resident were significantly more likely to know the correct recommended refrigerator temperature (54.9%) and personal refrigerator temperature (29.9%) than were urban residents (31.9%; P<0.001 and 15.9%; P<0.001, respectively)
• Respondents who knew the temperature inside their refrigerators were significantly less likely to eat chicken nuggets (OR=0.49; 95% CI=0.31, 0.75; P=0.001) than were those that did not know the temperature inside their refrigerators, but were significantly more likely to eat fresh basil (OR=1.55; 95% CI=1.03, 2.3; P=0.023).

• The survey assessed consumption of high-risk foods and evaluated consumer food safety knowledge and practice in the home, and the results suggest that certain high-risk food consumption patterns and unsafe food-handling and hygiene practices are fairly common among respondents and are often associated with demographic characteristics. Despite awareness of concepts such as hand-washing and high-risk food consumption, respondents still engaged in high-risk behaviors that varied according to demographic characteristics. These findings indicate that education messages should stress that consumer behavior is an essential component of the safety of foods prepared at home and additional research is needed regarding consumer perceptions of the nature and origin of food-borne illness in order to design educational programs that will motivate behavioral change.
• This study had also outlined the difference in views and knowledge between public health scientists and the public with respect to high-risk foods. Although high-risk foods highlighted in this study pose different risks, more work is needed through food attribution studies to better understand the role of particular food items and infectious disease. Once this information is available, educational campaigns could be better directed to make sure that the public is aware as much as possible of all high-risk foods that public health officials are concerned about and combine those messages with strategies to promote healthy diets with appropriate nutritional guidance.

Government:

Public Health Agency of Canada

Respondents were not well described, although many details were published elsewhere. The authors note the following limitations:

• Food safety questions on the preparation and handling of food were asked of the person most familiar with these practices in the household and are thus not representative of individuals in the general population
• Since consumer awareness of food safety issues reported by household food preparers is higher than that of the general population, it is possible the prevalence of positive food safety practices here were overestimated
• These results relied on self-reported practices and knowledge, which can lead to biases toward reporting socially desirable responses
• If respondents reported what they believed to be correct rather than what they practiced, the prevalence of unsafe practices among the general population may be even higher than reported here
• The results presented are not validated by observation. Since studies show discrepancies between self-reported practices and actual food-handling behaviors, it is possible that actual food safety practices are poorer than are the self-reported practices here.