Electrical power outages in Canada are mainly caused by extreme weather events such as wildfires, tornadoes, droughts, and flooding which can pose a threat to human health and life. Reference Dotto, Duchesne and Etkin1,Reference Casey, Fukurai and Hernández2 Several studies have anticipated that power outages could be stronger and more frequent due to climate change, high electricity demand, and aging infrastructure.Reference Bruce, Egener and Noble 3 –Reference Wang, Zhu and Huang 6 Aging infrastructure was one of the reasons for the 2003 power outage that affected the Northeast of Canada and the US.Reference Anderson, Santos and Haimes 7 Around 256 power plants experienced failures in both countries impacting approximately 50 million Americans and Canadians.Reference Minkel 8 The restoration time was around 4 days in the US and around a week in Canada with the cost of electrical repairs and restoration being between 4 to 10 billion USD.Reference Anderson, Santos and Haimes 7
Power outages can increase the risk of food- and water-borne illnesses, which often occur due to unsafe practices and exposures at home.Reference Byrd-Bredbenner, Berning, Martin-Biggers and Quick 9 –Reference Scott 11 During a power outage, refrigerated food can become warm enough to enable microbial growth and toxin formation, which if consumed can cause gastrointestinal and other food-borne illnesses. 12 –Reference Godwin, Coppings and Kosa 14 For example, after the 2003 power outage in North America, Klein et alReference Klein, Herzog and Smolinske 15 found that food contamination queries to poison control services in New York City increased by 54% and questions about water contamination increased by 183% compared to the previous year. Moreover, Marx et alReference Marx, Rodriguez and Greenko 16 concluded that more than 20 New York City emergency departments had an increase in diarrheal visits 2 days after the initial outage. Similarly, flooding can lead to an increase in gastroenteritis because of the consumption of foods contaminated with flood water. Flooding also damages sewage and water infrastructure which can contaminate drinking water sources with water-borne pathogens. Reference Andrade, O’Dwyer and O’Neill17,Reference Cann, Thomas and Salmon18 Food and water safety preparedness before, during, and after power outages and floods are necessary to reduce these risks. Reference Godwin, Coppings and Kosa14,Reference Dominianni, Ahmed and Johnson19– Reference Kosa, Cates and Godwin 21
The literature on individuals’ preparedness for food and water safety before, during, and after power outages is limited, especially in Canada.Reference Klinger, Landeg and Murray 20 Two studies by Kosa et al Reference Kosa, Cates and Ball13,Reference Kosa, Cates and Godwin21 in the US using web-enabled surveys and focus groups respectively revealed that multiple factors influence people’s preparedness level during a power outage (e.g., not thinking about preparing for power outages, low-risk perception of the dangers from power outages, uninformed about guidelines and being reluctant to follow them). There is a need to investigate these factors further in a Canadian context. Therefore, a qualitative study using focus groups was conducted to investigate the factors influencing Ontarians’ food safety preparedness during power outages at home. These findings can be used by public health and food safety agencies to improve risk communication with the public and reduce the likelihood of gastrointestinal illness during power outages.
Methodology
Study Design and Ethics
A qualitative descriptive study design was used for this research using focus groups to understand Ontarians’ preparedness during power outages. Qualitative descriptive studies use interpretive methods to understand individual experiences from a subjective perspective.Reference Doyle, McCabe and Keogh 22 This study was conducted by an interdisciplinary team with experience in qualitative research, food-borne diseases, water-borne diseases, and behavior change. This manuscript followed the Standards for Reporting Qualitative Research Guidelines (Supplementary File A).Reference O’Brien, Harris and Beckman 23 This study was reviewed and approved by the Toronto Metropolitan University Research Ethics Board (REB) with the File Number 2023-091. The approval was granted in accordance with the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (TCPS 2 – 2022) which is Canada’s national guidelines for research ethics. Documentation of REB approval is available upon request. Participants were selected based on responses to an initial screening questionnaire. Those who met the eligibility criteria were contacted via email. These individuals were provided with a link to an electronic consent form administered through Google Forms. The consent form was distributed in advance of the scheduled focus group to allow potential participants sufficient time to consider their involvement. The form clearly stated that participation was voluntary, that participants could withdraw from the study at any time without penalty or retribution, and that they could choose not to answer specific questions. It also outlined that data would be collected and analyzed upon commencement of the focus group, unless participants contacted the research team within 2 weeks post-session to request that their information be excluded. Anonymity and confidentiality were assured, and all data were stored securely. Written consent was obtained electronically through the Google Form prior to participation.
Study participants and recruitment
Focus groups were conducted with participants in 3 urban areas (Hamilton, Oshawa, and Toronto) and 3 rural areas (Elgin County, Huron County, and the United Counties of Prescott and Russell). These locations were chosen because they were known to have experienced power outages due to weather-related events in recent years. Participants were recruited through online advertisements and local library postings on each site. Recruitment was conducted by posting recruitment notices within Facebook groups specific to each community and by using local libraries within communities to distribute and post flyers on community bulletin boards. Additionally, participants who signed up and participated in focus groups were encouraged to pass along the recruitment materials to other potentially interested participants. Individuals had to meet the following inclusion criteria: be 18 years or older, be English-speaking, and live in one of the targeted study locations. Participants were also required to have a working internet connection for video conferencing through Zoom 5.17.11 (Zoom Video Connections Inc., San Jose, California, US).
Data collection
Participants were selected for the focus groups using the demographic information collected using a screening questionnaire. Participants’ age, gender, ethno-racial identity, education, place of residence, and household income were collected to ensure a heterogeneous sample within each group. To protect identities, pseudonyms were used in place of actual names for transcripts and throughout this article. Focus groups were conducted via Zoom in July 2023 using a semi-structured question guide (Supplementary File B). The guide was developed using questions from previous studies and government guidance documents for the public during power outages. Reference Kosa, Cates and Ball13,Reference Godwin, Coppings and Kosa14,Reference Kosa, Cates and Godwin21,24– 26 The focus group guide contained 16 questions focusing on power outages, flooding and information sources. VT served as the focus group moderator and ensured consistency across groups by following a standardized focus group guide. To prepare for this role, VT engaged in self-directed training by reviewing a focus group guidebook and relevant literature,Reference Stewart and Shamdasani 27 –Reference Krueger and Casey 30 and conducting training with AT and IY, both of whom have prior experience conducting focus groups. This preparation supported a consistent approach to question delivery, facilitation, and providing neutral responses to participant comments. Focus groups lasted an average of 98 minutes and ranged between 85-111 min. A $25 electronic gift card was provided to participants as an incentive to ensure sufficient participation and improve attendance reliability.Reference Krueger and Casey 30
Analysis
The analysis was informed and guided by the Theoretical Domains Framework (TDF).Reference Atkins, Francis and Islam 31 The analysis followed an abductive approach to coding. This means that coding consisted of a mix of inductive methods (i.e., researcher-generated codes based on the data) and deductive methods that included codes informed by the TDF. A total of 14 codes were applied to the data (Supplementary File C). After the coding process, relevant themes were generated based on 2 criteria. The first was the frequency of specific beliefs and the second was any indication of strong beliefs expressed by participants that affected their food or water safety preparedness during extended power outages. A belief was considered strong when participants overtly expressed their beliefs without hesitation or doubt. Moreover, a strong belief was also determined when participants agreed with each other or when one participant was able to add onto another participant’s explanation. A semantic approach was employed consistently across all focus groups for thematic analysis, whereby themes were generated based on explicit participant statements rather than on the underlying or inferred meaning of the data.Reference Braun and Clarke 32
Transcripts were reviewed multiple times by V.T. to ensure codes and themes aligned with participants’ explanations and emphasis on certain responses. Both the codes and themes were iteratively reviewed, discussed, and revised with I.Y. and A.T. for validation and to gain diverse perspectives. Two types of triangulations were used to increase collaboration and explore the credibility of the results.Reference Byrne 33 Researcher triangulation was performed by discussions and peer debriefing throughout the research process. Member checking was also completed, and respondents were asked to provide feedback on the overall findings. Analysis was conducted using NVivo 14.23.2 qualitative analysis software (QSR International, Doncaster, Australia).
Results
Forty-six participants participated in the focus group discussions, with approximately equal representation between men and women, and varied educational backgrounds (Table 1). Several factors contributed to participants’ behaviors and preparedness during an extended power outage. These factors were coded and integrated within 4 overarching themes discussed below.
Participants’ demographic characteristics (n = 46)
Theme 1: Information from Trusted Agencies, Lived Experiences, and an Expressed Desire to Change Behaviors during Power Outages
Most participants stated that they knew common actions and correct behaviors that should be followed during a power outage. Most of these behaviors that were mentioned were factually incorrect or not substantiated by the Government of Canada’s risk communication guidance. The most prevalent misconception was the duration of time perishable items would stay safe in the fridge after a power outage:
The moment it’s more than 24 hours then at that point that’s suspecting. Stop checking the smell and stuff like that to see if the food is good. Anything after 24 hours, I feel it’s, you know, it’s not safe. (George, 30-39 years old, rural area)
Another common incorrect claim was how participants checked if their foods had spoiled during a power outage:
Yeah, I could actually realize that when there is a change in texture, or should I say the smell actually is different from the original smell. It might actually be an odor. And then you would know that it’s not safe to eat and you need to dispose of it. (Kyle, 18-29 years old, urban area)
Only 2 participants within the same focus group stated the correct length of time perishable items were allowed to be in the fridge after a power outage, while most participants also followed the recommendations from public health to minimize the frequency of opening the fridge door:
Yeah, for me, I limit the time I open the fridge to preserve the coldness in it. I just limit the time I open it and then nobody goes to the fridge during that period. When you open it once you take out all you need and then just close it, so the temperature is preserved. (Fatima, 18-29 years old, urban area)
Most participants recounted behaviors and memories from previous power outages and trusted government organizations more than social media due to the fear of misinformation. These participants reported seeking information from organizations such as the Canadian Red Cross, Public Health Agency of Canada, and the Canadian Food Inspection Agency (CFIA).
Okay, I really think my major source of information regarding things like this, is, like the government-owned websites. Or, maybe, you know, organization websites. I don’t think they can give misleading information. So, I really go with them. (Aisha, 30-39 years old, rural area)
However, a few participants believed otherwise and trusted information from family and friends more than government sources:
I go to social media just to check on procedures that have to do with food safety and sometimes, my neighbor. She works with a food company and knows more about food preservation. She normally lectures me on that. (Livia, 18-29 years old, urban area)
Some wanted to change their own behaviors near the end of the focus group after they had heard from their fellow participants. They started to plan and set goals:
For me, I will say what stood out to me today was I’d like a solid plan for a power outage. So, it’s not something I really thought of before. (George, 30-39 years old, rural area).
Theme 2: Accessibility to Alternative Power Sources, Social Support, and Financial Resources Affect the Level of Preparedness
Most participants were worried that they did not have alternative power sources such as solar power or a generator in case of an extended power outage:
I don’t really think I have, you know, enough appliances to support myself in case of a power outage. But I really like…having a solar panel to…help myself out and I really think I’ll go into that, maybe make some plans within the month to get some things like that. (Jacob, 40-49 years old, rural area)
Financial resources were also salient while discussing the level of preparedness. Participants wanted to purchase an alternative power source but did not have the means to do so:
[Solar panel installation is] a bit pricey and…a lot of people barely have their basic needs met. (Carolyn, 30-39 years old, urban area)
Participants also stated that social support played a major role in their preparedness level for an extended power outage. Some participants stated that they did not have anyone to ask for help or information whereas others explained that they had family and friends to lean on during a power outage.
[The outage] was a bit of a problem because it lasted for an hour and at that time, we had no one to ask or rather to help… (Victor, 30-39 years old, rural area)
Theme 3: Factors Beyond One’s Control Including Geographical Location, Resource Distribution, and Population Density
When prompted about the difference between preparing for a power outage in an urban or rural area, most participants stated access and distribution to resources were limited in rural communities:
Personally, I believe that, for the open areas there is priority in resource distribution. In rural areas, they will not allow these resources. This will in turn affect our preparedness as it regards (sic) to our safety during an emergency. Period. (Harish, 40-49 years old, rural area)
However, some other participants stated that preparedness for power outages would be heightened in rural areas due to the inadequate infrastructure:
I think us being a rural area we ought to be very prepared for it. Yes, we have to experience more frequent outages and power emergencies compared to living in the city. We just have to be prepared. There is no option. (John, 18-29 years old, rural area)
Participants also commented on how population density affected their level of preparedness because they could anticipate the increased demand for resources in densely populated areas compared to rural regions:
If you live in a populated area, I guess I’ll be more prepared because whenever the power goes out, there will be more competition. I know there are more people here, so water and other stuff. The stores may be full of people trying to get stuff. Yeah, so I guess it makes you more prepared and more careful compared to when you’re living in a rural area where you can just go to the store and get it, and you wouldn’t meet many people. So, I guess yeah, it depends on the area. (George, 30-39 years old, rural area)
Theme 4: Wide Range of Perceived Risk, Personal Agency, Caregiver Roles, and Emotions
Participants had mixed responses when discussing their perceived risks of extended power outages. Some participants understood the dangers of power outages to themselves and others, and were willing to take action to prepare for an outage:
It’s good to always have a plan so I do that. I think that’s the most important thing that has been brought to my notice. Whether or not you are expecting something like that, you should have a plan. And I see how even more important it is for me since I have children, and I have to be responsible for other people. (Carolyn, 30-39 years old, urban area)
Other participants did not understand the risks of power outages and did not find it necessary to prepare for such an event:
I don’t have an emergency plan because it’s not of so much importance. It rarely happens and when it does it doesn’t last long. (Aylin, 18-29 years old, urban area)
Most participants did not have an emergency kit but had items such as candles and flashlights in case of an extended power outage:
It was really bad. So. I wasn’t prepared. I have learned a lot from the experience. I have a couple of flashlights at home now. Before I wasn’t prepared at all. (George, 30-39 years old, rural area)
Some participants also believed that there are no actions that can be taken to prepare for an extended power outage.
You can never be prepared even if you are being informed. Let’s say you’re being informed right? There is no way you’re gonna put everything together like, yes, maybe a power bank. That’s, that’s the easiest you can do. (Fiona, 40-49 years old, urban area)
Some participants were overconfident and claimed that they were as prepared as they could be whereas others did not have confidence in themselves and worried that they would not be prepared for such an event if it were to occur:
I don’t really think I’m prepared for, you know, 24 hours. If they have any kind of preparation information, then when it happens, I can get myself stuck up in the house. But I don’t think I can survive it in terms of an emergency. (Jacob, 40-49 years old, rural area)
Participants’ social roles and identities also influenced their behaviors and preparedness before, during, and after a power outage. Four participants who were also mothers expressed that preparing for power outages in a home with children was more complicated. These women also revealed that the safety and security of their children was important and influenced their preparedness during a power outage.
I think you can never be prepared enough to have a complete 24-hour power outage in your house especially when you have kids, and you have so many things powered with electricity. But I think the incident taught me a lot because I was totally unprepared, and I had to run to the mall to get things like torchlights and candles to be able to give to my household. (Fatima, 18-29 years old, urban area)
Participants’ emotions and feelings when experiencing a power outage also shaped their preparedness. Most participants who were uncomfortable and felt helpless experiencing past power outages were more prepared for future power outages:
Due to me going through such experience it has actually opened my mind to a lot. And I’ve actually equipped my home with some solar lights and battery lamps as well. And like in my basement and the living room I have a couple of lights that would actually last me, let’s just say for 2 days max. And, I do have… some meals that could actually be preservable until the light comes up again yeah. (Kyle, 18-29 years old, urban area)
Discussion
To the best of our knowledge, this is the first study conducted in Ontario, Canada about food and water safety preparedness using focus groups. The qualitative design allowed for the exploration of Ontarians’ opinions and beliefs about preparedness during power outages. One’s knowledge about the risks associated with power outages appeared to influence participants’ preparedness during power outages. Several studies have indicated that knowledge of the risks associated with disaster or emergency situations leads to greater perceived risk,Reference Kellens, Terpstra and De Maeyer 34 –Reference Ao, Zhang and Yang 37 which could lead to preparedness practices for a particular event. The focus groups found a lack of preparedness for power outages because most participants did not have an emergency kit. Participants mentioned that they have 1 or 2 items such as candles, flashlights, and/or canned foods. Participants lacked knowledge about food and water safety during power outages. For example, most participants believed foods in the refrigerator would still be safe to consume until 24 hours after a power outage. This belief is a misconception because both the Government of Canada and the US Food and Drug Administration strongly recommend that foods would be safe for only 4 hours after a power outage. 24 – 26 The finding also corresponds with Kosa et al. Reference Kosa, Cates and Ball13,Reference Kosa, Cates and Godwin21 who found that few US participants knew to discard perishable foods from the fridge after 4 hours without power.Reference Kosa, Cates and Godwin 21
Most participants also incorrectly claimed that smelling and tasting foods was the proper way to check if foods were safe to consume during or after a power outage. Kosa et alReference Kosa, Cates and Godwin 21 reported similar findings, where 20-65% of respondents smelled or tasted their foods to determine if they were still safe to consume. Similarly, Nesbitt et alReference Nesbitt, Majowicz and Finley 38 found that 64% of participants in a Canadian survey used visual inspection to determine that meat was ready to eat when cooked. Knowledge is a mechanism of action that could be prioritized to increase public knowledge and preparedness during power outages.Reference Carey, Connell and Johnston 39 Behavior change techniques such as instructions on how to prepare for a power outage might be useful to bring about change through this mechanism of action.Reference Carey, Connell and Johnston 39 The development of educational materials which target these misconceptions may be helpful to change people’s preparedness behaviors, such as assembling and providing emergency kits.
The majority of participants had a desire to seek information from trusted organizations such as the Red Cross or the CFIA. Given the knowledge and risk perception gaps identified in this study, more effective communication strategies and a greater reach of government guidance on this topic may be needed. The behavior change technique of informing citizens about the health consequences of their actions during power outages (i.e., the risk of developing gastrointestinal illness due to poor food handling) could be effective in increasing knowledge and perceived susceptibility to this issue.Reference Carey, Connell and Johnston 39 Studies conducted in the US also recommend presenting the public with the illness consequences of power outages using concrete, everyday terms instead of technical terms. For example, instead of terms like gastrointestinal illness, simple terms like vomiting and diarrhea could be used in future messaging.Reference Dransch, Rotzoll and Poser 40 – 44
Participants also stated that social support played a major role in their preparedness level for an extended power outage. Some participants stated that they did not have anyone to ask for help or information whereas others explained that they had family and friends to lean on during a power outage. Social rewards need to be in place to change behaviors influenced by social support.Reference Connell, Carey and De Bruin 45 A social reward is a positive reinforcement from other people that includes acceptance, recognition, and praise.Reference Michie, Richardson and Johnston 46 Therefore, positive feedback and encouraging messaging through social media and other platforms might help to increase social support on this issue.Reference Laycock and Caldwell 47 –Reference Rudolph-Cleff, Knodt and Schulze 50
Participants’ place of residence was a predominant environmental influence on food and water safety preparedness during power outages. Several studies have shown rural regions in North America are disproportionately affected by power outages and flooding depending on the population density and the distance to an urban city.Reference Andresen, Kurtz and Hondula 51 –Reference Moffatt and Struck 54 Similar findings by Prelog and MillerReference Prelog and Miller 55 found that smaller rural Texas communities were increasingly concerned about the distribution of resources within their community during a disaster. These resource distribution problems are not due to the disaster but are already existing constraints within the community,Reference Laycock and Caldwell 47 which partially explains why rural Ontarians have power interruptions that are double the amount of time as urban Ontarians. 56 Behavior change techniques of restructuring the built environment could be effective in bringing about change through this mechanism of action.Reference Carey, Connell and Johnston 39 In this case, prioritizing the improvement of existing electrical infrastructure in rural regions can provide upstream preventative approaches to combat power outages.Reference Kanduth, Clark and Dion 57 In addition, officials can engage rural communities in community resilience planning using participatory frameworks and tools to enhance preparedness. Reference Murphy, Anderson and Bowles58,Reference Amaratunga, Anderson and Bowles59
Study participants often expressed a desire to purchase another energy source such as solar power but were financially limited. Such alternative power sources would mitigate the impact of power outages in Ontario ensuring that more residents have backup power options during extended outage events. Barrington-Leigh and OuliarisReference Barrington-Leigh and Ouliaris 60 found solar photovoltaic (PV) systems were useful as alternative power sources, particularly in rural areas, but were expensive in Canada. Government-based incentives or rebates could support public motivation in this area and increase overall power outage preparedness.Reference Carey, Connell and Johnston 39 Such incentive programs exist in some municipalities, and there is an opportunity to expand this in Ontario. Studies in other jurisdictions have found that the addition of rebates increased the installation of residential solar panels.Reference Canfield, Egbue and Hale 61 –Reference Bagherzadegan 64
Limitations
This study employed non-random sampling methods, including snowball and purposive sampling, which are appropriate for qualitative research aimed at exploring diverse perspectives rather than producing generalizable results.Reference Khalid Ahmed 65 Therefore, the results of the study may not be generalizable to other provinces or countries. Some participants were encouraged to share the recruitment materials within their own networks. While this snowball approach helped broaden participation, it may also have introduced homophily, as participants are likely to refer individuals like themselves.Reference Khalid Ahmed 65 To mitigate this potential, participants were recruited using multiple methods to enhance sample diversity and recruitment was focused on geographically dispersed locations across the province. Race and gender data were collected and reported to demonstrate the heterogeneity of the sample, not for analyzing trends or patterns in opinions or behaviour by demographic category. The purpose of a heterogeneous sample was to include a range of perspectives which is in line with qualitative research standards, helping transferability and transparency.Reference Nowell, Norris and White 66 This multi-pronged strategy aimed to reach a broader and more varied participant pool beyond personal networks. Additionally, it was not possible to confirm the accuracy of participants’ statements and experiences (e.g., whether they had previously experienced a power outage).
A limitation of focus groups is social desirability bias in which participants may state ideas that would be socially acceptable to other participants.Reference Bergen and Labonté 67 However, this is not believed to be a predominant limitation in this study since the topic is not particularly controversial or polarizing in nature. Nonetheless, multiple steps were taken to reduce this potential bias, such as using pseudonyms instead of people’s real names. The moderator also ensured to provide neutral responses for any statements made during the focus group allowing participants to feel comfortable to state their opinions. The moderator also carefully monitored the discussions and intervened when appropriate to invite input from less vocal participants, ensuring that 1 or 2 individuals did not dominate the conversation. At the same time, space was allowed for natural interactions to unfold, facilitating the collection of rich, essential data.
Conclusion
This qualitative study used the TDF and iterative approaches to explore individuals’ food and water safety preparedness during power outages in Ontario, Canada. Some misconceptions and knowledge gaps were identified and require enhanced outreach and risk communication. Other factors were unmodifiable such as place of residence, resource distribution, and population density, and were mainly mentioned by rural residents. Participants had a strong desire to purchase alternative power sources but did not have the financial means to do so. As a longer-term measure, the electrical infrastructure should be improved with priority given to rural areas. Consideration should be given to incentive and rebate programs for rural communities to encourage alternative power options. Future research is recommended to investigate these issues among a representative sample of the population and to specifically investigate preparedness among higher-risk population groups (e.g., families with children and older adults).
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/dmp.2025.10155.
Author contribution
V.T. wrote the manuscript and was responsible for data collection and analysis. I.Y. formulated the focus group guide and provided conceptual and methodological oversight. A.T. provided guidance throughout the writing process and on the analysis. F.T. and G.O. provided methodological and conceptual guidance. All authors reviewed and edited the manuscript.
Funding statement
Vidursana Thiraviyarajah received partial salary support through I.Y.’s Ontario Early Researcher Award.
Competing interests
The author(s) declare none.
Open access
