Exploring perceived risk for COVID-19 and its role in protective behavior and COVID-19 vaccine hesitancy: a qualitative study after the first wave

Background The novel coronavirus pandemic (COVID-19) has had severe impacts on morbidity and mortality globally. Methods This study was set in rural central Kentucky and included participants recruited from public spaces. Fifteen qualitative interviews about personal experiences during the COVID-19 pandemic were conducted by phone from July 3 to July 24, 2020. Interviews were recorded, transcribed, and coded using a grounded theory approach. Results Participants who perceived COVID-19 to be a severe risk tended to have personal health concerns and therefore reported taking protective measures for themselves. A slightly smaller proportion of participants reported taking measures to protect others (particularly family). A minority of participants had an ambivalent attitude towards the risk and only took measures if required. COVID-19 vaccine acceptability was low with most participants expressing concerns regarding their need for a vaccine, safety of this vaccine, the value of personal rights, or future vaccine supply. Conclusions Most participants perceived some risk of COVID-19 and took steps to prevent infections in themselves and others. Mandates for mask use in certain locations were additionally useful for those who had an ambivalent attitude towards the risk of illness. There was surprisingly little connection between perceiving COVID-19 risk and a desire for the COVID-19 vaccine. In this setting, vaccine acceptability was low, with vaccine concerns outweighing perceived potential benefits. In conclusion, because the risk was often constructed in terms of worries for themselves and others, the framing of health education materials for protective behaviors in these terms may be effective. Furthermore, future COVID-19 vaccine education should address vaccine knowledge and concerns, such as the need for a vaccine and its safety, and emphasize how a vaccination would reduce their chances of severe disease if they were to get sick.

Patterson et al. BMC Public Health (2022) 22:503 behaviors, quarantine measures, widespread restrictions including mask-wearing (formerly), and vaccination. These strategies have been promoted by public health organizations and governments alike [5]. While protective behaviors are being promoted, exploring the nuance of how individuals perceive the risk posed by COVID-19 is critical. According to the Health Belief Model, protective behaviors are driven by attitudes and perceptions, including the perception of risk associated with enacting or not enacting the behavior in question [6]: more protective measures tend to be taken when individuals perceive greater risk and vice versa [7].
COVID-19 vaccine development began in January 2020 and advanced rapidly [8]. As of April 2020, there were 115 potential vaccine candidates in several development and testing stages [8]. While facing the reality of extensive protective restrictions and limited therapeutics for reducing morbidity and mortality of the COVID-19 pandemic, the use of protective measures and broad vaccine utilization will be necessary to mitigate the spread of the pandemic [5,9].
Mitigation of COVID-19 risks requires that COVID-19 vaccines not only be available but also widely accepted. Yet, when it comes to vaccine acceptability, it is important to understand components of vaccine hesitancy. Vaccine hesitancy describes a spectrum of beliefs and attitudes towards vaccines resulting in delayed vaccination uptake and outright refusal [10]. Vaccine hesitancy is complex and is often very specific to time, vaccine, and risk groups [11]. This suggests that barriers to vaccine uptake can be different in any given situation [12,13]. Exploring how vaccine acceptability or hesitation is shaped is critical to informing a national COVID-19 response.
Widespread vaccine hesitancy can limit vaccine uptake and effectiveness. Considering the declining confidence in Measles, Mumps, and Rubella (MMR) vaccines and subsequent outbreaks across the United States and Europe, in 2019 vaccine hesitancy was declared by the World Health Organization (WHO) as a top threat to global health [14][15][16] that may lead to limited vaccine coverage and ongoing SARS-COV-2 viral transmission. On June 28, 2020, Dr. Anthony Fauci, the director of the National Institute of Allergy and Infectious Diseases, spoke of the current state of vaccine hesitation in the United States, stating that if many citizens refuse a COVID-19 vaccine, attaining widespread individual immunity leading to herd immunity is unlikely [17]. In the time since Dr. Fauci's statements, the ongoing emergence of variants of concern and new data demonstrating that prior infection and vaccination may not be sufficient to prevent new infections with SARS CoV-2 in all cases makes herd immunity a much further target than previously hoped [18]. Nonetheless, the marked reduction in severity of disease in vaccinated persons and data showing that booster doses are associated with reduced risk of infection, including reducing hospitalizations leading to an overwhelmed health care system underscore the critical importance of COVID-19 vaccination [18][19][20]. As of May 2021, about 60% of Americans had received at least one dose of a COVID-19 vaccine and as of February 5, 2022 this number has increased to 76% [21,22]. However, as vaccine rates seemed to peak in mid-April 2021, vaccine uptake seems to have slowed due partly to the manifestation of vaccine hesitation [23].
Meanwhile, the effects of COVID-19 on patients with pre-existing chronic health conditions exacerbate the true risk within populations [24,25]. For example, populations in rural areas, such as rural Kentucky, tend to have high rates of obesity, diabetes, and hypertension compared to other regions of the U.S. [25,26]. Patients with hypertension or diabetes are more likely to experience severe cases of COVID-19 [27]. At the same time, those with body mass index measures (BMI) greater than 25 were nearly four times as likely to die and eight times as likely to need advanced respiratory support as a result of their COVID-19 infection [28,29].
Regarding Kentucky specifically, vaccination rates declined in March 2021, and by February 5, 2022, 6 4% had received one dose, 55% are fully vaccinated, 23% fully vaccinated with a booster/additional dose of a COVID-19 vaccine [31,32,37]. This difference between rural and urban areas was about 14% between December 2020 and August 2021 [38]. Vaccine acceptability can be thought of as a combination of the perceived risk and perceived severity of the infection, with the perceived benefit of vaccination [7]. For this reason, to prevent morbidity and mortality associated with COVID-19, we examined how COVID-19 risk perceptions are constructed in order to better address the uptake of protective behaviors and COVID-19 vaccine acceptability among rural residents of Kentucky.

Recruitment and study design
Participants were recruited by one of the authors with strong ties to the community. Recruitment took place throughout July 2020 in high-traffic public places in rural central Kentucky, including the parking lots of local retail establishments. Potential participants were asked if they would be interested in participating in a study about COVID-19 and if they were between the ages of 18 and 65. Due to the various infection control measures in place at the time of recruitment, a convenience sample was recruited in person, but interviews were conducted via telephone.
Using this process, 93 people were approached, and of those, 39 people agreed to participate and provided their contact information. In keeping with COVID-19 social distancing safety measures in place at the time, a study team member then reached out to conduct the interview by phone. Fifteen people ultimately were interviewed and were included in the study. Two study team members (NP and TM) conducted semi-structured phone interviews using an interview guide to ensure that three main themes were addressed: COVID-19 risk perceptions, beliefs and attitudes towards protective or preventative measures, and beliefs and attitudes towards future COVID-19 vaccines. Drawing from the Health Belief Model, questions within the COVID-19 risk perceptions theme sought to explore factors affecting individual perceptions of susceptibility to and severity of COVID-19, while questions within the other themes explored perceptions of benefits and barriers to each behavior, including notions of perceived safety and efficacy [6]. Interviews lasted between 30 and 45 min and provided sufficient data to approach redundancy and saturation. The recorded personal contact information (phone numbers and addresses for compensation) was disposed of after three failed attempts to contact the respondent or the completion of the interview and delivery of their incentive in accordance with the approved protocol.

Analysis
In-depth interviews were recorded and transcribed verbatim using pseudonyms. Each transcript was coded using inductive thematic content analysis utilizing the analysis software Dedoose [39][40][41]. Initially, three transcripts were double coded by the two team members who conducted the interviews, creating multiple initial code lists. To develop a working codebook, the coders then discussed and compared these initial lists to identify redundancies, categories, and subcategories of emergent themes. Discrepant codes were discussed and resolved by consensus, and decision trials were noted. This codebook was then applied to each successive transcript by one evaluator, with modifications to the codebook retroactively applied to prior transcripts. Main themes and categories are reported here, with relevant participant excerpts presented in-text.

Ethical considerations
Informed consent was given before each interview, and participants were compensated for their time. This study was conducted in accordance with the Belmont Principle guidelines for human-subjects research. The study design was approved by the Tulane University Institutional Review Board (IRB) (protocol 2020-738).

Results
Fifteen interviews were conducted in total. Ten participants were women, and five were men. The average age of the participants was 43, ranging from 27 to 65 years old. All participants were residents of rural Kentucky. Participants were employed by local businesses, retired, and some (n = 3) were unemployed at the time of the interview. The themes, categories, and subcategories that emerged from the data are provided here, where "I" indicates the interviewer, "P" the participant, and the information in parenthesis indicates presumed gender, age, occupation.
At the time of data collection, the COVID-19 morbidity and mortality in central rural Kentucky were low. Three of the 15 participants knew of someone personally who had been sick. One participant described it like this: "I-Why do you think that those people who aren't following recommendations are making that choice? P-Um, because it's not hitting close enough home. And like some of these big cities, the death rate and what they have heard, they might see 3 or 4 or 10 cases in these little small areas. Um, you know, compared to [bigger county] that's got like hundreds a day. " (W, 33, Nurse).
The health-related toll contrasted with the microeconomic toll where participants mentioned job losses for themselves and others that they know personally. Where jobs were kept, the workload was altered, sometimes significantly. These contextual specifics may have led to differences in the perceptions of risk and the acceptability of a future vaccine for COVID-19 between localities. Regarding perceived risk, the determination of risk seemed to be dictated by an individual's close social networks, illustrated by the quotes below.

Perceived risk associated with COVID-19
All of the study participants (n = 15) were aware of COVID-19 and acknowledged that COVID-19 has been causing sickness and death. It was a common understanding that many people had been affected, many have died, and there have been profound economic impacts. One participant described this:

" I-What does the pandemic mean to you? […]
P-It makes me like feel bad because I don't like   In general, based on the interview context, there was increased concern about COVID-19 over time. From when COVID-19 first emerged, to the time of the interview, perceptions seemed to evolve, though it was often difficult for participants to pinpoint when their perceptions changed. When the threat was perceived as severe, the recommended protective behaviors were also seen as beneficial. The temporality of risk was described as follows by one participant:

Levels of concern
In general, there were three main types of risk perceptions found: personal health concerns leading to the utilization of protective behaviors, interpersonal concern leading to protective behaviors, and little or no concern resulting in only implementing the mandated measures.

Personal level concerns
The utilization of protective behaviors for personal protection was common. Most participants (n = 10) took these measures with their health in mind. In this case, taking the time to do additional handwashing and cleanings, wearing masks, taking social distancing measures, and any other precautions to keep oneself safe was deemed to be the obvious choice. Interviewees who had personal concerns also indicated that taking these measures is now a part of normal life.
For example, this reasoning was described by some as follows:

Interpersonal level concerns
For some (n = 8), the use of protective behavior and materials was motivated by interpersonal concerns. People reported wearing a mask, increased cleaning, and adhering to social distancing measures to keep someone else from getting sick, particularly children and older adults. This was described as follows:

Minimal to no concern
Other participants did not seem to perceive much of a risk at all (n = 5). They may have used protective measures, but only because of local government, workplace, or store requirements. This was often marked with an ambivalent attitude towards the risk of COVID-19 in terms of morbidity and mortality that was generally based on religious beliefs. Furthermore, the perception of risk was at times influenced by the individual's opinion of government or the media, if either positive or negative (n = 8). Those who had a negative attitude towards the government or media tended to take minimal precautions. On the other hand, when perceptions about government or media were more positive, more protective measures were taken in general. However, most participants (n = 12) did not indicate strong feelings about the government or media to influence decision making. As opposed to the following sentiment.

Vaccine acceptability
While most participants perceived some risk associated with a COVID-19 infection and there were some worries that the pandemic would never end, the use of a COVID-19 vaccine was met with a high level of hesitation. This hesitation centered on four main themes: vaccine knowledge, an individual perceived need for a COVID-19 vaccine, vaccine safety, and infringement on personal liberties. These themes along with other COVID-19 vaccine acceptability findings are discussed here.

Vaccine knowledge
Several participants had difficulty understanding the role vaccines would play in preventing COVID-19, frequently confusing vaccines with some form of treatment or cure instead of prevention (n = 6). This was expressed in the context of asking questions about vaccines, revealing some misunderstandings that required the interviewers to explain the preventative nature of vaccinations.

Perceived vaccine needs
Almost half of the participants (n = 7) did not see the need for a vaccine to prevent COVID-19 for resuming a normal life. They indicated they would not need a reason for vaccine refusal, and the integrity of the information they have received was in question. This was discussed as follows:

"I-What would you have to hear in order for you to feel better about getting the vaccine? […] P-[sigh] Uh. I don't know if I would trust in it. I guess it has to come, an actual, personal who's who's involved with it. […] I mean I've been reading articles on Facebook and certain media, and even putting on CNN said that. It's hard to believe any of them anymore because anything could be tote-up and have a label on it.
I-Yeah.

P-So it would really take, be up to a doctor to truly know [phone ringing in background] to get the right source. " (W, 33, Nurse).
While the belief that the vaccine was unneeded was common, a minority (n = 4) of participants were enthusiastic about the opportunity to be vaccinated. Among these individuals, however, was a concern about the equitable delivery of a vaccine and concerns about possible future vaccine supply (n = 4).

Perceived vaccine safety
It was common that safety worries were perceived as barriers to vaccination and a driving force for vaccine hesitation. At times safety concerns were voiced generally. One participant referenced it like this: In addition to general safety concerns, because the vaccine would be newly developed, the worries about side effects (n = 9) and vaccine components (n = 7) seemed to be more pronounced than they would be otherwise. Many participants (n = 9) pointed to unknown side effects as a major concern.
"I-So on a scale, of let's say 1 to 10 again, with 1 being kind of not at all, 10

Perceived Infringement on Personal Liberties
Vaccine concerns in general and concerns that potential vaccine mandates may infringe upon personal rights made a vaccine for COVID-19 undesirable. Mandates were occasionally spoken about positively. For example, the quote below suggests that only a mandate could prompt the participant to be vaccinated. However, about half of the participants (n = 7) were against vaccine mandates for COVID-19, as well as other vaccines more generally, implying that a vaccine mandate would take away liberties in a way that shouldn't be allowed in the United States. Beyond personal rights potentially being infringed upon, it was also sometimes implied that desiring a mandated vaccine would be foolish.

Discussion
The perception of risk and associated protective behavior is likely closely tied with rurality, where risk perception and protective behaviors likely differ between rural and urban communities [42]. The connection between the perceived risk of COVID-19 and COVID-19 vaccine acceptability is not overly apparent here, contrary to some similar studies [43][44][45][46][47][48][49]. When it comes to the perceptions of risk in the context of the COVID-19 pandemic, the participants universally understood that they might get sick. When considering the temporality of perceived risk, there was inconsistent seriousness paid to the necessity of protection. However, after months of news stories, government regulations, and widespread economic impacts, the community members in this study were aware of protective measures that are recommended to be taken. This change in perception of the seriousness of the pandemic was often intangible when asked if thoughts have changed over time, yet at times marked when the local government or common storefronts ordered mandates for mask use.
The evaluation of the emerging data revealed three main groups of participants in terms of perceived risk: those with personal health concerns, those with interpersonal health concerns, and those with little to no concern. Protective behaviors were most commonly used for personal protection. According to Dryhurst et al., often, these measures were taken for oneself as well as when an individual's actions could affect a loved one [50]. Focusing on the protection of others is an important aspect of health promotion efforts and the way health warnings were communicated both locally and globally [51,52]. The emphasis that this behavior is not only for your health but also for protecting others seems to be effective when considering the uptake of protective behavior.
When the participants perceived the risk of contracting COVID-19 to be small or inconsequential, the participants tended to view getting sick as just a part of life. Those individuals tended to have little desire to change how they do things to prevent negative outcomes and only took precautions because they were mandated. It has similarly been shown that when the risk of illness seemed to be small, protective behaviors tended not to be taken [53]. In this participant group -perceiving little or no concern for COVID-19 --health messaging in terms of personal or interpersonal impacts seemed to be unmoving. This may give credence to the idea that mandates can induce behavior change during a pandemic when other forms of communication did not [54].
When it comes to the COVID-19 vaccine and related mandates, the themes that emerged included a lack of perceived need, safety concerns, and concerns over personal liberties. As a reason for vaccine hesitation, the perceived need seemed to at least partially point to a lack of understanding of a vaccine's purpose. Understanding the intended benefits of vaccination tended to be associated with increased vaccine acceptance [55,56]. However, acknowledging a lack of need seemed to be a strong motivator that may not be influenced by vaccine promotion efforts, as these participants seemed unwilling to vaccinate in any circumstance.
In addition to a low perceived need, many who might consider vaccination indicated they would prefer to wait and see if it is safe first. They indicated that they would likely wait until well after the vaccine was licensed and available to take it unless it was required to do so by an employer. This safety concern encompassed potentially dangerous vaccine components and unknown side effects. Many of these concerns seemed rooted in the fact that a COVID-19 vaccine would be new, though some mentioned safety concerns regarding other vaccines as well, regardless of when they were developed. Concerns based on safety, composition, novelty, possible side effects, and the desire to delay vaccination are common tropes in vaccine hesitancy literature. In fact, many of these same concerns were discussed in a 2021 review of factors contributing to vaccine hesitancy during recent pandemics, epidemics, and global outbreaks including the 2009 H1N1 pandemic [49,57]. Machingaidze et al., in particular suggest that worries surrounding vaccine hesitation globally are complex with several contributing factors [49]. Concern over vaccine safety, vaccine need, and logistical distribution worries indicated here align closely with concerns relating to vaccine hesitancy in low-and middle-income countries [49]. Sociodemographic factors such as political affiliation and location of residence also contribute to high levels of vaccine hesitation in some areas [36].
Nonetheless, mandated vaccination for themselves or their children was often described as unwanted. This is a common occurrence and not specific to a COVID-19 vaccine [13,58]. Considering how often concerns of future safety were voiced, the possibility that a mandate may force the use of a vaccine that the community had concerns about could lead to general mistrust [59].
The danger of COVID-19 was often evident, and most participants in varying degrees perceived the risk it posed to this population. This, however, did not seem to reliably translate to the desire for a new COVID-19 vaccine. In this instance, vaccine concerns appeared to outweigh the potential benefits of preventing sickness for most participants. There was general unease towards a potential vaccine, and only a minority would be enthusiastic in receiving one regardless of potential concern. This is consistent with the current state of vaccine hesitation, where the outlook is variable, and hesitation may or may not prevent eventual vaccination administration [11,12,60]. While desire for a vaccine usually comes along with recognizing the risk associated with future illness, for most participants in this study, the risk of illness as a result of COVID-19 infection did not outweigh the risk they associated with a future COVID-19 vaccine [13].
While perceiving greater risk or seriousness of a COVID-19 infection seemed to be closely tied to enacting protective measures that are less invasive than vaccines among this population, the relationship between perceived risk and acceptability of a future COVID-19 vaccine did not seem to be as strong, suggesting that messaging around the risk associated with a COVID-19 infection may resonate more strongly within this community when targeting preventative behaviors other than vaccination. Low levels of vaccine literacy, concerns around perceived need, safety, and personal liberties underpin why a COVID-19 vaccine may encounter hesitant beliefs and suggest that even in the context of a COVID-19 pandemic, public health officials seeking to promote COVID-19 vaccination should design messaging targeting healthcare provider recommendations, vaccine literacy and more generalized mistrust to address the low perceived susceptibility, safety, and concerns about personal freedoms.
This study adds to a large body of literature exploring COVID-19 risk perception and provides much needed detail about the relationship between those perceptions and protective behaviors, including vaccination in rural areas of the United States [35,61,62]. While the results contribute to a growing body of COVID-19 literature outlining factors that influence vaccine hesitancy, results suggesting greater acceptance of other preventative measures among this rural population highlight the differential way risk perceptions shape intention for different protective behaviors, an area of study that is likely to become more important as we think about future pandemics, pandemic preparedness, and the challenge of health communication during infectious disease outbreaks [63]. Considering the elevated awareness of infection control during the pandemic, this study may be poised to contribute to effective health communication specific to vaccination and other protective behavior [56].
This study has several strengths and limitations. First, as with all studies that rely on qualitative data, social desirability bias may have shaped the way participants responded during the interviews. However, the fact that interviews were conducted by phone instead of in-person, face-face, thereby allowing participants to maintain some distance and anonymity from the interviews, likely made it easier for respondents to respond. Secondly, the generalizability of this study is particularly limited. Not only was the sample chosen based upon convenience, which could have led to the introduction of selection bias, but while the study was open to individuals 18-65 years old, our sample skewed older. It didn't include populations 70 and older who may be more vulnerable to severe cases of COVID-19. Additionally, participant refusal rates were high resulting in a small sample size. Thus, efforts to extrapolate these findings to other rural communities and age groups should be made cautiously despite approaching saturation within this data. Furthermore, as the pandemic continues to evolve, viewpoints on protective behaviors and vaccine acceptability may have evolved since data was collected. Nonetheless, this approach did provide us with rich details about COVID-19 risk perceptions and their relationship to several preventative behaviors or vaccination intentions, using participants' own words, on a topic that would have been harder to assess in a quantitative study.

Conclusions
This study has found that in one rural area of the US, perceived risk is expressed in terms of worries for themselves, others, or an ambivalent attitude towards possible outcomes. Increased recommendations by trusted sources or framing interventions in terms of protecting themselves or others may be useful for promoting non-invasive protective behaviors such as wearing masks, social distancing, and increased cleaning. These perceived risks do not seem to correspond with a willingness for a COVID-19 vaccine. As vaccination is necessary to slow spread and reduce hospitalizations that overwhelm the health care system, promoting vaccine acceptability is of critical concern. This vaccine promotion effort may be positively impacted if vaccine safety was addressed, trusted public sources drove the messages, and messages emphasized how vaccinations help prevent the spread of COVID-19 to others and prevent severe disease.
Furthermore, as the pandemic continues to spread and new variants cause additional concern, well-known or personal experiences of sickness may influence individuals to become more accepting of protective behaviors or preventative vaccines. Exploring risk perception and vaccine acceptability using qualitative methods may be useful in guiding communication or policy during future pandemic events. However, as the interviews were conducted before a COVID-19 vaccine was available, the gap between behavioral intention and behavior suggests a need for longitudinal studies which take into consideration additional factors that may influence uptake in other rural [64].