Influenza vaccination hesitancy in large urban centers in South America. Qualitative analysis of confidence, complacency and convenience across risk groups

Influenza vaccination coverage in countries of Latin America is low among priority risk groups, ranging from 5 to 75% among older people. This paper aims to describe and analyze the determinants of influenza vaccination hesitancy through the lens of the 3C model of confidence, complacency and convenience among middle-class, urban risk group populations in Brazil, Chile, Paraguay, Peru, Uruguay, countries in South America with contrasting vaccination coverage. Focus groups were conducted among four risk groups: pregnant women, mothers of children aged <6 years, adults with risk factors, and adults aged ≥60 years in samples of urban residents. Adults with risk factors expressed the most detailed perceptions about confidence in the vaccine. A wide range of perceptions regarding complacency were expressed across risk groups and countries, with pregnant women and mothers showing greater concerns while convenience had a narrower and generally more positive range of perceptions. Participants from Chile and Paraguay expressed the most contrasts regarding confidence and complacency. Information and communication strategies need to be tailored for risk groups while confidence and complacency should be addressed in synergy.


Introduction
Influenza is a respiratory viral disease whose symptoms can be confused with those of a common cold, easily escalating to a more serious condition and even leading to death. Influenza cases are estimated at around 1 billion per year worldwide, of which 3 to 5 million are severe. Modelling studies suggest between 291 243 and 645 832 influenza deaths occur annually [1]. assigned to health as against other responsibilities and may stem paradoxically from a lowering of risk perception due to immunization program success. Self-efficacy-the self-perceived or real ability of an individual to take action to vaccinate-also influences the degree to which complacency determines hesitancy. Convenience of the vaccine is defined by availability, affordability, willingness-to-pay, geographical accessibility, ability to understand and accept vaccine-related information (language, culture and health literacy), the appeal of immunization services, and quality of care [13].
This paper aims to describe and analyze the determinants of influenza vaccination hesitancy through the lens of the 3C model across urban risk groups in five Latin American countries with contrasting coverage levels, selecting Brazil and Chile as high coverage countries and Paraguay, Peru and Uruguay as low coverage countries. Countries with different coverage rates were selected to identify the fullest possible range of expressions of confidence, complacency and convenience and to discuss their implications for vaccine uptake.

Methods
The protocol was reviewed by authorized ethics committees within each of the study countries, as follows: Brazil, Comissão Nacional de Ética em Pesquisa, 05215918. 6 The study design is multi-center and qualitative, contrasting knowledge, attitudes and practices through focus group technique with homogeneous members of four risk groups. Focus group methodology was chosen over other techniques to address the qualitative aspects of vaccine hesitancy given its capacity to obtain a greater amount of information in a shorter time elicited both through directed queues and the interaction between participants [14]. Furthermore, focus groups would enable us to obtain a first repertoire of perceptions on the basis of which to later address the hesitancy through more structured instruments. Interaction between persons sharing risk group characteristics would elicit the widest possible range of views based on knowledge, beliefs and practices regarding influenza vaccine confidence, complacency and convenience.
Focus groups were evenly distributed in one or two large cities within each country as follows Brazil: Porto Alegre (in the state of Rio Grande do Sul); Chile: Santiago and Valparaíso; Perú: Arequipa and Lima; Paraguay: Asunción and Ciudad del Este, and Uruguay: Montevideo and Salto (Table 1). Participants across the four risk groups lived in lower-middle class neighborhoods. A representative survey of health facility clients in the same neighborhoods undertaken as a part of the same research project to explore quantitative determinants of vaccine hesitancy showed that between 26.9 and 38.8% of elderly adults and adults with risk factors, respectively, had up to primary education, as against pregnant women and mothers of children, among whom only between 8.1 and 9.3% had such lower education levels. Participants from Paraguay and Uruguay tended to have lower education levels across risk groups when compared to participants from the other countries [15].
Focus groups were piloted and initiated in Peru in October 2018 and followed up in the four other countries between March and July of 2019, as soon as ethics committee authorizations allowed, and aiming to hold them as close as possible to influenza immunization campaigns. Table 1 describes the focus group recruitment strategy and characteristics. Individuals with homogeneous characteristics were recruited based on having the attributes of each of the following risk-groups: adults above 60 years of age (OA), adults with risk factors (ARF), pregnant women (PW) and mothers of children under 6 years of age (MC). Focus groups had between 6 and 10 participants with an average of 8, recruited upon leaving health facilities in the case of Brazil, Paraguay and Peru. In Chile, participants were recruited based on a sample frame developed for research purposes and registering risk group attributes and socioeconomic characteristics. In Uruguay, recruitment was based on a Facebook call, filtering for risk group and health service use characteristics. Heterogeneity within focus groups was sought with respect to the vaccination decision (accepted and declined) in order to understand indepth, the reasons and influences in both cases, as well as the obstacles and facilitators that led to their choice. Focus groups were undertaken in Spanish or Portuguese for the case of Brazil. All participants spoke Spanish or Portuguese and no translation was required. Focus groups were moderated in all cases by an expert focus group researcher, supported by an observer. Focus groups were held in meeting rooms or open spaces within or next to health facilities in Brazil, Paraguay (one focus group) and Peru, and in research meeting rooms or a hotel meeting room in Chile, Paraguay (three focus groups) and Uruguay. A discussion guide was developed and organized into knowledge, attitudes and practices sections including each the discussion of vaccine confidence, influenza complacency and vaccine convenience. (The guide is available as a S1 File). Within each section. Focus group sessions were audio-recorded after verbal informed consent was obtained from participants and information was transcribed verbatim. A first code list was developed by two of the authors (BP and MAGB) based on the theoretical framework and the pilot study focus group data from Peru. A coding manual was provided to researchers within countries (YC, PC, AFL, LB, DRK, BR and MR). Codes were inductively refined and detailed by country-level researchers based on focus group data. Detailed codes were reviewed by BP to arrive to a final set of codes for each country. Country data codification was reviewed by BP, DRK, AFL and disagreements involving the allocation of statement to the categories of confidence or complacency were resolved by consensus. Information was encoded with the support of the Atlas-ti v.8 software. Qualitative content analysis was carried out on thematic units of analysis based on the 3 Cs model categories by each of the country researchers. The main focus of analysis were risk groups, followed by cross-risk group analysis. Country case studies were produced by countrylevel researchers and reviewed by BP and MAGB. Final interpretation was based on comparing country case studies.

Results
Tables 2-4 present an overview of the range and contrast in perceptions regarding confidence, complacency and convenience by risk group.

Confidence: Effectiveness of the vaccine
The group of older adults and, to an extent, adults with risk factors showed a wide range of positive evaluations on the effectiveness of vaccines, including the recognition of the importance that vaccines be effective and a more nuanced perception of vaccine potential and limitations. As shown in Table 2, perceptions range from a clear sense of vaccine effectiveness, including limitations, lack of reliability, and a perception of trust or distrust in the context of negative experiences:  Mothers more than other groups expressed the need for specific information on vaccine effectiveness to enhance their trust:

Confidence: Safety
All groups expressed, paradoxically, that the vaccine is safe and that it has risks of its own. Except pregnant women, informants of all risk groups expressed concerns that the vaccine may be ineffective and cause adverse reactions: Aspects such as fear of needles, the adduced presence of mercury and its relations to autism, and ignorance on side effects the vaccine can have on babies contribute to raising doubts about safety. One mother from Chile with experience of pre-term births expressed having to be particularly careful in case contaminated vaccines cause autism: In the last resort, I prefer that [vaccines] be mercury-free. . . because I know of cases of people with autism who associate it with that.
Mothers and pregnant women also expressed concern with potential side effects from the vaccine:

Confidence: Trust in the system
Confidence in the information offered by service providers varies across risk groups. Confusion with the wide range in vaccine nomenclature i.e. "H1N1", "influenza", "swine flu" was widely shared. All groups also expressed the need to have more, and more specific, information on influenza and on the vaccine. Adults with risk factors expressed the most nuanced notions of trust (or distrust) with the health system, both as a whole and with vaccination services and campaigns specifically. The fact that influenza vaccination schedule needs to be updated every year engenders mistrust regarding the vaccine's effectiveness.

My boss works in health and says that every year the strains change, yet when the vaccine comes here it is [with] the old strain.
(OA, Uruguay).
Trust in public or government health services varied across countries, but in all groups, there was trust in general. Public services are seen as more reliable because they have historical expertise and know-how, besides having greater resources than the private sector: I've heard that the vaccine cold chain works better in the public health services than. . . in the private services.

(ARF, Paraguay)
Some participants expressed that the vaccine in private health services might be of a better quality than in their public counterparts, while also having shorter waiting times.
Health providers were by far the most common source of trusted information on vaccination across risk groups.
When the doctors explain things to me, I understand them better, and they told me: get your shot for the flu, or else it will be very bad. And so that was it, I went and got the vaccine. Really, that was the main reason I got the vaccine.
There are doctors who. . . tell you where you should go to get a vaccine-and they follow up on you, they ask you if you've got your shot. This is why I believe this is the best vaccine, I feel confident, I am sure that vaccines are good for everybody.

(ARF, Paraguay)
Despite being a trusted source of information, medical advice regarding the vaccine is not always followed through. Reasons for not getting the vaccine in these cases are not due to a lack of confidence and may be associated with factors regarding convenience and complacency.
Another concern expressed with trust in the system was that vaccination program priorities could be driven by pharmaceutical industry interests. In an extreme, some participants suspected the pharmaceutical industry had financial gains with the process of updating influenza strains. The commercial interest related to the vaccination programs raises doubts the credibility of vaccination program guidelines and of public health professionals' advice: It is a business of the laboratories that treat us like guinea pigs.
Comparing the perception of confidence across countries, more often questioning of vaccine effectiveness was expressed in Paraguay than in other countries across risk groups. In this country participants reported commonly the use of traditional medicine and herbs in the prevention of influenza and for the treatment of disease symptoms, a qualitative association that suggests that trust in traditional medicine could rival trust in the vaccine Participants from Brazil, Paraguay, Peru and Uruguay questioned vaccine effectiveness only in the case of some risk groups, especially mothers of young children.
Participants from all risk groups in Chile expressed positive views about vaccine safety. In Paraguay, by contrast, the idea that the vaccine can cause the disease was widespread. In Peru and Brazil, mistrust concerning the safety of the vaccine occurs more among adults with comorbidities, while in Brazil questions arose about the safety of the vaccine during pregnancy. Greatest distrust in the health system was expressed in Paraguay, especially regarding the view that the vaccine is, in general, offered late. Related to this, participants in this country expressed more trust in traditional medicine and in pharmacies than participants from other countries. In Uruguay and Brazil, participants in some risk groups, particularly adults with comorbidities and pregnant women, say that doctors have different positions concerning vaccine recommendation, raising doubts about the safety of the vaccine.

Complacency: Influenza risk and severity
Influenza was perceived by most risk groups as a serious illness, aggravated by the fact that there are no pharmaceutical cures and that it may be life-threatening (Table 3).

[Influenza is of] high risk, a person can die. My niece got sick last year and badly, [she had] seizures, high fever. [She was given] very expensive antibiotics and was hospitalized. (ARF, Chile).
And curing the flu, it does not cure; it has its process, you can alleviate the symptoms with all that is sold in the market.
Pregnant women were most ambiguous in discussing the risks regarding the vaccine and the disease. Some participants argued that it would be best to take the vaccine, because if they had the flu they would have limited options due to drug contraindications during pregnancy: Other pregnant women perceived influenza with ambivalence with respect to its severity:

I think influenza can be easily cured. . . I don't know anyone close to me with a flu that knocks you down. But from everything I've read about it, influenza can be very complicated, there are cases of death. But in my family environment and in my work, I work in a salon, it is not very big and there has never been a case of someone who had to be bedridden. (PW, Brazil)
Confusion between influenza and other respiratory diseases led also to complacency, where influenza is seen as a cold (resfriado) with more intense symptoms or as a complication of a common cold:

It is a flu disease that, when not cared for, can become pneumonia, it has more risks and it is a little difficult to cure and that is why you have to be vaccinated every year.
(ARF, Peru).

Let them tell us the specific benefits of the vaccine; we do not know in writing what the benefits are [for children]. Not catching the flu, but what else? Maybe it is also useful for respiratory infections, to make them milder, I don't know, anyway. [We need] something that one can read, short and clear, not technical information. (MC, Uruguay).
Individuals did not perceive themselves to be in a special risk category, except in a few instances with pregnant women-as seen above-and with older adults. Rather, everyone is deemed equally exposed to contracting influenza. Among mothers, what makes children vulnerable to catching influenza are certain behaviors, such as walking barefoot or going out without proper clothing: It's impossible to avoid [the flu] with children. . . children will be children! They go about in their bare feet, they drink ice cold water, they don't want to put on proper clothes. Not even socks, even when its winter! (MC, Brazil).

Complacency: Alternatives for prevention and treatment
All risk groups expressed that there are multiple ways of preventing influenza besides the vaccine, although the vaccine was perceived as the most important, clearly so by mothers and pregnant women (Table 4). Elderly adults and mothers expressed that you could mitigate some influenza symptoms; this, in turn, may moderate risk perception and induce alternative preventive actions.

A home remedy, a good lemon tea with honey and a splash of alcohol, so that the body perspires, relieves itchy throat, a couple of anti-flu [medications] and do not go out to the cold. (OA, Uruguay).
Comparing the perception of complacency with influenza across countries, adults with risk factors in Uruguay perceived themselves as particularly vulnerable. With regard to pregnant women, in Brazil, Paraguay and Uruguay they did not see themselves as particularly vulnerable to the flu, as against Chile and Peru where vulnerability was clearly expressed. Participants across risk groups from Paraguay believed influenza can be cured, a perception that could compete with that of vaccines as an alternative to address the disease. Participants from Paraguay and Uruguay privileged behavioral measures for prevention of influenza such as hygiene and diet as well as home or traditional remedies. In Uruguay, pregnant women were more willing to take the vaccine because they anticipate they will have limited resources if they eventually fall ill, considering restrictions due to their pregnancy.

Convenience: Availability and geographical accessibility
All risk groups perceived relatively good access to influenza vaccination in general. The vaccine is referred as available within health facilities during campaigns.
You can get [the shot] anywhere, when they launch the national campaign. . . You have to look for the schedules and see when its best for you. It is easy to access.

(MC, Uruguay).
In Brazil there was consensus among all risk groups regarding availability and ease of access in public health services with the health system, the Sistema Ú nico de Saúde (SUS). However, mothers complained of closures of those facilities outside the health system that no longer offered the vaccines as well as changes in primary health daily operations: Even during the campaign, there were some complaints of stock-outs, but these were not shared by most participants.

[The vaccine] ends quickly and you need time to go; most [of us] work, to get a leave is very complicated because you have to estimate [when] you will have to go again.
(MC, Paraguay).

The mutualists [health providers] ran out of doses very quickly. This time I got it at a polyclinic at the municipality, which was also [open] in specific days. (PW, Uruguay).
The campaign duration seems to be too short for many who want to use these health services. There were complaints regarding vaccine services scheduling restrictions, especially for participants who work throughout the day and find it difficult to attend health services.
In Brazil expressions were prominent regarding access restrictions for specific risk groups, and the putting in place of stratagems to gain access and avoid payment of private providers: Focus group participants across all countries highlighted that there are no problems regarding access to the influenza vaccine during campaign periods and for priority risk groups. They know where to get the vaccine and, in general, health services are close by and easily accessible.

Convenience: Affordability and willingness-to-pay
The vaccine was generally perceived as free at the point of delivery in government facilities during the campaigns, but expensive in private facilities: Influenza vaccine was generally perceived as accessible from the economic perspective in all countries, with participants reporting that it could be obtained free of charge within public health facilities.

Convenience: Understanding and appeal
All groups except adults with risk factors complained of a lack of information on the vaccine and its availability.
There is no certain information that in such a place they can go whenever they want [to be vaccinated].
The little vaccine information is mostly reported to be received across risk groups from vaccination campaigns and health posts. In some cases, telephone reminders were mentioned as facilitators. Access to insurance was also a facilitator of convenience:

My mom gets [the influenza vaccine] every year in the doctor's office; they call her on the phone and tell her that she has to get vaccinated on that day.
(ARF, Chile).
As I have insurance, they know that I am lacking the vaccine and they simply tell me 'sir, you need to be vaccinated for influenza' and they. . . vaccinate me.
Comparing the convenience of the influenza vaccine across countries, participants from Paraguay expressed least knowledge about influenza and the vaccine. Participants from Chile, on the other hand, expressed a higher level of knowledge. In other countries, participants expressed the need the have more information on the disease and on the vaccine.

Discussion
To our knowledge, this paper analyzes for the first time the range of knowledge, attitudes and practices expressed with respect to influenza vaccine hesitancy by members of high-risk groups living in large urban areas of countries in South America. In Latin America the only previous qualitative studies on influenza vaccine hesitancy focused on Peru [16], while studies in Brazil have focused on pediatric vaccine hesitancy [17].
Our results suggest that confidence is the most important construct associated with influenza vaccine hesitancy across the four risk groups, mostly through perceptions of mistrust in vaccine safety and effectiveness and, to a lesser extent, mistrust in the health system (Fig 1). Moreover, some groups expressed the fear that the vaccine itself might pose risks, having side effects, specially to those who were thought to be in a vulnerable condition, such as pregnant women and adults with risk factors, a perception that was particularly acute in Brazil. Our study suggests that confidence in the vaccine in the five countries studied and across risk groups was higher in Chile and lower in Paraguay, with marked contrasts regarding vaccine effectiveness, safety, and trust in the health system. Both our qualitative and quantitative results are congruent with the national influenza vaccination rates observed across countries, where Brazil and Chile have considerably higher uptake for all risk groups as against Paraguay and Uruguay, and also to a lesser extent in comparison to Peru (Table 5).
Focus group participants generally expressed trust in their health care providers, with exceptions in Paraguay. Even when they had doubts about the vaccine, their confidence in health professionals leads them to gain confidence in the vaccine. However, the medical advice was perceived as ambiguous across risk groups, expressing that sometimes they were instructed to be vaccinated, others not.
Our study also suggests that less information or knowledge is associated to a greater mistrust of the vaccine. Paraguay stands out as a country where focus group participants had very little knowledge of both the disease and the vaccine.
The importance of confidence as a factor in vaccine hesitancy was also observed by the survey arm of our study which obtained quantitative information on the 3Cs for the same risk groups, countries and socioeconomic groups, reporting that individuals with a higher composite indicator of influenza vaccine confidence also had higher probability of being vaccinated in the last year [15]. This study also suggested higher vaccine confidence and vaccine uptake levels in Chile and lower in Paraguay and Uruguay, supporting the qualitative study results we are now reporting.
Other studies in South America also evaluate confidence as the critical component of vaccine hesitancy. Brown et al. found among a survey sample in Brazil that the commonest reasons for hesitancy were issues with confidence (41.4%), efficacy/safety of the vaccine (25.5%) and concerns about adverse events (23.6%) [16]. Reinders and collaborators found in a randomized household survey in urban areas near large hospitals in Peru that confidence and specifically "being afraid of vaccination and its effects" is the most critical factor behind influenza vaccine hesitance. A world-wide systematic review of seasonal influenza vaccination intention and behavior undertaken by WHO between 2005 and 2016 found confidence and specifically a perceived low effectiveness of the vaccine as the most frequently reported reasons for hesitancy across risk groups [16]. Complacency with not being vaccinated was found in our study to be also critical for vaccine acceptance, with a large and diversified set of perceptions such as the importance of a healthy diet lowering the perception of influenza risks while providing alternatives for prevention. We also found a low level of perception of at-risk individuals as members of a risk group, except for older adults and to some extent adults with risk factors, especially those who had a condition such as asthma. The tendency was to perceive greater risk in others, particularly among pregnant women who perceive vaccination to be more important for their babies than for themselves. Across countries, complacency of influenza was higher in Paraguay and Uruguay than in other countries, possibly due to the high value given in these countries to traditional measures to prevent illness, a situation that could take the vaccine out of the spotlight or relegate as a secondary preventive measure.
Our data indicate that confidence and complacency are intimately related. We found that complacency with the risks posed by influenza in the context of trust in traditional medicine and distrust in the health system-as observed for the case of Paraguay-was associated to a perception of side-effects and even of the view that the vaccine can cause influenza. Furthermore, our study suggests, as in the case of older adults, that low levels of complacency with influenza risks can support higher levels of confidence in the vaccine. These possible associations across confidence and complacency require further research to ascertain cognitive and affective links.
In our study, less information and knowledge regarding influenza and the vaccine are qualitatively related to a greater mistrust of vaccination. Focus group participants from Paraguay expressed the most doubts on both topics. The more people distrust the vaccine, the more they will have recourse to traditional or alternative strategies to prevent influenza, such as diet, hygiene, clothing, environment. Those who do not perceive influenza to be a serious or lifethreatening disease tend to be more vaccine hesitant.
Findings on confidence and complacency point to the importance of health promotion and communication strategies to improve knowledge about influenza risks and to establish that hygiene measures do not replace vaccination. While all risk groups acknowledged that influenza could be a potentially mortal disease, they also expressed confusion between influenza and other respiratory infections. In this sense, influenza is seen either as a complication of the common cold or as something very prevalent-ordinary even-posing a small threat to the health of most people who have it.
Reinders and collaborators found particularly low levels of self-perception of pregnant women and older adults as high-risk groups for influenza in Peru [16]. The perception of individuals as belonging to risk groups can be enhanced through conditioning social program benefits, as in the case of children in Brazil, where having a vaccination card up to date is a requirement to receive Bolsa Família, a government program for income redistribution [18].
For the past decades, social scientists have discussed how different Explanatory Models (EM) for a disease are important for understanding how individuals explain their causes, symptoms, and treatments, and hence for the role of complacency in vaccine hesitancy. The flu and the common cold are associated by participants to feeling cold or exposing parts of the body to lower temperatures (bare feet specially), and prevention involves dressing warmly or consuming warm foods and liquids. A healthy diet is cited by participants as another important preventive measure, and that involves consuming fruits rich with vitamin C or associated with boosting the immune system. In our findings, influenza vaccination is perceived as competing with other practices that are also considered preventative, whether by themselves or as complements to vaccination. Mothers of young children and pregnant women tend to over rely on hygiene and a healthy diet-which are unquestionably positive for pregnancy and child development-as preventative of influenza. For the older adults and adults with risk factors, their perceived vulnerability to influenza may be associated with the observed expressions of interest in vaccination, making them less complacent. Future studies focusing on folk medicine and existing EM for the flu and the cold in urban contexts of Latin America would be important to expand on our findings [19,20].
Our study suggests that individuals belonging to specific risk groups for influenza lack examples of the severity and hence the uniqueness of influenza, possibly due to the paradoxical effects of vaccination programs as they reduce mortality. The exception were older adults, who had both more personal experience with the vaccine and knew someone who had been terribly ill due to influenza. Reinders and collaborators found greater vaccination rates among persons who perceived the severity of influenza in Peru [16]. Our survey study found complacency to be the component of the 3C model with the weakest evaluation given prejudices, knowledge and risk perception. However, as mentioned already, complacency was not as strongly associated with vaccination rates as confidence was. Further afield, in Australia extreme events lived by communities pivoted parental responsibility for child vaccination [21].
The convenience of influenza vaccination was the most positively evaluated component in our study, both across risk groups and countries. None-the-less, perceptions on restrictions with information, particularly in Brazil, were observed, while opportunities to improve access were found across all risk groups and countries. Perceptions of program appeal were not so much negative as absent, together with examples of negative appeal in terms of possible disrespectful and culturally inappropriate treatment of indigenous population in Peru, as found for health care in general for indigenous populations in other countries [22,23]. In another arm of our research, we explored the balance given by country influenza vaccination programs to strategies focused on increasing confidence and convenience and on lowering complacency [24]. Congruent with the qualitative findings, we found that programs across the five countries privilege convenience through supply-side strategies. Our qualitative findings suggest the need for influenza vaccination programs to address confidence and complacency through directed strategies.
A limitation of our qualitative research is its primary focus on the situation of vaccine hesitancy within risk groups rather than at the country level. Furthermore, our results cannot be generalized to urban populations across countries, given the primary aim of the research of identifying qualitative patterns. Even so, our sample aimed to include countries with both high performing vaccine programs (Brazil and Chile) and low-performing programs (Paraguay, Peru and Uruguay, see Table 5). A possible limitation of the focus group technique employed is that it could have inhibited responses that differed markedly from those held by the majority of the participants.
In spite of the limitations stated, our analysis did suggest that influenza vaccination hesitancy has greater similarities than differences across the countries studied, possibly due to similar social, health system and cultural characteristics, as in our analysis of national vaccination programs [24].
The specific findings regarding each of the 3Cs in the vaccine hesitancy model across risk groups and countries provide valuable information for program planning and to guide intersectoral policies and strategies. Given that vaccine convenience perceptions are structured and focused on limited problems, efforts to improve this dimension of hesitancy can be readily implemented and can support strategies to improve confidence and complacency. Effective and assertive communication strategies can be developed to target specific risk groups and health workers involved in vaccination. Messages should emphasize the risks posed by influenza and the benefits of being vaccinated, and these messages need to be divulged all year, not only during the influenza vaccination campaigns. Spokespersons can be designated and trained to demonstrate the value of vaccination as well as to solve all vaccination doubts through interactive, digital-based media as available. However, vaccine hesitancy is very often based on emotional responses and countering emotion with reason (e.g. unbiased information) typically doesn't work [24]. Bolstering vaccine confidence on the basis of emotive as well as rational concepts adapted to specific risk groups should therefore be the top priority to reduce hesitancy. Given that the most trusted sources of information are health professionals, efforts should be made to provide them with training, incentives and tools focusing on the needs and sensibilities of specific risk groups. Three groups of actions have been proposed: addressing opportunities to modify confidence in vaccine effectiveness and concerns about safety; bolstering altruism in specific social contexts through normative messaging; and pivoting change in health professionals through incentives, sanctions and requirements [25,26]. Communication strategies must be precise to avoid counterproductive results which could lead to reinforcing hesitation to vaccination [27].
The fact that focus group participants across risk groups from Paraguay and Uruguay expressed greater influenza vaccine hesitancy suggest the importance of further analyzing country-specific factors such as the reliance on traditional medicine as well as differences in health system trust.
The COVID-19 pandemic will likely affect uptake for seasonal influenza vaccination given changes in risk perception associated with mortality and economic impact of viral respiratory diseases [28]. Influenza vaccination has been recommended to reduce hospitalizations and relieve resources to care for COVID-19. The COVID-19 vaccine may also alter seasonal influenza vaccination programs and influenza vaccine uptake. Seasonal influenza vaccination in Brazil was anticipated one month in March 2020 to help reduce respiratory diseases and help to cope with COVID-19. It is important to prepare for COVID-19 vaccination addressing lessons with influenza vaccine hesitancy. Influenza and COVID-19 vaccination strategies may converge in the future so our findings can be useful to prepare for such a scenario.

Conclusions
The study of influenza vaccine hesitancy in five countries of South America revealed a wide range of perceptions regarding confidence and complacency across risk groups and, to a lesser extent, across countries. While risk groups share some common perceptions, they also have specific conceptions that present intervention opportunities. Information and communication strategies need to be tailored for specific risk groups to address their concerns and the take advantage of their varied relationship to health providers. Country-specific measures should consider contextual factors. Confidence and complacency should be addressed in their own right, through strategies that can increase one while reducing the other, in synergy.