|Home | About | Journals | Submit | Contact Us | Français|
Vaccine acceptance is a critical component of sustainable immunization programs, yet rates of vaccine hesitancy are rising. Increased access to misinformation through media and anti-vaccine advocacy is an important contributor to hesitancy in the United States and other high-income nations with robust immunization programs. Little is known about the content and effect of information sources on attitudes toward vaccination in settings with rapidly changing or unstable immunization programs.
The objective of this study was to explore knowledge and attitudes regarding vaccines and vaccine-preventable diseases among caregivers and immunization providers in Botswana, the Dominican Republic, and Greece and examine how access to information impacts reported vaccine acceptance.
We conducted 37 focus groups and 14 semi-structured interviews with 96 providers and 153 caregivers in Botswana, the Dominican Republic, and Greece. Focus groups were conducted in Setswana, English, Spanish, or Greek; digitally recorded; and transcribed. Transcripts were translated into English, coded in qualitative data analysis software (NVivo 10, QSR International, Melbourne, Australia), and analyzed for common themes.
Dominant themes in all three countries included identification of health care providers or medical literature as the primary source of vaccine information, yet participants reported insufficient communication about vaccines was available. Comments about level of trust in the health care system and government contrasted between sites, with the highest level of trust reported in Botswana but lower levels of trust in Greece.
In Botswana, the Dominican Republic, and Greece, participants expressed reliance on health care providers for information and demonstrated a need for more communication about vaccines. Trust in the government and health care system influenced vaccine acceptance differently in each country, demonstrating the need for country-specific data that focus on vaccine acceptance to fully understand which drivers can be leveraged to improve implementation of immunization programs.
Vaccines are one of the most important public health achievements in history resulting in significant decreases in the prevalence of many childhood diseases . Vaccination rates have steadily increased in both developed and developing countries through the World Health Organization’s Expanding Program on Immunization and various public-private partnerships. However, disparities in rates of new vaccine adoption and sustained vaccination rates across different regions persist. In addition, an increase in the number of communities in both developed and developing countries with low or decreasing vaccination rates has resulted in disease outbreaks [2,3]. While access to health care services can lead to disparities in vaccination rates, vaccine acceptance is another critical component of sustainable immunization programs. Studies have suggested that vaccine-hesitant individuals, who hold various degrees of indecision about specific vaccines or vaccination in general, can significantly contribute to decreases in immunization rates and slow uptake of newly introduced vaccines [4,5,6,7,8].
In light of these changing trends, there is increased focus on defining vaccine hesitancy and identifying key determinants that can be addressed by immunization programs, although there is not significant evidence supporting specific interventions to reduce hesitancy [9,10]. A recently published review suggests that region- or country-specific research is crucial given the complex array of individual, sociocultural, and political factors that influence vaccine acceptance . In international studies, several factors have been identified that affect vaccine acceptance, including lack of knowledge about the need for vaccines, misperceptions about vaccines and vaccine-preventable diseases, fear of side effects after vaccination, inconsistent vaccine availability, lack of trust in the health system, and history of a negative interaction with immunization providers [11,12,13,14,15,16,17]. In a multinational study of immunization managers in low- and middle-income countries, a variety of views were presented, but few managers were concerned about vaccine hesitancy . A study of health care workers in Greece, France, Croatia, and Romania demonstrated the predominance of concerns about vaccine safety . In developed countries in Europe facing financial crisis, health inequalities are expected, with concern that vaccination coverage will decline . Reliable sources of information about immunizations vary by country. In settings with expanding immunization programs, access to information about newly introduced vaccines may be particularly important. Globalization has resulted in a rapid increase in information exchange. The effect of pro- and anti-vaccine news from developed countries on worldwide attitudes toward vaccination may significantly impact attitudes toward vaccines [21,22].
We explored attitudes and beliefs related to vaccine acceptance and the methods of communication about vaccines leading to vaccine acceptance in three countries in diverse regions of the world and with immunization programs in various stages of development: Botswana, the Dominican Republic (DR), and Greece. Botswana and the DR are middle-income countries in different parts of the world with expanding immunization programs within a developing public health system, whereas Greece has an established health care delivery system that has been acutely affected by significant resource limitations . Because current studies suggest that vaccine acceptance in different regions is dependent on factors specific to that sociopolitical and cultural environment, independent evaluation of these factors in different regions is essential . Investigating attitudes toward childhood vaccines and the perceived influence of information sources in these settings can provide important insights regarding vaccine acceptability as new vaccines are introduced to ensure ongoing sustainability of immunization programs.
This qualitative study was conducted in two middle-income countries, Botswana and the DR, where immunization programs are primarily supported through the public health system, and Greece, an upper-income country with an immunization program supported through both public and private health care systems. These three countries were selected to identify both similarities and differences in vaccine acceptance in light of the varied cultures, sociopolitical environments, and stages of development of each nation’s vaccination program, as detailed in Table 1. Due to this complexity of factors, it is important to measure attitudes in countries that have different immunization practices. Our established relationships with these countries allowed us to carry out this research and prepare for next steps in increasing vaccine acceptance.
Botswana has a rapidly growing economy but is challenged by the third-highest rate of HIV/AIDS in the world . The DR also has a growing economy but has significant income disparities. While national estimates for immunization rates are high in both Botswana and the DR, there is significant variation in coverage rates by district, suggesting that implementation of current vaccination recommendations is not uniform [29,30]. In each location, a small proportion of children receive care through the private sector where they may also receive vaccines, particularly those that are not included in the national vaccine program. Both settings have a large immigrant population—from Zimbabwe in Botswana and from Haiti in the DR—and these children are eligible to receive immunizations through each country’s National Immunization Program (NIP).
In Greece, the NIP includes universally recommended vaccines free of charge to all resident children, including immigrants . While children can receive vaccines in primary health centers and non-governmental organizations at no additional cost, the majority receive vaccines through the private sector . An economic crisis has engulfed Greece since 2008, which has undoubtedly been detrimental to the health care system . For example, many families lost public health insurance coverage because of unemployment, creating significant gaps in insurance benefits. Although research at the beginning of the crisis demonstrated that incomplete vaccination was not associated with insurance status but rather was driven by socioeconomic factors such as parent age and occupation, this relationship is incompletely understood .
We implemented focus groups among caregivers of children younger than 5 years and immunization providers to identify salient themes associated with vaccine acceptance and hesitancy, to investigate knowledge regarding childhood vaccines, and to explore content of and access to information about vaccines and vaccine-preventable diseases. Qualitative methods with focus groups were utilized to encourage candid responses, to allow participants to build on each other’s ideas, and to allow participants to focus the discussion on topics most relevant to their experiences . For caregivers, we included any individual who was the parent or guardian of a child 5 years of age or younger who made decisions regarding vaccination for that child, regardless of the caregiver’s attitudes toward vaccination. For immunization providers, we included any clinician (public health practitioners, nurses, or physicians) who administered vaccinations to children 5 years of age or younger. Participants were compensated for their time with refreshments and reimbursement of transportation costs at all sites. In Botswana and the DR, caregivers and providers were recruited from clinics in low-, middle-, and high-income neighborhoods across urban, peri-urban, and rural communities with a range of sociodemographic characteristics and immunization rates to capture a comprehensive array of attitudes toward vaccination. In Greece, caregivers were recruited from kindergarten networks in communities of low-, middle-, and high-income neighborhoods, while providers were recruited through public and private hospitals and government agencies. All focus group participants signed an informed consent document, with review and approval by the Institutional Review Boards of the University of Pennsylvania and Agia Sofia Children’s Hospital in Greece, Conabios in the DR, and the Botswana Ministry of Health.
The focus group guide was developed based upon a conceptual framework grounded in the health belief model but also integrating theoretical models related to vaccine hesitancy [33,34] (Fig 1). The conceptual framework and focus group guide were developed in consultation with a study team member with expertise in qualitative research methods (JS). Themes explored included three domains: (1) knowledge, communication, and the health care system; (2) attitudes and beliefs; and (3) logistics of vaccine delivery. Questions for both the focus groups and interviews were read from a standardized guide eliciting knowledge and beliefs about vaccines and vaccine-preventable diseases, sources of vaccine-related information, influence of popular media on vaccine-related beliefs, facilitators and barriers to obtaining vaccines, and attitudes toward the health care system. All focus group moderators and interviewers were trained by a qualitative researcher (JS). The focus group and interview guides were translated and back-translated into Setswana, Greek, and Spanish.
Focus groups led by trained research associates were conducted in Setswana, English, Greek, or Spanish, depending on the preference of the participants. Focus groups for providers and caregivers were conducted independently. Those groups with one or two participants were conducted as semi-structured interviews. Sessions lasted approximately 90 minutes and were audiotaped. Sessions were then transcribed and translated into English. Facilitators of focus groups fluent in native languages validated translation quality.
Translations uploaded into NVivo 10 (QSR International, Melbourne, Australia), a qualitative data analysis software program, were analyzed for common themes through a deductive approach based on the conceptual framework. As new themes emerged, an inductive approach was used to add to the coding schema. Two independent coders coded the transcripts, with discrepancies rectified on the basis of review in an iterative process. Kappa values were calculated between the initial two coders after reviewing and modifying discrepancies. If there was lack of agreement between the two coders after review, a third coder evaluated the discordance to reach a Kappa value of 1.
The coding schema on completion was reorganized from three domains into four domains: (1) knowledge and communication, (2) interactions with the health care system; (3) attitudes and beliefs; and (4) logistics of vaccine delivery. Within each theme, comments were categorized into nodes within NVivo that were neutral toward, supported, or presented a barrier to vaccine acceptance.
We conducted 37 focus groups and 14 semi-structured interviews among 96 providers and 153 caregivers across three countries (Table 2). The majority of participants received or provided care within a public sector setting. Male and female providers in different practice settings in all involved communities were recruited.
Within the four domains, there were 23 themes, although two were specific to Greece. In the Botswana focus groups, 945 comments were included in the final analysis with 1122 attributions to nodes (~20% of comments were attributed to >1 node). Ninety-one percent of Kappa values using NVivo software were above 0.6, with a mean of 0.9. In the DR focus groups, 1076 comments were included in the final analysis with 1236 attributions to nodes (~15% of the comments were attributed to >1 node). Ninety-eight percent of Kappa values using NVivo software for the primary coders were above 0.6 with a mean of 0.97. In the Greece focus groups, 1080 comments were included with 1225 attributions to nodes (~14% of the comments were attributed to >1 node.) Ninety-seven percent of Kappa values using NVivo software for the primary coders were above 0.6 with a mean value of 0.97. The frequency with which each theme was discussed, either as supporting vaccine acceptance, creating a barrier to vaccine acceptance, or without impact on vaccine acceptance, was tallied to identify prevalent themes (S1 Appendix). The numbers next to each node are the total number of comments elicited across all groups. As study methodology allowed participants to speak freely about topics, these values should not be considered precise measurements of importance. The domains of logistics of vaccine delivery and attitudes and beliefs confirmed published literature, while the domains of knowledge and communication, as well as interactions with the health care system, are highlighted in this manuscript.
In recognizing that knowledge about vaccine-preventable diseases, vaccines, vaccine efficacy, and vaccine safety influences vaccine acceptance, it is critical to assess how caregivers and providers build their knowledge base. Results demonstrate the relative impact of information from health care providers, the media, community members, and society in each country. A dominant theme that emerged was the recognition of health care providers as the main source of vaccine information for caregivers with indications that they relied on them more than other information sources. Similarly, health care providers recognized their role as information sources for their patients and looked to medical literature primarily as an information source to maintain their knowledge. Representative quotes are highlighted in Table 3.
Within this theme in Botswana and the DR, information from health care providers was described as positively supporting vaccine acceptance, while in Greece, many participants felt that they received information from health care providers that negatively affected vaccine acceptance. Similar trends were noted for both community members and media as information sources. These themes were mentioned less frequently across all sites, but participants from Greece were more likely to consider messages from community members or popular media as negatively affecting vaccine acceptance.
Forms of media used by study participants varied between countries, with emphasis on radio, television, and newspapers in Botswana; television and Internet in the DR; and the internet in Greece. Representative quotes from caregivers and providers are highlighted in Table 4. In Botswana, participants noted that media delivered positive messages about vaccines usually in the form of public service announcements or general updates on new vaccines that became available. If media messages were negative or raised questions, participants indicated that they would confirm the information with health care providers rather than trust the media’s negative message about vaccines. In the DR, while the majority of information received through media was reported as supporting acceptance, many participants pointed out that access to television or the internet is not consistent throughout the country, possibly limiting the impact of this information. In Greece, caregivers reported pediatricians as their main source of information for vaccination. However, caregivers were frequently exposed to media messages, especially through the internet. A significant number of caregivers expressed a desire to discuss any negative concerns with their family doctors before making vaccine decisions. However, some caregivers also described a direct influence of media messages on their decision-making regarding vaccination, regardless of consultation from their family doctor.
Across all three sites in both caregiver and provider focus groups, the theme of insufficient or ineffective communication about vaccines was prevalent, especially in Greece (Table 5). Caregivers were asked the question, “Overall, do you think there is enough information provided about the vaccines your children need?” while providers were asked, “Overall, do you think there is enough information provided about the vaccines you administer to children? In Botswana, providers indicated that increasing education through media with positive messages might be helpful. Caregivers suggested a number of methods to increase information availability, including increasing information on “Under 5” cards, increasing media presence such as billboards, and having more information sessions presented by health care providers. In the DR, more information was requested to increase caregivers’ knowledge and for clarification of confusing messages. In response to the question, “What else would be helpful to know about vaccines?,” caregivers requested more information through radio, television, and internet sources, whereas providers suggested increased information dissemination from the public health system. Providers also noted the distance that many people live from health care settings and the need to disseminate information through newspapers or television more frequently than just during outbreaks.
In Greece, caregivers described the need for more information on vaccination and for better dissemination strategies. Many participants reported inconsistent messaging coming from health care providers often creating confusion among caregivers. Ineffective communication by the media, specifically on the internet, in the form of either unclear or too much information, was also cited. Caregivers suggested that they would appreciate more reliable, consistent information from media sources (such as television) as well as information circulating through the educational system. To minimize conflicting statements, they also expressed the desire for a single responsible body that would coordinate and communicate up-to-date and consistent information about vaccines.
In Botswana, participants expressed an overwhelming trust in the government and closely aligned the health care system and government as one unit. Interactions with health care workers were equally positive and negative, with reliance on health care providers for information yet concerns about providers’ knowledge, skills, or compassion. In contrast in the DR, there were few comments related to trust in the system or government. Similar to Botswana, participants reported experiences with health care workers that both supported and set up barriers to immunization. In Greece, participants reported significant barriers related to access to vaccination and a lack of public health infrastructure. This resulted in caregivers choosing to obtain vaccines from the pharmacy and pay for vaccine administration in the private sector rather than trying to navigate the public system. Representative quotes are highlighted in Table 6 and Table 7.
In this qualitative study, we explored the role of knowledge and communication on vaccine acceptance in countries with both resource limitations and rapidly changing immunization programs. While there is a growing body of literature investigating drivers of vaccine acceptance, few studies have focused on the impact of different methods of information dissemination on vaccination programs in middle-income countries with expanding immunization programs. Major findings from this work demonstrate that while both caregivers and immunization providers from a wide range of communities in Botswana, the DR, and Greece report high levels of support for vaccination because of knowledge of vaccines and vaccine-preventable diseases, a need still exists for increased access to accurate and reliable information about vaccines. Contrasts in responses between countries highlighted the importance of information sources including health care providers, medical literature, and the media in supporting vaccine-related knowledge. For example, while the media primarily supported positive vaccine beliefs and was viewed as a way to disseminate messages to encourage vaccination in the DR and Botswana, media messages were largely viewed as negative in Greece. There were also differences in the degree of trust in the public health system across sites. These differences highlight the influence of context on vaccine acceptance and emphasize the importance of region- or country-specific research to understand local drivers of vaccine acceptance.
While disseminating scientifically sound information about vaccines is critical, reliable information alone does not necessarily promote sufficient trust in vaccines to influence decision making about vaccines, especially among individuals with strongly held beliefs . Recognizing which source of information is most relied upon by caregivers and providers and which communication method is most effective is essential when considering strategies to improve local vaccine acceptance. Provider recommendation is known to be a key predictor of vaccine acceptance [35,36]. As demonstrated in the theme of health care providers and medical literature as an information source, participants across all three sites relied on health care providers as their primary source of information, although many participants from Greece identified conflicting messages from providers as an issue, highlighting the importance of consistency. In contrast, in Botswana, health care providers were overwhelmingly recognized as the trusted, consistent information source for caregivers, especially when confronted with negative media messages.
Beyond direct communication with health care providers, there are multiple ways to disseminate information about vaccines. Participants’ comments about the use of multiple types of media highlight this idea. However, research establishing the effectiveness of different communication strategies on changing vaccination behavior is in its infancy [22,37]. Mass media campaigns are often used by public health systems to promote health messages and have been associated with improved child health outcomes . This study focused on the role of media given the proliferation of social media outlets and rapid dissemination of both pro- and anti-vaccine messages. In Botswana and the DR, popular media as a tool for the dissemination of positive messages was perceived to be a potentially effective tool. Negative vaccine messages in the popular media were not viewed as reliable and, therefore, had little to no influence on attitudes and beliefs about vaccines. In contrast, in Greece, the media, specifically the internet, was more often described as disseminating negative messages that did not support vaccine acceptance. Additionally, such messages were often considered in vaccine decision-making. While the comments we elicited demonstrate the varied types of media available to participants in different regions, the theme of insufficient communication highlights that both caregivers and providers are eager for more substantial, accurate information from reliable sources, regardless of the medium through which it is communicated.
A lack of confidence in vaccines in the face of a knowledgeable public is known as the vaccine confidence gap and indicates that an additional mix of factors affects the public’s trust and subsequent vaccine acceptance. In fact, citizens’ trust in immunization programs and the health care system requires understanding the complex religious, socioeconomic, historical, and political contexts of different regions . The multiple references to trust in health care workers and the health care system reflects this concept. In Botswana, our results showed a high level of trust in the government and heath care system as a provider of resources and information. In contrast, trust in the health care system and government was very low among Greek respondents, which may reflect the recent economic instability and loss of many government services for a large proportion of the population. Our results illustrate the complexity of factors influencing vaccine acceptance and also suggest potential targets to enhance the success of immunization programs.
This is a qualitative study among a sample of caregivers and immunization providers. While we recruited from regions of each country and from both private and public health care settings, our population is a non-random sample of each country’s population. Recruitment strategies in each country were designed with local collaborators, but certain populations in each country may be underrepresented in our sample. In the DR and Botswana, we primarily recruited through clinics. There may be unique facilitators of and barriers to vaccination in other locations where health care services may be more dispersed or among families that do not interact with the health care system for preventive care. In Greece, we were able to capture a range of social classes in Athens as well as in the semi-urban area because of recruitment through the kindergarten networks.
We have identified dominant themes but qualitative research cannot quantitatively measure predictors of vaccine acceptance. Counts of dominant themes as presented are hypothesis-generating and can inform the development of survey instruments that can be used in future quantitative studies across a broader range of communities and health care delivery settings in each country for more generalizable data. Our focus group guide was based upon a conceptual model grounded in behavioral theories, but it is possible that we may have missed additional salient themes. However, we attempted to address this by adding themes that emerged during the coding process.
Prior studies have indicated that knowledge about vaccination is important but not sufficient to change behavior . In this study, participants communicated an understanding of the need for vaccination and had a significant understanding of vaccines and vaccine-preventable diseases; however, participants also expressed interest in receiving more information about vaccination. To address some of the challenges identified by participants related to knowledge and attitudes, the vaccine delivery system has the powerful ability to be an influential and cost-effective purveyor of vaccine information.
In a country such as Botswana where trust in government is high, this trust can be leveraged to bolster communication about vaccines. In the DR, such messages may be more effective if disseminated through an individual’s health care provider or social network, as study participants did not voice significant trust in the public health system. In Greece, participants emphasized the dependency on health care providers to help guide decision-making about vaccines yet also reported inconsistent information. To support effective communication and ensure consistency, providers can be targeted for vaccine education interventions. This may be an important mechanism to strengthen trust in health care systems and help close the vaccine confidence gap. Future quantitative work based on this study should focus on identifying characteristics of populations that correlate with vaccine acceptance and on identifying strategies to remedy the identified barriers.
Vaccine delivery and acceptance are complex issues affected by a variety of local and personal factors. These data reinforce the notion that while some facilitators and barriers to vaccination may be generalized, an understanding of local determinants is crucial to supporting effective implementation of immunization programs. For these reasons, working with local teams and developing materials that address local situations and concerns are imperative to successful implementation of new or maintenance of existing vaccine programs.
We acknowledge the work of Marc Callender, MD, Caitlin Fasano, and Rodney Finalle, MD; in addition to support from the Pincus Fellowship; the Vaccine Education Center at Children’s Hospital of Philadelphia, and Theoklis Zaoutis, MD, MSCE.
We acknowledge the work of Marc Callender, MD, Caitlin Fasano, and Rodney Finalle, MD; in addition to support from Theoklis Zaoutis, MD, MSCE. This publication was made possible through core services and support from the Penn Center for AIDS Research (CFAR), an NIH-funded program (P30 AI 045008). Lori K. Handy was supported by T32-AI096345 and L40-AI-107932. Additional funding was provided by Mabel Freeman Fund and the Vaccine Education Center at Children’s Hospital of Philadelphia. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
All data are available within the paper and its Supporting Information file.