Abstract
Parental concerns about vaccine safety have grown in the United States and abroad, resulting in delayed or skipped immunizations (often called “vaccine hesitancy”). To address vaccine hesitancy in Washington State, a public–private partnership of health organizations implemented and evaluated a 3-year community intervention, called the “Immunity Community.” The intervention mobilized parents who value immunization and provided them with tools to engage in positive dialogue about immunizations in their communities. The evaluation used qualitative and quantitative methods, including focus groups, interviews, and pre and post online surveys of parents, to assess perceptions about and reactions to the intervention, assess facilitators and barriers to success, and track outcomes including parental knowledge and attitudes. The program successfully engaged parent volunteers to be immunization advocates. Surveys of parents in the intervention communities showed statistically significant improvements in vaccine-related attitudes: The percentage concerned about other parents not vaccinating their children increased from 81.2% to 88.6%, and the percentage reporting themselves as “vaccine-hesitant” decreased from 22.6% to 14.0%. There were not statistically significant changes in parental behaviors. This study demonstrates the promise of using parent advocates as part of a community-based approach to reduce vaccine hesitancy.
Introduction
In recent years parental concerns about vaccine safety have grown in the United States and abroad, resulting in delayed or skipped immunizations (often called “vaccine hesitancy”). When fewer children are immunized, more people are at risk of getting dangerous infectious diseases. In 2011, Washington State’s school entry exemption rate was 6.2%, a figure that had doubled over the previous 10 years and was 3 times the national average (Immunization Program CHILD Profile, Statewide Summary of Immunization Coverage for Kindergarten SY 2009-2010, 2010).
Background/Literature Review
The literature on vaccine hesitancy is emerging, as the rise in parental vaccine hesitancy is a relatively recent phenomenon. Many researchers have acknowledged the paucity of multifaceted, evaluated interventions to address vaccine hesitancy (Dube, Gagnon, & MacDonald, 2015; Jarrett, Wilson, O’Leary, Eckersberger, & Larson, 2015; Kaufman et al., 2013; Sadaf, Richards, Glanz, Salmon, & Omer, 2013; Saeterdal, Lewin, Austvoll-Dahlgren, Glenton, & Munabi-Babigumira, 2014). Our review found very few similar programs that had comprehensive descriptions and/or evaluations; therefore we focus on existing literature regarding educational outreach and social marketing efforts to address vaccine hesitancy.
Interventions involving written educational materials have shown mixed results regarding impact on parents’ knowledge and attitudes about vaccines and on their intent to vaccinate. A number of studies have shown positive results (Chan, Cheung, Lo, & Chung, 2007; Clayton, Hickson, & Miller, 1994; Gillespie et al., 2011). However, Nyhan, Reifler, Richey, and Freed (2014) found that traditional messaging, especially those with images and dramatic narratives, may factually reinforce negative attitudes about vaccines among some groups.
Less traditional modes of education (e.g., telenovelas, graphical information, and Web-based decision aids) have shown some impact (Cox, Cox, Sturm, & Zimet, 2010; Gowda, Schaffer, Kopec, Markel, & Dempsey, 2013; Kepka, Coronado, Rodriguez, & Thompson, 2011; Wallace, Leask, & Trevena, 2006). Kaufman et al. (2013) assessed four randomized controlled trials evaluating the effects of face-to-face interventions delivered by community health workers or midwives to parents to educate about early childhood vaccination. Evidence for the efficacy of all four interventions was found to be overall of “low quality,” and at risk for bias. In a study based in Pakistan, trusted community members attended informational meetings to discuss immunization rates, costs, and benefits of childhood vaccines, and were encouraged to spread positive messages through the community. A year later, a follow-up survey showed a significant increase in the uptake of diphtheria pertussis tetanus vaccination in the study group compared to controls (Andersson et al., 2009).
Using social marketing techniques to address vaccine hesitancy has produced promising results for promoting human papilloma virus (HPV) in adolescents. A 3-month campaign in rural North Carolina that used posters, brochures, and a website to raise HPV vaccine awareness among mothers, and health care providers of adolescent girls found a slight increase in vaccination rates in intervention communities compared to nonintervention communities (Cates, Shafer, Diehl, & Deal, 2011). Another campaign, also aimed at parents and medical providers, used brochures and posters, radio public service announcements, and an online continuing medical education course to promote uptake of HPV vaccines among adolescent boys. Boys in the intervention group were 34% more likely to get vaccinated in an intervention county than in a control county (Cates, Diehl, Crandell, & Coyne-Beasley, 2014).
Recent studies using social marketing approaches for childhood vaccines are less prevalent in the literature. The “Immunise Australia” (Caroll & Van Veen, 2002) and the “I Immunise” (Attwell & Freeman, 2015) campaigns (both implemented in Australia) showed improvements in awareness and attitudes toward vaccines. “I Immunise” (the more recent of the campaigns) used a values-based social marketing approach to target vaccine-hesitant parents in a community known for lower than national vaccine coverage. The campaign used community members as “community advocates” in Web-based video narratives, posters, billboards, Facebook memes, and newspaper advertisements, distributed throughout the community. A survey found three quarters of parents identified as having an “alternative lifestyle” felt “more positively” about vaccination after exposure to the campaign (Attwell & Freeman, 2015). This intervention was similar to the Immunity Community in that real people served as advocates and shared their own values and personal stories; however, “I Immunise” did not use the “community advocates” for direct advocacy with other parents. We are not aware of any interventions similar to the Immunity Community that mobilized parents to be agents of social change.
This article presents evaluation results from a 3-year community-based intervention called the “Immunity Community,” implemented in Washington State. The intervention mobilized parents who value immunization and provided them with tools to engage in positive dialogue about immunizations in their communities. This article provides a brief description of the intervention and summarizes key evaluation findings, including the level of parent advocate (PA) engagement and the impact of the program on local policy and parental knowledge, attitudes, and behaviors.
Method
Program Description
A public–private partnership, called Vax Northwest (the partnership), was formed in 2008 to address vaccine hesitancy in Washington State, recognizing the need for creative, evaluated interventions. The partnership operates under the collective impact model (Kania & Kramer, 2011), with one organization serving as the backbone organization. This backbone organization also served as the direct implementer of the program (program staff). All partners provided some funding and/or offered expertise.
The Immunity Community was the primary focus of the partnership’s early work and continues to be a cornerstone program. The goals of the Immunity Community program were to address parental vaccine hesitancy by empowering parents to be immunization advocates, improve awareness of immunization as a social norm among parents at participating sites, and change those parents’ attitudes and behaviors. The program also had an intentional focus on influencing organizational and/or local policies associated with communication about and monitoring of children’s vaccination status, whether at participating sites or in the larger community. The desired long-term impact was reduced vaccine hesitancy and exemption rates. The program recruited, engaged, and supported volunteer PAs, to spread positive messages about vaccines in an effort to provide a counterbalance to the strong negative rhetoric about vaccines being circulated on social media and in the press.
The program was based on the principles of social marketing, a process that uses marketing principles and techniques to influence target audience behaviors that benefit society as well as the individual (Kotler & Lee, 2011). The principles have been successfully applied to address a range of public health goals such as reducing tobacco use, encouraging the use of bike helmets, and stopping the spread of communicable diseases (Kotler & Lee, 2011). The partnership was motivated to develop a program to address vaccine hesitancy using social marketing principles. They also recognized that in times of high parental concern, expertise delivered by a medical provider may be perceived as less valuable than information delivered by an open, trustworthy, and likeable messenger, perceived to be working towards the same goal as a parent (Opel, Diekema, Lee, & Marcuse, 2009).
A creative communications agency specializing in the development of social marketing strategies for nonprofit organizations was hired to help design the program, in collaboration with representatives from the partnership and a community advisory panel including parents and pediatricians. The result was the Immunity Community brand and logo along with materials and resources that included the following: a Parent Action Guide (a resource for PAs), posters, postcards, stickers, and branded giveaway items (e.g., first aid kits, notebooks), which were used by PAs at their schools, at events, and on social media. The partnership also created a website that included a password-protected site for PAs and public facing information pages to which PAs could refer friends and family (www.vaxnorthwest.org). The program’s suite of materials was revised and expanded each year, based on formative feedback from parent focus groups and key informant interviews with program staff and participants.
The program was designed to focus efforts in defined geographic communities. The first community was a suburban school district outside a major metropolitan area. Subsequently, the program expanded to a second community—a small city in a rural county. Both focus area populations are relatively more affluent and more educated than their respective counties overall. New program sites within the focus areas were engaged each year. Table 1 provides an overview of the intervention’s evolution.
Immunity Community Components and Evolution Over the 3-Year Program
NOTE: PA = parent advocate; TA = technical assistance.
The program staff first recruited interested elementary schools, child care centers, and preschools to participate. A site director or school nurse (at elementary schools) committed to be the Immunity Community site representative and support the PA. Program staff worked with the sites’ leadership to find one to two interested parents to be PAs. Program staff provided a comprehensive 4-hour initial training for PAs. Training included background information on vaccines and the issue of vaccine hesitancy and emphasized how to communicate effectively and respectfully in conversations with peers, how to plan immunization-positive activities in the community, and tips on engaging on social media. After initial training, program staff continued to provide PAs with resources, technical assistance, and support, including group meetings, one-on-one in-person check-ins, a monthly newsletter, and a private Facebook group.
Evaluation Design
The evaluation was mixed-methods, with an uncontrolled pre–post survey designed to measure impact on parent knowledge, attitudes, and behavior, supplemented by interviews, focus groups, and observation to document implementation and provide richer information on parent outcomes.
Data Collection
Data were collected from a variety of sources. Methods that are pertinent to results presented in this article include key informant interviews, parent surveys, and tracking of PA activities, and are described below. In addition, the evaluation team conducted focus groups with 29 parents; observed trainings, meetings, and community events; and monitored the program’s media presence.
Key Informant Interviews
Semistructured interviews were conducted with Vax Northwest partners and program staff, PAs, and site representatives at the end of each year. All respondents were asked about activities and accomplishments, perceptions of impact, challenges, lessons learned, and suggestions for program improvement. A total of 80 interviews (54 unique individuals) were conducted over 3 years.
Parent Survey
An online survey of parents was administered at each site prior to their beginning participation in the intervention, and then to all sites at the end of the second and third year of implementation. The instrument included several items from the Parent Attitudes About Childhood Vaccines survey (Opel et al., 2011) and questions developed by the evaluation team. The survey was implemented using Survey Monkey. PAs or site representatives sent out the web link to parents via e-mails, newsletters, or handouts and were asked to send weekly reminders during a 4- to 6-week time frame. A raffle entry incentive was offered to all those completing the survey. A total of 460 parent responses were received at baseline and 238 at the final follow-up. Response rates to the parent survey varied considerably across sites. The overall average response rate was 24% for the baseline and 13% for the follow-up. The median response rate was 36% at baseline and 24% at follow-up as larger sites tended to have lower response rates.
PA Activity Tracking
PA activities were tracked using a Google form embedded on the password-protected section of the program website. The form asked PAs to choose from an activity category, indicate how many minutes they spent on each activity, and briefly describe each activity. E-mails and Facebook reminders to complete activity logs were sent to PAs every 2 weeks.
Analysis
Qualitative data from key informant interviews were analyzed using a template approach (Crabtree & Miller, 1999). Atlas.ti, a qualitative analysis software package, was used to help code and manage the data (Atlas.ti Scientific Software Version 7.1.8; http://www.atlasti.com).
Quantitative data from the parent survey were analyzed with IBM SPSS Statistics for Windows (Version 22), using t tests and chi-square tests to test for pre–post changes. As mentioned above, new sites were enrolled in each of the 3 years and not all sites participated for the same length of time. The pre–post analysis compared each site’s baseline to the most recent follow-up data available for that site (when multiple follow-up surveys were conducted). We excluded survey responses that were largely incomplete and responses from sites with only 1 year of participation.
PA activity information, media mentions, and unique visitors to the program website were tallied using simple coding categories and reported using descriptive statistics. Facebook posts were analyzed to determine frequency of PA Facebook use and influence of immunization messaging (whether posts were “liked” or garnered positive comments, how often messages were shared).
This evaluation was reviewed by the group health research institute’s institutional review board.
Results
Parent Advocate Activities
Figure 1 summarizes PA activities. PAs used multiple modes to raise awareness and educate parents at their sites and in their communities, including social media advocacy, hosting events, distributing immunization-related materials, engaging in one-on-one conversations, and calculating and publicizing site immunization rates. About 80% of the PAs in Years 2 and 3 consistently reported their activities (data not available for Year 1), so information in Figure 1 likely underestimates the overall number of activities. The estimated average number of hours per month spent on reported activities, was 5.9 in Year 2 and 2.2 in Year 3. In Year 2, there was a marked increase in activities, including an increase in the number of events, engagement in social media (especially Facebook), and the creation of content for school and PTA newsletters. In Year 3, PAs were less involved with social media but distributed a greater number of vaccine-related materials. They also organized and participated in more events, including two large community events.

The Number and Type of Activities Conducted by the Parent Advocates in Each Year of the Program
Impact on Parent Vaccine-Related Knowledge, Attitudes, and Behaviors
Parent Vaccine-Related Knowledge and Awareness
Table 2 shows increased knowledge of local and state vaccination rates, familiarity with vaccine hesitancy, and understanding the concept of herd immunity1 on the pre and post surveys. The only statistically significant pre/post change was an increase in knowledge of the vaccination rates at their children’s child care or school. Qualitative interviews (data not shown) revealed that PAs and site representatives believed the program did a good job of raising awareness and knowledge about vaccine-related issues among the parents at their sites. They also agreed that parents at sites were more amenable to conversations about immunizations as a result of the program. PAs specifically commented on the importance of having a presence and being a source of accurate vaccine information for their peers. The primary ways knowledge was increased, according to PA and site representative report, was through posting materials, outreach events, peer-to-peer communications, and school newsletters.
Parents’ Vaccine Knowledge, Attitudes, and Beliefs
Percentage answering the question “What percentage of people in your community need to be vaccinated for everyone to be protected from disease?” with the response “almost all (95% to 100%)”: possible responses—“at least half (50%),” “somewhat more than half (51% to 65%),” “close to three quarters (66% to 80%),” and “almost all (95% to 100%). bNumber of valid responses for each question varied from 384 to 436 (pre) and 206 to 233 (post). cOpel et al. (2011).
.05< p < .10 testing for change from pre to post. **p ≤ .05 testing for change from pre to post.
Parent Attitudes and Beliefs About Vaccination
Table 2 shows pre/post results from 16 survey questions pertaining to vaccine-related attitudes and beliefs. Of these, 13 were trending in the desired direction, and five of those were statistically significant (p ≤ .05). Statistically significant improvements included an increased concern about other parents not vaccinating their children and a decrease in the belief that vaccines are given to children when they are too young. Table 3 shows there was also a statistically significant increase in the number of parents reporting confidence in vaccinating their child as a good decision, and a statistically significant decrease in the number of parents reporting themselves to be “vaccine–hesitant.” Qualitative interview data from the program’s first and second years revealed that PAs and site representatives were unsure whether attitudes had been changed among parents at their sites. In the third year of the program; however, more PAs believed that the program had influenced parents to think more positively about vaccinations, and fewer indicated they were unsure.
Parents’ Confidence in Vaccination, Safety Concerns, Vaccine Hesitancy, and Refusal/Exemption
Number of valid responses for each question varied from 452 to 454 (pre) and 228 to 232 (post). bRated on a 1-10 scale, where 1 = not at all confident and 10 = extremely confident. cOpel et al. (2011).
p ≤ .05 testing for change from pre to post.
Parent Vaccine-Related Behaviors
Survey results show a decrease in the number of parents who reported signing or planning to sign a vaccine exemption form; however, this result was not statistically significant (see Table 3).
Policy and Organizational Changes
The program also was designed to influence organizational- and community-level policies. Impact in this area included a statewide policy change and several organizational changes in policies and procedures. The Organization of Parent Education Programs (OPEP), a statewide organization that sets program standards for coordinators of cooperative preschools, was made aware, through the Immunity Community program, of the potential problem of not having complete and accurate vaccination information on students in cooperative preschools. As a result, OPEP revised its risk management manual to designate individuals at each cooperative to collect immunization records, calculate immunization rates, and maintain records. The revised manual also provides guidance on reporting disease outbreaks and controlling the spread of disease. This policy change has the potential to affect all cooperative preschools in Washington, which serve over 10,000 families. Program staff and PAs were instrumental in making OPEP leadership aware of this issue and supporting the revision of the manual.
At the organizational level, a majority of sites reported taking action to increase the tracking of immunizations, and calculating and publicizing rates. Rates were publicized mostly at the site/school level, though a school nurse did report results at the district level, and the Washington State Public Health Association newsletter reported on the program. Some sites made sustainable changes to procedures and processes, including the development of a spreadsheet for calculation of rates, addition of a computer system change to flag missing immunization data, and increased attention to parental education regarding the topic of immunizations.
Discussion
Our evaluation of the innovative Immunity Community program shows it is possible to engage, train, and support parent volunteers to be vaccine advocates in their communities. PAs implemented a multifaceted campaign including engaging in social media advocacy, hosting school and community events, distributing immunization-related materials, engaging in one-on-one conversations, and working to calculate and publicize site immunization rates. The program was also instrumental in a statewide policy change that will influence thousands of Washington State families. Finally, the evaluation demonstrated statistically significant increases in parental knowledge, and changes in vaccine-related attitudes. The Immunity Community program is the first intervention we are aware of that examines the impact of a range of approaches implemented by mobilizing volunteer PAs.
A key question that needs to be addressed is whether the program can be implemented and sustained with fewer resources. The current program model included a program staff member working at 80% of a full-time position for a 16-month program cycle. This person was the primary staff member responsible for recruiting, training, and providing technical assistance, as well as leading the process of materials development and updates. It is unclear how effective the program would be in contexts with different vaccination norms and values, different resources available to support vaccination, and/or different state and county vaccination policies. The issue of how to obtain better vaccination rate data from child care organizations is also a consideration.
The study had several limitations. The online parent survey was cross-sectional. While some of the same parents may have completed both the pre and the post surveys, we do not know how many respondents were matched from pre to post. Response rates to the survey were low, especially for the postprogram samples, where there were more challenges in working with staff to administer surveys near the end of the school year. Low response rates overall make the results potentially biased and less generalizable. The lower post survey response rates may result in an anticonservative bias if those responding have more favorable attitudes toward vaccinations and are more familiar with the program. In addition, the number of respondents at each site was too small to analyze for a dose–response relationship—that is, assessing whether schools with greater exposure to the program (greater number of years participating) showed a greater effect on key outcome measures. Limited and incomplete vaccination data at the participating sites meant we could not examine the ultimate outcome of increased vaccination rates. Not all sites are required by law to report immunization statistics to the Department of Health, so the reliability of data and our ability to track data systematically and longitudinally are limited at this time. We acknowledge limitations in our ability to assign causality to the intervention because of survey design and small sample. We also are not able to generalize findings broadly due to the small sample size and the focus on relatively affluent urban and suburban communities.
Conclusion
This study demonstrates the promise of a new and innovative approach to reducing vaccine hesitancy: engaging parent volunteers to be advocates in their own communities. The approach builds on findings that parents’ social networks are a strong predictor of vaccine acceptance (Brunson, 2013) and that likeable, trustworthy messengers are positively received (Jarrett et al., 2015; Opel et al., 2009). Results show statistically significant shifts in certain vaccine-related knowledge and attitudes among parents in prioritized communities. In order to test the long term impacts of such a program, like decreased vaccine exemption rates, researchers must identify ways to accurately calculate immunization rates at participating sites, especially sites where there is no requirement to report to the state Department of Health. Future research is planned to assess whether program impact can be maintained with fewer resources.
Footnotes
Acknowledgements
We would like to thank the parent advocates, site leaders, members of the community advisory board, our funders, and our partners in Vax Northwest, including the Washington State Department of Health, Seattle Children’s Hospital, BestStart Washington, Group Health Foundation, and WithinReach, who implemented the program.
