Development of an Interactive Social Media Tool for Parents With Concerns About Vaccines

Abstract

Objective.

Describe a process for designing, building, and evaluating a theory-driven social media intervention tool to help reduce parental concerns about vaccination.

Method.

We developed an interactive web-based tool using quantitative and qualitative methods (e.g., survey, focus groups, individual interviews, and usability testing).

Results.

Survey results suggested that social media may represent an effective intervention tool to help parents make informed decisions about vaccination for their children. Focus groups and interviews revealed four main themes for development of the tool: Parents wanted information describing both benefits and risks of vaccination, transparency of sources of information, moderation of the tool by an expert, and ethnic and racial diversity in the visual display of people. Usability testing showed that parents were satisfied with the usability of the tool but had difficulty with performing some of the informational searches. Based on focus groups, interviews, and usability evaluations, we made additional revisions to the tool’s content, design, functionality, and overall look and feel.

Conclusion.

Engaging parents at all stages of development is critical when designing a tool to address concerns about childhood vaccines. Although this can be both resource- and time-intensive, the redesigned tool is more likely to be accepted and used by parents. Next steps involve a formal evaluation through a randomized trial.

Parental vaccine hesitancy is a growing public health concern (Freed, Clark, Butchart, Singer, & Davis, 2010; Kennedy, Basket, & Sheedy, 2011; Omer, Salmon, Orenstein, deHart, & Halsey, 2009). An estimated 13% of parents choose to either delay or refuse vaccines for their children (Glanz, Newcomer, et al., 2013). When compared with parents who fully vaccinate their children, parents who refuse or delay childhood vaccines are more likely to use the Internet for vaccine information and less likely to trust vaccine information from health care professionals, health departments, and the government (Downs, de Bruin, & Fischhoff, 2008; Freed, Clark, Butchart, Singer, & Davis, 2011; Jones et al., 2012; Kennedy, LaVail, Nowak, Basket, & Landry, 2011). To combat this lack of trust in vaccine information (Glanz, Wagner, et al., 2013), providers describe having to spend increasing amounts of time discussing vaccine-related topics during routine well-child visits, leaving less time to cover other important topics such as nutrition, development, behavior, and injury prevention (Diekema & the Committee on Bioethics, 2005; Wilson & Keelan, 2013). In some instances, pediatricians dismiss families from their practices when parents choose not to vaccinate, potentially further undermining public safety (Flanagan-Klygis, Sharp, & Frader, 2005; Larson, Cooper, Eskola, Katz, & Ratzan, 2006).

One promising approach to address this emerging public health concern is through the use of social media (Betsch, Renkewitz, Betsch, & Ulshöfer, 2010; Korda & Itani, 2013), defined as web-based technologies that facilitate social interaction and enable users to share content (Kaplan & Haenlein, 2010). Social media represents a relatively new medium in which risk communication messages can be delivered to parents. Through social media, parents can be empowered, informed, and engaged (Chou, Hunt, Beckjord, Moser, & Hesse, 2009; Eysenbach, 2008; Hawn, 2009). Over the past 5 years, there has been a proliferation of vaccine-focused websites that use social media technologies (Wilson & Keelan, 2013). Many of these web-based resources, however, are using these technologies to disseminate erroneous information, discredit the medical community, and create fear (Kata, 2012). Such free-form technologies are also prone to misinformation and vandalism. In many instances, the online forums are not moderated by experts, and the sources of health information are anonymous (Sudau et al., 2014). The proliferation of these antivaccine websites highlights the challenges of using social media for vaccine and health information in general. In addition, despite its wide use, the evidence base for using social media to positively influence vaccine-related health behavior is lacking (Chou, Prestin, Lyons, & Wen, 2013; Witteman & Zikmund-Fisher, 2012). Clearly, interventions designed to evaluate the effectiveness of using social media to reduce parental vaccination concerns are needed (Sadaf, Richards, Glanz, Salmon, & Omer, 2013).

For this study, we describe a process for designing, building, and evaluating a theory-driven social media intervention tool to help reduce parental concerns about childhood vaccination. Our objective was to develop the tool so that it could be formally evaluated with a randomized intervention trial. First, we describe our process for building and designing the information content and architecture of the web-based tool. We then provide results from individual interviews, focus groups, and usability testing that were used to further develop and refine the design and functionality of the intervention tool.

Method

Theoretical Basis for Using Social Media to Address Vaccine Concerns

Our objective was to create a web-based tool that provides evidence-based information in an interactive environment where parents can contribute content and discuss concerns with other parents and vaccine experts. The conceptual framework for this dynamic, online environment has been described as a multidirectional communication model (Thackeray & Neiger, 2009). This model represents a social marketing strategy with three primary components. The first component is a conventional top-down process in which researchers and web developers create content and deliver messages to users. The second component is a bottom-up approach where users create information and contribute messages. The third component is a horizontal side-to-side process in which users share information with other users. In this multidirectional model, users receive, develop, share, and actively request information using various social media technologies. Users become active, rather than passive, participants in the communication process.

Study Setting and Participants

The study was conducted between April 2010 and July 2013 at Kaiser Permanente Colorado (KPCO), a managed care organization with approximately 500,000 members in the metropolitan Denver area. KPCO uses an electronic health record (EHR), which captures demographic, medical encounter, and vaccination data on all members. Participants of the study included pregnant mothers and parents of children younger than 4 years. The local institutional review board approved the study, and written consent was obtained from participants.

Cross-Sectional Survey to Assess Potential Use and Acceptability of a Vaccine Social Media Tool

To gauge interest from our target population of parents, we developed, pilot-tested, and mailed a survey to assess their hypothetical trust in and use of a social media web-based tool for vaccine and health information. The questions on trust in and use of social media were part of a larger survey on the vaccine decision-making process (available as online supplementary content at heb.sagepub.com/supplemental). The details of how the survey was developed and pilot-tested have been reported previously (Glanz, Wagner, et al., 2013).

Using the EHR, we created a sampling frame of parents who either accepted, delayed, or refused vaccinations for their children. We first identified all children enrolled in the KPCO health plan between the ages of 4 and 24 months. A manual medical record review was then conducted on the children to determine if parents had delayed or refused vaccination for personal, nonmedical reasons. Surveys were subsequently sent by mail to a random sample of parents who accepted vaccines (n = 500), all parents who delayed vaccines (n = 227), and all parents who refused vaccines (n = 127). Parents who delay vaccines, in particular, are the primary target population for the intervention because these parents are continuously weighing the risks and benefits of vaccines in their child’s first 2 years of life (Glanz, Wagner, et al., 2013). Surveys were sent up to two times to each participant, and no incentive was provided. A subanalysis was conducted on the demographics of the survey nonresponders to assess the representativeness of the final study population.

Building and Designing the Vaccine Social Media Intervention Tool

The tool was developed in two phases. The overall development process was informed by the adapted mental models approach, as described by Morgan, Fischhoff, Bostrom, and Atman (2002). In Phase 1, we built the initial prototype, which included developing the content, design, layout, and information architecture. In Phase 2, we evaluated the prototype by conducting focus groups, one-on-one interviews, and usability testing with a representative group of parent end users. Results from this evaluation were used to further develop and refine the content and design of the tool.

Phase 1: Development of Content, Social Media Applications, Design, and Information Architecture

Content.

During the first 12 months of the study, our multidisciplinary research team met weekly to develop the content for the tool. Our team included pediatricians, vaccine safety epidemiologists, behavioral scientists, risk communication specialists, web developers, and information technology specialists. The content was guided by our prior qualitative work on vaccine hesitancy, as well as the theory of planned behavior and the health belief model, which are theoretical frameworks that have been used to describe vaccine-related health behaviors (Brewer et al., 2007; Ernsting, Schwarzer, Lippke, & Schneider, 2013). The factual information focused on childhood vaccine safety, vaccine effectiveness, vaccine preventable diseases, vaccination laws, recommended immunization schedules, and vaccine ingredients. Our primary goal was to present the information in a way that helps establish trust and credibility with parents. Although the content is provaccine, we used a conversational tone that conveys openness and empathy. We also included descriptive information about our research team to convey dedication, commitment, competence, and expertise, all of which can be further emphasized through interactive social media components as described below (Cohrssen & Covello, 1989; Covello, 2003).

Interactive and social media applications.

We incorporated the following technologies to facilitate interaction with parents: a blog, a discussion forum, a chat room, and a portal for parents to ask questions of our team of experts. We anticipate that these interactive forums will empower parents to express their vaccine concerns, to which we can respond in real time. We developed an operational plan to regularly update the tool’s content, quickly respond to questions, and actively participate in the interactive forums. At minimum, the interactive features of the website will be monitored for new activity twice a day. To keep the content current, we will conduct weekly searches on the Internet and scientific databases, such as PubMed and Google Scholar, for new media stories and publications relevant to parents’ concerns about vaccines. Importantly, when communicating with parents, our tone will be respectful, we will treat parents as experts, and we will explicitly acknowledge parents’ concerns as valid. Our goal is to provide reliable information, resources, and support to a community of parents and experts.

A majority of the tool’s updated content and social interaction will be conducted through the blog. Blog content can be text or multimedia posts, such as photos, video, and audio (i.e., podcasts). Although the primary content will be developed and posted by the research staff, parents will be encouraged to post comments, questions, and opinions in an open conversational format. To sustain engagement with users, we plan to update blog entries twice per month. We will also circulate a newsletter to inform parents about new content and other updates.

The online chat room and discussion forum will enable direct communication between parents and the research team. The objective is to facilitate in-depth conversation on vaccine-related topics. Both parents and experts will have the opportunity to propose and discuss topics with other experts and parents. For the monthly chat room, topics will be scheduled in advance. An expert member of the research team will initiate discussion, answer questions, and ensure that users stay on topic. Chat discussions will be conducted in real time as open question and answer forums. In contrast, the discussion forum will allow parents to hold in-depth conversations as posted messages. Each posted message will be organized by topic, known as “threads.” Each thread will be posted to the site’s forum archive where it can be searched and viewed.

The final interactive application is a simple portal through which parents can privately ask the research team questions. The portal has two purposes. One, it allows parents who would rather not interact publicly to interact privately with the research team. Two, it allows parents who are reluctant to engage with the social media components to first use this portal to build trust and rapport with the research team and ultimately ease into a more public role in the future.

Moderation plan for the interactive vaccine resource.

We developed a detailed, institutional review board-approved social media protocol for monitoring and managing all social interaction on the website (available as online supplementary content at heb.sagepub.com/supplemental). Activity will be monitored on a daily basis to identify and filter abusive or threatening language, bullying, and disclosure of personal identifying health information. If users break rules, they will be placed on “moderation” and potentially removed from the website as described in the Terms of Use section on the tool. In addition, we will use WordPress (Version 3.5.1; Word-press.com) open-source software to automatically filter the discussion board, blog, and chat room for abusive language.

Design, layout and information architecture.

Over a 6-month period, our team worked iteratively with a web programmer and a visual designer. The web programmer used WordPress as the architectural platform for the tool and its interactive components. WordPress is a flexible web publishing system that allows nonprogrammers to make changes and updates to the tool. In collaboration with the visual designer, we developed the aesthetic look of our interactive web-based tool, including the colors, imagery, fonts, and placement of content on the pages. The designer applied graphic design theory to manage the spacing, proportion, balance, grouping, flow, and topography of the elements on the various pages of the tool. The designer was also trained in graphics software (Adobe Photoshop, Illustrator, InDesign [Adobe Systems Incorporated]) and coding languages (HyperText Markup Language 5 and Cascading Style Sheets).

After deciding on the design and general layout, we focused on information architecture. As a team, we organized the expert-driven content to try to anticipate users’ childhood vaccine information needs and search behavior. Specifically, we developed a content schema of subcategories and links to simplify the process by which people navigate and find information. Once our schema was implemented and the tool was developed, we evaluated its functionality through focus groups, interviews, and usability testing.

Phase 2: Focus Groups, Interviews, and Usability Testing.

We conducted a two-step approach to evaluate the tool’s functionality and usability. The evaluation methods are described below. The participants for these focus groups, interviews, and usability testing were selected in a similar manner as described for the survey.

Step 1: Focus groups and interviews.

Early in the development process, we conducted two focus groups and 16 one-to-one interviews with a total of 23 participants across a range of vaccine behaviors. Of the 23 participants, 3 were pregnant, 12 delayed vaccines, 6 accepted vaccines, and 2 refused all vaccines for their children. The objective was to elicit feedback on the overall appeal, design, layout, functionality, and perceived usefulness of the various web-based components. Interviews were conducted as brainstorming sessions to generate ideas on how the tool’s content, layout, and design could be improved. We also asked parents how they had made their decision about vaccinating their child, what interactive features they would use on a childhood vaccine information website, and what additional features they might want to access on the tool.

Each focus group and interview was audio recorded and transcribed by an independent transcriptionist. The qualitative data were analyzed using a team-based inductive approach (Ryan & Bernard, 2003). Specifically, three researchers (JS, NW, JG) used a constant comparative analytic method, in which the transcribed data were coded and categorized in a multistep process (Glaser, 1965). The analysis was aided by Atlas.ti 6.2 qualitative analytic software (Scientific Software Development GmbH, Berlin, Germany).

Separate interviews were conducted with four parents as a card-sorting exercise to help us understand how well parents could navigate the content of the tool. Each parent was given index cards covering the various topics covered on the tool. Parents were then asked to organize the index cards into categories that were easy to navigate.

Step 2: Usability testing.

We conducted seven 60-minute usability sessions with parents to evaluate the functionality of a prototype tool. This was a new sample of parents that did not participate in the focus groups. Usability is defined as how effective, efficient, and satisfied users are when performing specified website tasks (International Standards Organization, 1994; Nielsen, 1993). In these sessions, parents were instructed to perform specific tasks and complete surveys. Tasks included logging on to the tool, finding specific vaccine information, posting comments to the blog, and starting a discussion forum. As parents performed tasks, an interviewer encouraged them to “think aloud” and describe what they liked about the tool, what they found difficult or frustrating, and what they thought could be improved. The interviewer took detailed notes, and usability was assessed quantitatively by the following four metrics: (a) completion rate, (b) error free rates, (c) time on task completion, and (d) user satisfaction.

At the end of each session, user satisfaction was measured with the System Usability Scale (SUS), a reliable and widely used survey instrument for quantifying subjective assessments of usability (Brooke, 1996; Bangor, Kortum, & Miller, 2008; Sauro, 2011). The SUS survey comprises 10 questions measured on a 5-point Likert-type scale ranging from strongly agree to strongly disagree. Each SUS survey provides a single score ranging from 0 to 100; scores above 70 are considered passable.

Results

Cross-Sectional Survey Results

The overall survey response rate was 52% (n = 443), with a 37% response rate for parents who refused vaccines, 60% for the parents who delayed vaccines, and 52% for parents who accepted vaccines for their child. Table 1 shows the descriptive statistics for the 12 questions. Approximately 50% of parents in all three vaccine behavior groups reported that they would use the web-based tool often. More than 60% of parents who delay or accept vaccines reported that they would trust the information about vaccines presented on the tool. Regardless of their vaccine decisions, a high proportion of parents reported that they would use the tool to ask questions, to receive current vaccine information, and to review the childhood vaccination schedule. Approximately 50% of parents who delay vaccines also said they would discuss their experiences and vaccine concerns using the web-based tool. In the subanalysis comparing survey responders to nonresponders, there were no significant differences in age, income, or home clinic (data not shown). Overall, these results suggested that social media may represent an effective intervention tool to help parents make informed vaccination decisions for their children.

Table 1.

Trust in and Use of Web-Based Information by Vaccine Group.

	Proportion of parents (%)
Question/statementa	Refused vaccines (n = 47)	Delayed vaccines (n = 136)	Accepted vaccines (n = 260)	p
I would trust the website in providing vaccine information.  Strongly agree/somewhat agree	36.17	61.03	68.46	.0001
I would trust the website if it was hosted by my health care provider.  Strongly agree/somewhat agree	36.17	53.68	83.85	<.0001
I would be too busy to use the website.  Strongly agree/somewhat agree	12.77	11.76	20.38	.07
I do not have home computer or internet access.  Strongly agree/somewhat agree	6.38	0.00	2.31	.02
I would use the website often.  Strongly agree/somewhat agree	44.68	65.44	48.85	.003
I would use the website to ask questions to vaccine experts.  Strongly agree/somewhat agree	51.06	75.00	68.08	.009
I would be likely to post something on the website.  Strongly agree/somewhat agree	34.04	39.71	20.38	.0001
I would be likely to discuss experiences with other parents.  Strongly agree/somewhat agree	46.81	50.00	26.92	<.0001
I would be likely to express vaccine concerns on the website.  Strongly agree/somewhat agree	48.94	52.94	30.00	<.0001
I would use the website to receive current vaccine news stories.  Strongly agree/somewhat agree	74.47	82.35	71.15	.05
I would use the website to see what vaccinations my child will be receiving at the next provider visit.  Strongly agree/somewhat agree	46.81	81.62	84.62	<.0001

Note. Proportion of parents endorsing strongly agree/somewhat agree to each question is displayed.

^a5-point Likert-type question scale (strongly disagree, somewhat disagree, neither agree or disagree, somewhat agree, and strongly agree).

Focus Groups and Interviews

Of the 23 parents who participated in the focus groups and one-to-one interviews, 1 was male and 22 were female. Seventeen percent of the parents were Hispanic, and 94% were White. The average age of parents was 34.6 years. Parents evaluated the design, content, and specific features of the web-based tool. Four main themes emerged from the qualitative analysis. The first theme was balance. Parents were adamant that we present information on risks as well as benefits of childhood vaccination. In addition, parents requested more information on alternative vaccination schedules and vaccine ingredients. Transparency was the second theme. Parents thought it was important that we prominently display our source of funding for the study and provide citations for all sources of information. The third theme was moderation. Parents stressed the need for all the interactive forums to be carefully monitored by experts such as nurses, doctors, researchers, and complementary alternative medicine providers. Last, diversity emerged as a theme. Parents requested that we include more pictures of fathers and increase racial and ethnic representation on the website. In addition to these themes, parents suggested that we increase the font, improve the color contrast, and make changes to the overall layout of information.

Usability Testing

The seven parents were all female, married, and White. Twenty-nine percent were Hispanic; parents were well educated and had high household incomes. Most parents used social media and looked for health information on the Internet. Between 71% and 100% of the parents were able to complete specific tasks and between 14% and 100% of the tasks were completed without error (Table 2). The longest mean time to completion and lowest completion rate involved finding information about hepatitis B vaccine. The two most difficult tasks for parents to complete without error were posting to the discussion board and finding specific information about hepatitis B vaccination, with only one parent completing these tasks without error.

Table 2.

Usability Testing Task Results of Parents Using the Prototype Web-Based Tool (n = 7).

Task	Mean time to complete task in seconds	Task completion, n (%)	Error-free task completion, n (%)
Interactive vaccine schedule taska	74	7 (100.0)	7 (100.0)
Post a comment to the discussion forumb	125	7 (100.0)	1 (14.3)
Find information about hepatitis B vaccinec	146	5 (71.4)	1 (14.3)
Determine the next live chat with an expert sessiond	46	7 (100.0)	5 (71.4)
Find the most recent news story and add a commente	46	7 (100.0)	7 (100.0)

^aUsing your child’s date of birth, determine what vaccines are due at your child’s next pediatric clinic visit.

^bFind the discussion board and post a question asking other parents about breastfeeding to calm a baby during receipt of vaccines.

^cLocate information about why the hepatitis B vaccine is given at birth.

^dDetermine when the next live chat with an expert will be.

^eFind the most recent news story and add a comment about additional information on the topic.

Parents also provided qualitative feedback about the tool’s functionality. Overall, parents commented that the tool appeared “clean” and easy to read. However, some parents noted that the graphics were too basic and that the colors were too light. Parents also did not like having to scroll to the bottom of the page to find information and felt the tone of the messaging was too scientific. Last, some parents commented that the link structure of the website was inconsistent and difficult to navigate.

Total scores from the SUS survey ranged from 67.5 to 95.0. The overall mean usability score was 85.7 out of 100, which was above the threshold cut point of 70 (Table 3). Results of the card-sorting exercise suggested changes in the layout and categorization of the information on the tool. For example, a parent noted, “Website rules make me think of what I’m being told I cannot do” and suggested the heading “ground rules” or “website expectations.”

Table 3.

System Usability Scale (SUS) Responses of Parents About the Prototype Web-Based Tool (n = 7).

Question	Mean scorea	Minimum scorea	Maximum scorea
I think that I would like to use this website frequently.	2.43	2	3
I found the website unnecessarily complex.	3.71	3	4
I thought the website was easy to use.	3.71	3	4
I think that I would need the support of a technical person to be able to use this website.	4.00	4	4
I found the various functions in this website were well integrated.	3.00	2	4
I thought there was too much inconsistency in this website.	3.86	3	4
I would imagine that most people would learn to use this website very quickly.	3.14	2	4
I found the website very cumbersome to use.	3.71	3	4
I felt very confident using the website.	3.14	1	4
I needed to learn a lot of things before I could get going with this website.	3.57	2	4
Overall converted scoreb	85.7	67.5	95.0

Note. Negatively stated items were reverse scored.

^aSurvey response scale: strongly disagree = 0; disagree = 1; not sure = 2; agree = 3; strongly agree = 4.

^bScale converted from 0-40 to 0-100.

New Design, Layout, and Information Architecture

Based on the focus groups, interviews, and usability evaluations, we made additional revisions to the tool’s content, navigation, and overall look and feel.

We rewrote and organized several of the informational sections to emphasize risks as well as benefits of childhood vaccination. We added sections on adult, adolescent, maternal, and travel immunizations in response to emerging public health concerns and changing immunization policies. A Vaccines in the News section was added to feature current media articles on vaccines. For the blog, we developed a series of podcasts presented by a pediatrician researcher who also practices within the integrated health care system. These audio podcasts are approximately 3 to 4 minutes in length and cover topics ranging from discussing vaccines with parents at well-baby visits to how vaccines are chosen and financed within the health care system.

In Figure 1, we show the prototype tool prior to the usability evaluation and the finalized tool after incorporating parent feedback. We changed the font type and size, used a different color palate, and created a logo. To the home page, we added a sliding set of pictures (“slider”) to highlight the different topics available on the tool. We shortened the information pages and changed the overall theme to a tabbed design that provides access to vaccine information areas featured in the slider (Figure 2). This design addressed the findings of the usability testing by allowing users to easily navigate the content without having to scroll up and down the page. In addition, to address parents’ difficulty posting to the discussion board, we redesigned the appearance of the discussion board by changing the font, by adding icons with text summarizing each topic when hovered over, and by relocating new discussion topics to a side menu that allows for easier navigation. Last, to facilitate interaction, we added interactive buttons to each page that link directly to the discussion, blog, and “ask an expert” forums (Figure 1).

Figure 1.

Prototype welcome page changes: (A) Premodification; (B) Postmodification: Navigation buttons added, color and font changed, logo added, graphics changed, and page shortened.

Figure 2.

Vaccine ingredients spreadsheet information changes: (A) Premodification; (B) Postmodification: Change in visual layout, adding tabbed buttons for each vaccine (shown is DTaP vaccine).

Discussion

In this study, we describe a theory-driven approach for designing, building, and evaluating an interactive web-based tool for parents with concerns and questions about childhood vaccines. Conducting this study within a large integrated health system allowed us to use a well-developed EHR system to prospectively identify pregnant women and parents with a range of vaccine beliefs and behaviors. We used quantitative and qualitative methods to engage parents, who represent prospective end users of the tool. We now believe our newly developed web-based tool is ready to be formally evaluated using a randomized trial study design.

Engaging the target population is critical when developing a tool to address complex preventive health behaviors. In our population-based survey, we learned broadly that concerned parents who delay childhood vaccines would use and benefit from an interactive web-based tool embedded within the health care delivery system. After designing a prototype, we reengaged parents from this target population through focus groups and interviews to help us further develop and refine the tool. This engagement process revealed several important areas that needed to be addressed, including the look and feel of the tool, navigation, balanced content, transparency, and moderation. Collecting these qualitative and quantitative data increases the likelihood that the final iteration of the tool will be a useful resource for our target population.

It is important to note that social media technologies are not in and of themselves interventions that can elicit behavior change. Rather, they are technologies that facilitate engagement among individuals who have a propensity to interact with one another, such as pregnant women and parents of young children (Duggan & Brenner, 2013; Duggan & Smith, 2013). When designing a social media intervention, several important factors should be considered. The tool should contain a base of information relevant to the target population, the health messages should be crafted by individuals with experience in risk communication, and the interactive components should be carefully moderated by topic area experts. Additionally, trust must be established between the target population and the research team using a variety of methods, including presenting information on both risks and benefits, acknowledging participants’ concerns, avoiding scientific jargon, being transparent about funding sources, carefully referencing all sources of health information, and quickly responding to questions with tailored, personalized information. Adhering to these criteria will increase the likelihood that the tool will have a positive impact on health beliefs, attitudes, and behaviors.

Our process for developing an interactive web-based tool has been resource-intensive and time-consuming. This represents a potential challenge with research focused on technology-based interventions. First, a long development period implies that it may take several years before the tool is available to practitioners who have an immediate need for the intervention. Second, web-based technologies can significantly evolve over the course of a study period, implying that a tool may be outdated by the time it is ready to be implemented. It is therefore imperative that future research focus on developing efficient methods for developing technology-based interventions so that they can be quickly implemented into practice.

This study has several potential limitations. First, the response rate for our survey ranged from 37% to 60%. Although this may have limited the representativeness of the study population, our subanalysis comparing survey responders to nonresponders showed that these groups did not differ significantly by age, income, or home clinic. Second, the participants in the usability testing tended to be White and highly educated. Although this may have affected the generalizability of our results, it is important to note that the target population for our tool—parents with concerns about vaccinations—is strongly associated with these sociodemographic characteristics (Dempsey et al., 2011; Salmon et al., 2005).

Although we designed this tool to meet the informational needs of parents who delay vaccines, the tool would be available in practice to all parents regardless of their vaccination attitudes and decisions. In our survey work, we found that more than 60% of parents who delay or accept vaccines indicate that they would trust information from this tool. However, only 36% of parents who refuse all childhood vaccines report that they would trust the information from this tool. This is consistent with prior survey research showing that parents who refuse vaccines for their child are highly distrustful of the government, medical community, and pharmaceutical industry (Freed et al., 2010). Researchers, as a consequence, face significant challenges when designing interventions that target parents with strong concerns about childhood vaccination.

Our social media tool was not specifically designed to address the concerns of ethnically diverse and non–English-proficient parents. Social media applications, such as this one, raise a unique set of concerns around the applicability to non–English-proficient individuals. A culturally concordant tool for each non–English-speaking population would need to address the specific vaccine concerns of that cultural group. In addition, the interactive features—such as chat rooms, discussion forums, and blogs—would require language-concordant experts to provide input and moderate the various forums. We believe future research should focus on developing and testing interactive tools designed to address the vaccination concerns and barriers of diverse populations.

Although social media represents a promising new approach to engage parents about childhood vaccines, data on its effectiveness are lacking (Chou et al., 2013). It is currently not known if these interactive technologies can be used to positively affect parents’ vaccination knowledge, attitudes, beliefs, and behaviors. For these reasons, we believe that before our tool can be implemented into practice, it should be formally evaluated with a randomized intervention trial within an integrated health care system. The trial should be designed to target vaccine-hesitant parents and compare the use of interactive technologies with usual well-child care. Throughout the trial, the tool would be available to parents online, which they could access with their own electronic devices (smart phone, iPad, laptop, desktop computer) at their convenience. To capture clinically relevant outcomes, survey instruments could be used to measure vaccine knowledge, attitudes, and beliefs at multiple time points over their children’s first 2 years of life, and vaccination data could be extracted directly from the EHR. Analyzing such outcome data would provide direct evidence of the intervention tool’s effectiveness. If effective, we believe our newly developed social media tool would represent an innovative and broadly applicable resource to reduce parental vaccination concerns and improve childhood immunization rates over time.