Abstract
Few national studies have asked parents how to improve the childhood vaccination process. We surveyed a nationally representative online panel of parents on how to improve this process, rationales for nonstandard approaches, and alternatives to the standard schedule. Twelve percent of the 1222 respondents reported using nonstandard approaches: 3.2% used a specific schedule, 6.0% had no specific schedule, and 2.5% declined all vaccinations. The most common rationales were that too many vaccines are given at once, and discomfort with vaccine ingredients. Regarding how to improve the process, parents using the standard schedule most often said nothing could be improved (51%), or better vaccine information (22%). Those using nonstandard approaches most often would have liked more choice (40%) or better vaccine information (26%). Parents’ experiences with the vaccination process could be improved by offering information prior to visits, giving more information about side effects, and allowing more flexibility about vaccine scheduling.
Background
Recent epidemics of measles and whooping cough in the United States have drawn attention to the risks of parental refusal of childhood vaccines.1,2 While only about 2% of parents refuse all vaccines, more than 1 in 10 refuse or delay specific vaccines.3-6 Even among parents who have used the standard vaccination schedule, many express concern that children are receiving too many vaccines.4,7
The process of receiving or declining childhood vaccinations affects virtually every family in the nation multiple times during a child’s early years. Clinicians as well as parents feel there is room for improvement. A recent survey found that virtually all physicians who treat children receive requests to spread out vaccine doses more than the standard schedule, but few feel their strategies for responding are effective. 8
To improve the vaccination process, we need a deeper understanding of parents’ perspectives on what factors to prioritize, as well as their rationales for hesitancy about multiple vaccines. Understanding parents’ perspectives is challenging because varying beliefs and practices exist. 9 While many national surveys have described parents’ concerns about vaccines,3,4,7,10 few have directly asked their opinions on what factors are most important to improve in the overall process.
This study’s objective was to assist clinicians, policy makers, and parents by identifying priorities for improving the vaccination process. Our aims were to describe parents’ (1) perspectives on how to improve the process, (2) rationales for use of nonstandard vaccination approaches, and (3) reactions to hypothetical alternatives to the standard schedule.
Methods
Design and Participants
We conducted a cross-sectional survey of a nationally representative sample of US parents of children aged 7 to 48 months. We developed the survey with input from our project’s National Stakeholder Committee, and from a qualitative study we completed. 11 We conducted cognitive pretests with 4 parents to refine the draft survey.
We administered the final survey in September 2014 to members of KnowledgePanel, an online panel representative of the US population that is maintained by the GfK Group. Panel members are recruited by random selection of telephone numbers and addresses. Persons who agree to participate are provided with a laptop computer and Internet connection if needed. This panel has served as the sampling frame for other national studies of parents’ preferences about child health issues.3,12
Survey Content
The survey was composed of closed-ended questions, most of which were forced-choice to allow for targeted follow-up questions. The first question asked each parent to indicate his or her approach to the vaccine schedule via 6 mutually exclusive response options. The question noted that a schedule “specifies the vaccines a child should get at each visit.” The response options were: “I have generally or always followed my doctor’s recommendations,” “I have followed a different schedule that specifies which vaccines my child gets and when he gets them,” “I don’t follow a specific schedule but I have a general approach in mind for choosing which vaccines my child will get and when,” I don’t have a predetermined plan for vaccines—I decide about each vaccine at each visit,” “My child has not had any vaccinations,” and “I have done something else.”
We asked each parent a series of follow-up questions that elicited details about their vaccination approach, and asked them to choose the most important reason they chose this approach. We asked questions on how to improve the experience of vaccinations for children. We posed 2 questions describing hypothetical alternative vaccination schedules and asked parents to rate whether they would like these schedules for their own child (see the appendix). The survey instrument, which took a median of 15 minutes to complete, is available from the corresponding author on request. The Institutional Review Boards of Kaiser Permanente Northern California and the Harvard T.H. Chan School of Public Health approved this study.
Statistical Analyses
We conducted bivariate analyses of the association between schedule approaches and parent characteristics using the chi-square test for categorical variables and the Kruskal-Wallis test for ordinal variables. We created multivariate models to test the hypothesis that parents using nonstandard approaches were more likely to be white and to have higher socioeconomic status. These models used forced-entry logistic regression, with schedule approach as the dependent variable. Except where noted, we used probability sampling weights provided by the survey vendor to weight the sample to represent the US population of parents with children 0 to <5 years of age.
Results
Participants
We invited 3558 people from the survey panel who had previously reported having a child aged 0 to 5 years to participate. Of these, 1891 (53.1%) returned initial responses; 1222 (64.6%) of these were eligible based on having a child within the designated age range and completed the survey (estimated 53.1% participation rate). Participants were diverse in race/ethnicity and socioeconomic status (Table 1). Most respondents (61%) were 30 to 45 years old, and most (79%) had just one child who was 7 to 48 months old. Compared with nonrespondents, respondents were more likely to be male, older, white, and to have higher educational attainment and higher income (differences are significant at P < .01). To reflect national patterns, we weighted all subsequent analyses and tables to reflect the population eligible for the survey.
Demographic Characteristics of Study Parents and Comparisons Between Subgroups Using the Standard Vaccination Schedule or Alternative Approaches. a
Respondents had to have reported having at least one child between 7 and 48 months of age.
Six respondents who had missing or “other” responses to the question on what vaccination approach they used were excluded.
P values are from the Kruskal-Wallis test for age, education, income, and number of children; and on the chi-square test for sex. NA: Small cell counts precluded calculation of valid chi-square for race/ethnicity. See the text for results of multivariate modeling of these factors.
Among all children in the household, regardless of their age.
Prevalence of Nonstandard Vaccination Approaches
Of the study sample, 142 parents (12%) reported using a nonstandard approach to vaccine scheduling (Table 2). These represented 4 groups: 39 (3.2%) used a specific schedule, 39 (3.2%) used a general approach but not a specific schedule, 34 (2.8%) had no predetermined plan, and 30 (2.5%) declined all vaccinations. Of the 39 parents using a different specific schedule, 6 reported using a schedule recommended by Dr Robert Sears, 13 and 23 were following a different schedule designed to limit the number of shots at each visit.
Use of Standard and Alternative Vaccination Approaches.
For each question, respondents were asked to select the one answer that best describes their approach, or what was most important to them in relation to their approach to vaccinations.
Six respondents were excluded due to missing or “other” responses to the question on their approach to vaccination. Responses were weighted back to the eligible population and subgroup numbers may not add to the total numbers due to rounding.
Parents using nonstandard vaccination approaches differed demographically from those using the standard schedule (Table 1). In the multivariate model, low or moderate income (odds ratio [OR] = 1.54, 95% CI = 1.01-2.34) for household annual income <$75 000 vs higher), and having 2 or more children aged 7 to 48 months (OR = 1.71, 95% CI = 1.15-2.53) remained significant predictors of using a nonstandard approach while adjusting for parent age, sex, race/ethnicity, and education.
Improving the Vaccination Process
We asked all parents to indicate the most important way the process of vaccination could be improved (Table 3). Among parents using the standard schedule, 51% said that nothing could be improved, 22% selected vaccine information as the most important factor to improve, and 17% selected the amount of choice they had as the most important factor to improve. Parents who used nonstandard approaches were more likely to select the amount of choice they had (40%) than vaccine information (26%) or nothing (19%) as the most important factor to improve. Only 5% of parents using the standard schedule and 8% of those with nonstandard approaches selected interactions with their doctor as most important to improve.
Parent Perspectives on the Most Important Ways the Vaccination Process Could Be Improved.
Responses are weighted back to the eligible population. Six respondents who had missing or “other” responses to the question on what vaccination approach they used were excluded. Numbers may not add to 100% of responses due to rounding and omission of other responses.
Among all parents who selected this option as the most important way to improve the process.
For parents who selected improving vaccine information as most important, we asked them to select the most important way to improve it. Among all parents, the most common response was that they would like to receive information before visits instead of at them (20%), followed by wanting more information on vaccine side effects and safety (19%). Compared with other parents, those who used nonstandard approaches were more likely to want more information on vaccine ingredients (24%) or to want the doctor to give better and more complete information overall (20%).
For parents who selected the amount of choice they had as most important to improve, we asked them to select the most important aspect. Parents most commonly said they wanted more choice about which vaccines their child received (34%), followed by having noninjectable options (28%), or more choice about when their child received vaccines (19%).
Rationales for Nonstandard Approaches
Among the 142 parents who had used nonstandard approaches, the most common reasons selected as most important were the belief that too many vaccines are given at one time (32%), and discomfort with some of the ingredients in vaccines (21%) (Table 4). Parents who used a different specific schedule or a general approach were more likely than others to select too many vaccines being given at one time. In contrast, parents who had no predetermined plan or who declined all vaccines most commonly chose discomfort with vaccine ingredients as their most important reason for using an alternative approach.
Most Important Reasons Given by Parents Who Used Nonstandard Vaccination Approaches.
Abbreviation: NA, not applicable (this response option was not offered to this subgroup).
The question asked the respondent to indicate only one “most important” reason for their approach. Counts were weighted back to the eligible population and subgroup numbers may not add to the total numbers due to rounding.
Hypothetical Alternative Schedules
We asked parents to consider 2 different hypothetical alternative vaccine schedules (see the appendix), and rate how much they would like each one for their own child. The first alternative schedule was a shot-limiting schedule. The child would get the same total number of shots recommended but would get fewer injections at each visit and would have to make more visits. Among all parents, 120 (10%) liked this alternative a lot; 435 (36%) liked it somewhat, and 659 (54%) did not like it. The second alternative schedule was a vaccine-skipping schedule. The child would not ever receive certain vaccines at all, would receive fewer total vaccines and would be unprotected for certain diseases. Among all parents, 62 (5%) liked this alternative a lot, 237 (20%) liked it somewhat, and 915 (75%) did not like it.
Overall, more parents reported liking the shot-limiting schedule somewhat or a lot (46%) than the vaccine-skipping schedule (25%). This pattern was consistent within strata defined by parent age, sex, race/ethnicity, education, income, and number of children aged 7 to 48 months. In multivariate modeling, parents who liked the shot-limiting alternative approach did not differ from those who did not in their sex, age, race/ethnicity, education, income, and number of children aged 7 to 48 months. In contrast, parents who liked the vaccine-skipping approach were more likely than those who did not to have an educational level of high school or less (OR = 1.72, 95% CI = 1.18-2.51) or 2 or more children (OR = 1.58, 95% CI = 1.16-2.14), after adjusting for other demographic factors. Parents using nonstandard vaccination approaches were more likely than others to endorse the shot-limiting schedule (75% vs 42%) or the vaccine-skipping schedule (65% vs 19%).
We asked parents who said they liked the vaccine-skipping approach (n = 299) to name all vaccines they would skip. They most commonly named influenza (33%), varicella (27%), and human papillomavirus (27%). Fewer (43, or 14%) named the measles-mumps-rubella vaccine; this represented only 3.5% of all 1222 survey respondents.
Discussion
Key Findings
This study was unique in that we asked a national sample of parents about the most important ways to improve the vaccination process. While almost half of parents said nothing could be improved, half identified potential improvements. Parents said they would like vaccine information earlier, before office visits, and would like more information about side effects. Parents also indicated interest in more flexibility in vaccine scheduling, both in response to direct questions on what could be improved and when presented with hypothetical alternative schedules.
Our findings indicate that improving the vaccination process is likely to require different approaches for different subgroups of parents. Those using the standard schedule who identified a factor to improve most commonly selected vaccine information as their desired focus. In contrast, parents using nonstandard approaches were most likely to identify having more choice as the key possible improvement. Very few said interactions with their doctor were the most important aspect to improve.
Comparisons With Other Studies
The current study goes beyond prior research in 2 important ways. First, while many national surveys have described parents’ vaccine attitudes, few have directly elicited parents’ perspectives on how to improve the process. Our findings identified specific, potentially modifiable factors that could lead to better experiences.
Second, our study adds new insight by identifying contrasts in the concerns of varying subgroups of parents based on their vaccine decision-making approaches. Our overall finding that 12% of parents were using nonstandard vaccination schedules was nearly identical to that of a national survey conducted in 2010. 3 We also confirmed prior observations that parents who use nonstandard vaccination approaches fall into several distinct subgroups.3,9
This study goes beyond past studies by defining the key concerns of parents with different types of nonstandard vaccination approaches. We found variation among these subgroups, suggesting that clinicians may wish to develop different approaches that target the concerns of each. Parents who used a specific nonstandard schedule or a general approach were most likely to identify too many vaccines being given at one time as their key concern. In contrast, those who had no predetermined plan or who refused all vaccines were most likely to identify discomfort with vaccine ingredients as their key concern. Clinical program leaders may wish to develop ways to rapidly identify individual parents’ priorities—ideally in advance of office visits—in order to offer tailored approaches.
Implications for Practice and Policy
Of the 2 areas parents rated most important to improve—vaccine information and the amount of choice they had—the first may be simpler to address. Our finding that some parents identified receiving vaccine information before office visits as the most important possible improvement is consistent with studies that suggest that parents may form opinions about vaccines before they see their child’s doctor, even before or during pregnancy.14,15 Ongoing studies are evaluating ways of offering such information during these critical time windows via multiple information channels including the internet and social media networks.16,17
Many parents who said vaccine information needed improvement identified more information about vaccine side effects and safety as their highest priority. The United States has relatively robust active surveillance of the safety of individual vaccines, but existing research has not been designed to compare the outcomes of the standard vaccination schedule with nonstandard schedules. A recent Institute of Medicine review recommended that the government incorporate study of the safety of the overall childhood immunization schedule into its research priorities. 18
Regarding parents’ interest in more flexibility in vaccine timing, it may be possible to meet some parents’ desires within the existing standard schedule, which allows age windows for some vaccines. However, in day-to-day practice, many clinicians recommend 3 or more vaccine doses at some visits to achieve high immunization coverage rates, which are designated as key metrics by the National Committee for Quality Assurance. 19 The emphasis on immunization rates as quality measures may deter some clinicians from offering more flexibility within the existing standard schedule.
Both policy makers and clinicians have to mediate the inherent tension between individual parents’ freedom to decline vaccines and society’s need for high vaccination rates to maintain herd immunity. 20 Clinicians might be able to guide some parents toward societally preferred decisions by presenting the standard schedule as the default, 21 but such presumptive communication approaches are associated with lower parental ratings of the visit experience than participatory approaches. 22
The possibility of changing the standard schedule to increase flexibility, or offering an alternative that increases spacing between vaccines, is controversial. Giving all possible vaccines simultaneously at the earliest opportunity reduces the risk of subsequent undervaccination. 23 It is not clear whether an alternative schedule that spaced vaccines over more visits would lead to more missed visits and undervaccination. It is conceivable that a more flexible schedule could result in higher immunization rates if fewer parents felt the need to skip vaccines. Research on the effects of alternative schedules on vaccine-preventable disease incidence would be useful, but rigorous studies may be difficult because these diseases are so uncommon. 24
Our findings suggest that it may not be possible to improve the vaccination process for some parents who are using nonstandard vaccination approaches without giving them more choice. Many of the reasons they articulated for choosing these approaches were not necessarily amenable to being changed by more information. This is consistent with a recent study’s finding that physician communication training may not reduce vaccine hesitancy. 25 Concerns about exposure to vaccine ingredients were the second most commonly cited rationale among parents who had adopted nonstandard vaccination approaches. If clinicians wish to offer information tailored to these parents, they may want to focus on vaccine ingredients and about how to monitor for vaccine side effects.
Limitations
The method of fielding surveys used in this study has been endorsed as useful in providing timely data from national samples. 26 The survey panel was highly representative of national demographics. While participants did not perfectly reflect the US population, the panel was recruited using methods that reduced sampling bias, and the probability weighting based on the survey respondent group helped adjust the results to be representative of national perspectives.
Nonresponse bias is a potential limitation of most surveys. The recruitment rate into the original Internet survey panel was higher than the response rate to a typical telephone survey, for which the tendency to volunteer is not strongly related to political preferences. 27 Since prospective survey panel members did not know the topics of surveys in advance of signing up for the panel, those who were not recruited most likely represented people who did not want to take any survey, rather than being biased toward or against interest in vaccination issues.
Our study asked parents to report their vaccination practices, and elicited opinions about hypothetical alternative vaccination schedules. However, people’s actual behavior does not always mirror their reports, nor their responses to hypothetical scenarios. Our results may be affected by social desirability bias and reporting bias. Our survey did not evaluate parents’ knowledge about the potential risks and benefits of specific changes in the vaccine schedule. This suggests caution in making policy changes based on these results.
Conclusions
Parents’ experiences with the childhood vaccination process might be improved by offering vaccine information earlier, providing more information about vaccine side effects and safety, and allowing more flexibility about vaccine scheduling. Parent with different vaccination approaches express different priorities, suggesting that clinicians and policy makers may wish to develop methods of rapidly identifying individual parents’ approaches and create tailored ways to address their needs.
Author Contributions
All authors participated in this study’s conception and design, interpretation of data, drafting and revision of the manuscript. CC supervised the acquisition of data and participated in its interpretation. GTR led the analysis and interpretation of data. TAL, BJZF, and EW obtained funding, participated in the interpretation of data, and contributed supervision.
Footnotes
Appendix
Survey Questions on Hypothetical Alternative Vaccination Schedules
| (Question A) |
| Your child gets the total number of injections (shots) that the doctor recommends but he/she gets fewer injections at each visit. This would mean you need to take your child to the doctor for more visits overall. It would also mean that your child would be unprotected for a longer time than if he/she followed the standard schedule. |
| (Question B) |
| Your child does not ever get certain vaccines or injections (shots) at all. This would mean your child gets fewer total vaccines. It would also mean that your child would be unprotected for certain diseases. |
| The response options for each question were: |
| ❑ I like it a lot |
| ❑ I like it somewhat |
| ❑ I do not like it |
Acknowledgements
We are very grateful to our National Stakeholder Committee for their thoughtful guidance: Allison Kennedy Fisher, MPH, Walt Orenstein, MD, Daniel Salmon, PhD, Vicky DeBold, PhD, RN, Tawny Buck, and Jana Horne. We thank Adrianna Saada, MPH, for her work in the qualitative phase of this project, Grace Lee, MD, for helpful comments on the study design and draft survey, and Jason Gerson, PhD, and many other Patient-Centered Outcomes Research Institute staff members for their support. We are grateful for the excellent assistance of Karen Silva. We greatly appreciate the crucial contributions of the parents who participated in these surveys.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a Patient-Centered Outcomes Research Institute award.