Demographic differences in US adult intentions to receive a potential coronavirus vaccine and implications for ongoing study

Introduction

The public health and economic impacts of the coronavirus pandemic have led to much hope by US policymakers, news media, and the general public about the prospect of a safe, effective vaccine to either prevent infection or minimize symptoms and reduce mortality risk. However, recent US polls indicate a concerning level of hesitancy that will likely lead to suboptimal uptake if such a vaccine becomes available. (e.g., 1,2) Nevertheless, many polls only offer limited information via reporting findings in aggregate or, at best, broken down into crosstabulations according to select demographic subgroups (versus multivariable models that adjust for confounding factors and incorporate a fuller range of the data for deductive or inductive analyses).

Furthermore, when asking participants if they would be willing to get a coronavirus vaccine, many surveys only provide or report on two options—yes or no—excluding an important “uncertain/unsure” category that aids in more comprehensively assessing the breadth and prevalence of hesitancy attitudes. (1,3) This omission is significant, as vaccine hesitancy is a complex phenomenon that encompasses a spectrum of vaccination decision-making, including degrees of uncertainty and complete refusal. (4)

In addition to these methodological issues, several substantive factors unique to a coronavirus vaccine further complicate understanding public hesitancy towards it. First, any public concern, uncertainty, or interest in its receipt centers on a vaccine that is, as of the time of this writing, hypothetical versus an existing product with a known safety and effectiveness profile like all vaccines currently recommended for children and adults. Second, the development of a coronavirus vaccine has been extensively hyped and financially supported by the US Federal Government’s “Operation Warp Speed” initiative, which aims to significantly expedite the development, approval, and production of a vaccine. Promotion of this initiative to the news media and public by US President Donald Trump and his administration during an election year and amidst a pandemic that has witnessed the politicizing of public health and scientific expertise have raised concerns that this expedited process may lead to the approval of an insufficiently evaluated and potentially unsafe and/or inadequately effective vaccine. (5,6) Lastly, though several vaccine candidates exist in different phases of development and evaluation, their safety and effectiveness are already being challenged by anti-vaccine and COVID-19 science denialism disinformation regarding a coronavirus vaccine and other public health measures that have been spreading online for months. (7)

Motivated by these concerns and issues, the present study aims to contribute to current understanding about potential hesitancy towards a COVID-19 vaccine by evaluating demographic differences in US adults’ intentions to receive a potential coronavirus vaccine and the different motivations for such intentions. Existing evidence on adult vaccination provides much reason to be concerned not simply about overall hesitancy for a coronavirus vaccine, but also how such hesitancy may be more prevalent in specific demographic subgroups. These include racial-ethnic and lower socioeconomic status (SES) groups with lower adult vaccination uptake rates and histories of marginalization and surveillance that contribute to their distrust of government, the health care profession, and medical treatments. (8,9) Such hesitancy could risk exacerbating existing disparities in COVID-19 cases and mortality—especially among African Americans, Latinos, and other racial-ethnic groups as well as lower SES groups. (10,11)

Beyond just race-ethnicity and SES, there are empirical and theoretical reasons to conjecture that other demographic groups may be more likely to receive the vaccine due to their COVID-19 risk profiles: older adults, who experience the highest hospitalization and mortality rates (12); men, who experience a higher COVID-19 mortality rate than women (13); married or cohabitating individuals wanting to protect one’s partner and self; persons who have had a relative or close friend diagnosed with COVID-19; employed individuals, whose occupational conditions may necessitate a vaccine to reduce their exposure risk; and geographic locations, like cities and regions of the country where COVID-19 has been more prevalent. Lastly, given the extent to which scientific information and expertise has been politicized and contributed to most COVID-19 related attitudes and behaviors being divided by political ideology (14), one can hypothesize that persons holding liberal or moderate political beliefs will be more likely to express intent to receive the vaccine than will their more politically conservative counterparts, even after controlling for other demographics.

In addition to examining differences in intention to receive a potential COVID-19 vaccine, it is also important to understand potential motivations underlying people’s intentions to vaccinate and how people holding specific motivations might differ from others who are either unsure or have no intention. This study examines five prominent motivations. In addition to self-interested, personal protection reasons, vaccination intentions include the prosocial motivations of protecting one’s family and community (15) as well whether one’s physician or health care provider recommends vaccinations (4). Likewise, having a chronic health condition may be another significant motivation to receive the vaccine due to such conditions being risk factors for severe COVID-19 illness and mortality. (16) This reason may be particularly salient for older adults and persons of lower socioeconomic status and racial-ethnic minority status—all groups that experience disproportionate rates of chronic conditions identified as COVID-19 risk factors. Together, these five intentions may be useful for designing eventual vaccine messaging and campaigns to promote uptake.

Methods

The data are from the May 14-18, 2020 Associated Press (AP)-NORC Center for Public Affairs Research poll, publicly accessible at https://apnorc.org/projects/expectations-for-acovid-19-vaccine/ and determined by the University of California, Riverside Institutional Review Board to be exempt from review. Stata code for this analysis is available from the author for transparency and reproducibility. While data collection procedures are detailed in-depth elsewhere (17), briefly, data were collected using the AmeriSpeak Omnibus®, a monthly survey using NORC’s probability-based panel designed to be representative of the U.S. household population. Adults ages 18 and older from the 50 states and the District of Columbia were randomly drawn from the panel and 1,056 completed the survey via the web and telephone, with interviews conducted in English and Spanish. The analytic sample for this study consisted of 1000 respondents who had complete information for all study variables.

Results

Descriptive Statistics

Figure 1 shows that a substantial percentage of respondents are hesitant about receiving a potential coronavirus vaccine: nearly 20% expressed no plan to get it and approximately 30% were unsure. Among the 50.1% who answered “Yes,” the most common reasons were to protect one’s self (93.3%), family (88.6%), and community (78.0%), while intent due to a chronic health condition (36.2%) or what one’s doctor might endorse (45.3%) were less prevalent.

• Download figure
• Open in new tab

Fig. 1: Distribution of the vaccine intention responses (n = 1000).

NOTE: Reported n’s in parentheses are unweighted. Though specific reasons were only asked of those respondents who answered “Yes” to whether they plan to get a coronavirus vaccine, the weighted percentage reported for each reason is based on the entire sample (n = 1000), which includes all respondents offering that reason as well as those reporting that they (a) plan to get the vaccine, but for a different reason than the one specified reason, (b) do no plan to get a coronavirus vaccine, and (c) are unsure about getting the vaccine.

Table 1 reports the weighted percentages for each independent variable across the overall sample and for the three categories of vaccine intent.

View this table:

• View inline
• View popup

Table 1: Descriptive statistics for independent variables across coronavirus vaccine intention options (n = 1000).

Discussion

The prospect of a safe, effective coronavirus vaccine holds promise for enabling the US to mitigate the substantial COVID-19 morbidity and mortality it has been experiencing. However, the present findings indicate that such enthusiasm for a vaccine is unequally distributed along sociodemographic lines. This analysis of a national sample reveals substantial demographic subgroup differences not only among those expressing intent versus none, but also those expressing intent versus uncertainty. At the time of this writing, only a few vaccine candidates have even entered Phase III trials. (19) Hence, any potential vaccine is a hypothetical consideration. However, these findings raise important concerns for what may lie ahead if a vaccine is ultimately approved by the US Food and Drug Administration (FDA) and efforts are made to design and implement effective, nation-wide vaccination campaigns.

Most broadly, they point to the challenges that lie ahead for achieving significant population uptake to ensure herd immunity. Considering the coronavirus pandemic’s substantial health, social, and economic impacts, it is troubling that only approximately 50% of respondents expressed intention to get a vaccine that would potentially protect them and others from disease and death, and, collectively, would facilitate a re-opening of society and re-engagement in many activities currently curtailed due to the risk. One may be heartened that an additional 30% reported being unsure, as that may indicate a significant number of “fence-sitters” (20) who are willing to potentially consider it and may be easier to convince than the approximately 20% who said, “No.” However, as Table 3 shows, there is a comparatively larger number of differences observed between the unsure and no groups than there are for the yes versus unsure groups. Hence, while there are some important differences between the unsure and no groups, these findings highlight how convincing unsure respondents to say “yes” will remain a challenge.

The subgroup differences also show the potential for widening health disparities. Extensive attention has been paid to the disproportionate COVID-19 exposure risks and morbidity and mortality burdens that non-White and lower income subpopulations have experienced to date. (10,11) Yet, the present findings suggest that the opportunity to substantially mitigate such disparate risks via an effective vaccine will be unequally realized—notably by racial-ethnic and SES groups already significantly impacted by COVID-19—with the troubling potential for current disparities widening.

The findings on specific intentions to vaccinate are helpful in this regard. High prevalence was observed across three of the five specific reasons that, altogether, lend themselves well to health promotion efforts. Nevertheless, having one’s doctor recommend vaccines had the lowest endorsement, which is concerning as health care provider expertise is often a focus in addressing vaccine hesitancy. Also, despite the variation in support across these five items, they still produced very similar patterns of findings in the multivariable models. This suggests that it is important to start considering a range of local, county, state, and federal strategies for outreach, trust building, and education/promotion to marginalized and other demographics. Vaccine mandates are not a simple answer to uptake disparities—they can be politically fraught (especially in the current US climate where science has been heavily politicized) and counter-productive to the goals of maximizing vaccine uptake. As such, they should be viewed as a last resort. (21) Furthermore, it is important to consider how such hesitancy attitudes may exist due to context. For example, expressing uncertain or no intention to receive the vaccine may reflect previous or current barriers to obtaining preventive health care (i.e. people reporting little to no intention to get the vaccine due to a belief that it will be difficult to obtain) and distrust of medical professionals and/or the vaccine itself. (22)

Limitations

Several study limitations should be acknowledged. First, these data were collected in mid-May 2020. Thus, responses could have been affected by specific media coverage at that time period about vaccine development or Operation Warp Speed, the severity of cases and deaths, and extent of re-opening policies throughout the nation.

Second, this analysis only focused on adults. Vaccination will also be important for children in order to ensure their safety and achieve herd immunity for others. However, parents’ attitudes towards having their children receive a COVID-19 vaccine are unclear and require more research.

Third, respondents were asked about receiving a hypothetically available vaccine. It is possible that people may have different attitudes if a vaccine is eventually made available, as it will have a reported safety and effectiveness profile.

Fourth, while the analysis focused on different reasons for intentions to receive the vaccine, it did not examine why certain respondents stated they either had no intention to receive the vaccine or were unsure. This was due to survey design and data limitations. The survey included a series of nine items asked of those respondents who initially stated no intention to receive the vaccine. Yet, due to the small proportion of such respondents in the overall sample and the small number of endorsements for each specific reason (all but one reason endorsed by less than 69 respondents), it was not possible to do a systematic modeling of subgroup comparisons of those responses. This is instructive for other studies in that comprehensive analyses of hesitancy will require (a) sample sizes larger than the 1000 respondents in this analysis and (b) specific reason items for those answering unsure as well as yes and no.

Implications for Future Research

Overall, these findings offer useful insights—particularly with respect to three valid categories of intention: no, unsure, and yes. First, future surveys need to include an “unsure” category, as efforts to promote vaccination among the uncertain will likely need to be very different in approach and design than efforts aimed at those who have no intention of getting vaccinated.

Second, in terms of measurement, it is important to be mindful that a “yes” response in this or any future survey is capturing potential to receive the vaccine (under hypothetical circumstances) and is just one of multiple factors that a person’s receipt is contingent upon. Hence, such “yes” estimates should not be taken for granted. How such intentions may fare if a vaccine is released will depend on access issues that may facilitate or hinder receipt. Translating such willingness into actual effort to receive the vaccine will require designing programs that offer the public sufficient information about its safety and effectiveness and make its receipt as convenient as possible. Likewise, per research on diffusion of innovation (23), heterogeneity likely exists among the yes and uncertain responses in willingness to receive the vaccine. Though people may express willingness to receive this new innovation, they may not necessarily intend to be the “innovators” and “early adopters” (i.e. first in line), but rather wait for others to receive it first to observe from people in their networks what the side effects might be like.

Third and finally, COVID-19 vaccine-related attitudes and receipt intentions need to be assessed repeatedly by national polls. Public health efforts would greatly benefit from survey firms and polling organizations making their data easily available to the public health community (like NORC has) to conduct more extensive analyses than what are typically presented in polling organization’s public audience- and news media-focused reports. At a time when significant amounts of COVID-19-related disinformation are circulating and science is being heavily politicized and challenged amidst the tremendous burdens of the pandemic itself, quality information on the public’s vaccine intentions and the reasons for such intentions are crucial— especially as multiple Phase III trials proceed and the design of effective vaccination campaigns nation-wide becomes increasingly more urgent.