Unraveling the Attributes of COVID-19 Vaccine Acceptance and Uptake in the United States: A Large National Study

A total of 36,711 users responded to the question about vaccine acceptance. The largest number of respondents came from Connecticut and California with 8697 and 4668, respectively (Fig. 1A). HWF’s user base is approximately 79% female (Fig. 1b) and 83% white (Fig. 1vs). The users are aged 18 or over and are also divided into age groups (Fig. 1d). Over 68% of respondents were non-essential workers and users span a wide range of income groups. All descriptive statistics of study participants are available in the supplementary table 1.

A total of 30,618 (83%) were willing (“likely” or “very likely”) to be vaccinated (Fig. 1A). After applying a census-based post-stratification weight (see Methods), Vermont (92%) and Washington DC (88%) had the highest rates of vaccine hesitancy, while Dakota South (27%) and Louisiana (23%) had the highest rates. rate of undecided users (Fig. 1b). Weighted bar charts of vaccine hesitancy by demographic characteristics revealed that “undecided” users accounted for the largest proportion of unwilling users across all demographic groups (Fig. 2a, supplementary table 2).

COVID-19 Acceptance Rate: (A) (Left) Number of responses and (Right) unweighted and weighted percentages. (b) Weighted average willingness and undecided rate by state.

Demographic Acceptance Rate: (A) Weighted percentages of reluctant responses by race/ethnicity, occupation, location, age, income and use of protective measures. State-Weighted Reluctance Rates by (b) cumulative case rate (/100 individuals), (vs) cumulative mortality rate (/1000 individuals), (d) and the average number of users practicing protective behavior.

State-level reluctance rates (“Undecided,” “Unlikely,” or “Very unlikely”) were negatively associated with the average number of users who practiced transmission-mitigating behaviors and were positively associated with Cumulative COVID-19 case and death rates as of January 10, 2021 (Fig. 2b–d). Unweighted graphs are available in Fig. 2.

To assess demographic associations with vaccine acceptance, we fitted univariate logistic regression with sociodemographic, occupational, pre-existing medical, geographic, and COVID-19-related predictors (Supplementary Table 3) and a multivariate logistic regression model to fit potential relationships between predictors (Fig. 3Additional table 4). We implemented post-stratification weights using census estimates of gender, age, race, and census location (see Methods). People of color reported higher rates of vaccine hesitancy compared to non-Hispanic white users (African American OR, 3.94; CI, 3.47, 4.48; p= 1.26e−96). Vaccination hesitancy was more likely in women than in men (OR, 1.67; CI, 1.51, 1.83; p= 4.09e−25); younger users than those over 65 (18–30 OR, 2.17; CI, 1.86, 2.53; p= 1.03e−22); those with three or more pre-existing conditions than those with zero (OR, 1.19; CI, 1.06, 1.34; p= 0.0036); and parents than non-parents (OR, 1.26; CI, 1.15, 1.38; p= 9.61e−7). People on leave or looking for work were also more reluctant to be vaccinated than those working full-time or part-time (OR, 1.48; CI, 1.29, 1.70; p= 4.04e−8). Southern respondents (OR, 1.25; CI, 1.05, 1.48; p= 0.0105), from less densely populated or low-income areas were all more likely to be vaccine hesitant. Users who responded before Pfizer’s emergency use authorization (EUA) on December 11, 2020 were more reluctant to be vaccinated than users who responded after Pfizer’s EUA (OR, 1.48; CI , 1.37, 1.60; p= 9.96e−23), users who practiced protective behaviors against COVID-19 such as mask-wearing or social distancing were less vaccine hesitant (OR, 0.78; CI, 0.72, 0, 85; p= 6.12e−9), and users who received a COVID test were less reluctant to vaccinate (OR, 0.79; CI, 0.71, 0.89, p= 5.88e-5).

Results of association analysis based on logistic regression of vaccine acceptance: forest plots for unweighted (left) and weighted (right) multivariate logistic regression analyzes for vaccine hesitancy with confidence intervals by 95%. Variables not significant at the 0.05 level (white), significant positive associations (red), and significant negative associations (blue).

Nominal logistic regression (see Methods) assessed whether vaccine hesitancy was driven by ‘Undecided’ versus ‘Unlikely/very unlikely’ responses (Supplementary Table 5) and was also conducted with a weighted analysis (Supplementary Table 6). The reluctance of healthcare workers, people aged 55-64, Asian users, and those with a median income between $70,000 and $100,000 was driven by the “Undecided” group, while the reluctance unemployed, people with 3 or more pre-existing conditions and people from the South users was motivated by the “unlikely” group. Sensitivity analyzes were performed for multivariate and nominal weighted regression analyzes with a less restrictive threshold for trimming weights (Supplementary Table 7–8) and found similar results. We performed a sensitivity analysis to assess differences in reluctance among people who tested positive for COVID-19 and found no differences in intent based on test results (see Methods, Supplementary Table 9–ten).

Of the 36,711 users who answered the vaccine uptake question, 23,429 also answered the vaccine uptake question and its distribution is provided in Fig. 4A. Demographic distributions remained similar to those of respondents to the vaccine acceptance question with a slight increase in the proportion of users aged 65 and over (Fig. 3). Vaccination rates by state are shown in Fig. 4b for all users who answered the question on vaccine use and the subset of respondents who were offered a vaccine. Users with lower levels of vaccine acceptance had lower vaccination rates (Fig. 4c), and black and Hispanic/Latin users reported lower vaccination rates than non-Hispanic white users (Fig. 4d). Graphs of weighted and unweighted vaccination rates for all demographic characteristics are available in the Supplementary Figures. 4–5.

Vaccination rate: (A) Answers to questions on the use of vaccines for all users. (b) State-weighted vaccination rates of (left) all users who answered the vaccine use question and (right) users who were offered a vaccine. (vs) Weighted vaccination rate of users who were offered a vaccine based on vaccine acceptance and (d) race/ethnicity.

To formally identify demographic characteristics associated with differences in vaccination rates, we performed unweighted and weighted multiple logistic regression analysis (see Methods, Fig. 5Additional tables 11–12). All age groups reported lower vaccination rates than users over 65 (18-30 OR: 0.10; CI: 0.06; 0.18; p= 1.43e−16); Black users reported lower vaccination rates (OR, 0.58; CI, 0.38, 0.91; p= 0.0165) compared to non-Hispanic white users; non-healthcare essential workers reported lower vaccination rates (OR, 0.64; CI, 0.44, 0.92; p= 0.0162) compared to non-essential workers; parents reported lower vaccination rates (OR, 0.63; CI, 0.45, 0.89; p= 0.0086) compared to non-parent users; users in areas where the median household income (MHI) is $40–70,000 (OR, 0.56; CI, 0.37, 0.85; p= 0.0066) and $70–100,000 (OR, 0.63; CI, 0.42, 0.96; p= 0.0316) reported lower vaccination rates than regions with an MHI of $100,000+; users connecting from areas of 0 to 149 people/m². mi reported lower vaccination rates (OR, 0.53; CI, 0.34, 0.82; p= 0.0049) compared to users in densely populated areas; and users who responded “unlikely/very unlikely” (OR, 0.02; CI, 0.01, 0.03; p= 2.07e−114) and “Undecided” (OR, 0.08; CI, 0.06, 0.12; p= 1.06e−39) to the vaccine acceptance question reported lower vaccination rates than consenting users.

Results of association analysis based on logistic regression of vaccine uptake: forest plots for unweighted (left) and weighted (right) multivariate logistic regression analyzes for vaccination uptake with intervals 95% confidence. Variables not significant at the 0.05 level (white), significant positive associations (red), and significant negative associations (blue).

While vaccination rates were lower in the reluctant group than in the accepting group, 86% (2,157/2,520) of the reluctant users were vaccinated. In a formal multiple regression analysis examining demographic associations with vaccination among hesitant users, similar associations were found (see Methods, Supplementary Table 13). Younger age groups, healthcare workers, people from low-income households, and residents of sparsely populated areas had lower vaccination rates. Users who answered “Undecided” to the vaccine acceptance question reported higher vaccination rates than those who answered “Unlikely/very unlikely” (OR, 4.57; CI, 3.47 , 6.03; p= 2.26e−26).