Abstract
This internet survey study describes trends in respondents saying they were likely to accept vaccination for COVID-19 overall and by age, sex, race, and education between April and December 2020.
The coronavirus disease 2019 (COVID-19) pandemic is causing enormous morbidity and mortality across the US and is disproportionately affecting racial/ethnic minority populations and elderly persons. High acceptance of COVID-19 vaccines will be instrumental to ending the pandemic.
Four cross-sectional internet surveys1,2,3,4 (3 using convenience samples1,3,4) from April2 and May1,3,4 2020 found that 58% to 69% of adults intended to get vaccinated against COVID-19, with higher percentages reported in April2 than in May.1,3,4 These studies did not track the same individuals over time, making it difficult to assess whether intent to get vaccinated has truly declined.
We analyzed biweekly survey data from a nationally representative longitudinal study to describe changes over time in the public’s likelihood of getting a COVID-19 vaccine and across demographic subgroups.
Methods
The Understanding America Study (UAS) is a probability-based internet panel survey of approximately 9000 noninstitutionalized US adults.5 Respondents are recruited using address-based sampling, allowing for valid statistical inferences and avoiding coverage problems from convenience web panels. Internet-enabled tablets are provided if needed.
Beginning March 10, 2020, the entire UAS panel has been invited to participate in biweekly tracking surveys about COVID-19; consenting respondents are invited on a rolling basis (590 invited daily) to participate in each wave and have 2 weeks to complete surveys.
Beginning April 1-14 and continuing through November 25–December 8, we asked: “How likely are you to get vaccinated for coronavirus once a vaccine is available to the public?” Response options (very unlikely, somewhat unlikely, somewhat likely, very likely, unsure) were dichotomized into “somewhat or very likely” vs all others. A multivariable Poisson regression model with robust standard errors was used to estimate adjusted risk ratios (aRRs) and 95% CIs of the likelihood of getting vaccinated. Changes in likelihood of getting a COVID-19 vaccine (overall and by demographic subgroups) were analyzed using linear regression models with cluster-robust standard errors at the respondent level. Analyses (SAS version 9.4) accounted for survey sampling weights. Statistical significance was defined as a 95% CI that did not cross 1. The UAS panel oversamples residents of Los Angeles County and, to a lesser extent, residents of the rest of California; survey sampling weights adjust for this.
Participants provided written informed consent and the University of Southern California’s institutional review board approved this study.
Results
To date, 8167 UAS panel respondents have consented to participate in the biweekly tracking surveys. The number of consenting respondents who completed the surveys varied across 2-week periods (range: 5259 to 6139; range of completion rate per 2-week period: 75%-97%).
During November 25–December 8 (Table), the self-reported likelihood of getting COVID-19 vaccination was lower among women than men (51% vs 62%; aRR, 0.9 [95% CI, 0.8-0.9]) and Black vs White individuals (38% vs 59%; aRR, 0.7 [95% CI, 0.6-0.8]), and higher among adults aged 65 years and older vs those 18-49 years (69% vs 51%; aRR, 1.4 [95% CI, 1.3-1.5]) and those with at least a bachelor’s degree vs a high school education or less (70% vs 48%; aRR, 1.5 [95% CI, 1.3-1.6]).
Table. Percentage of Adults Who Stated in April 2020 and December 2020 That They Were Somewhat or Very Likely to Get Vaccinated for COVID-19, and Change Over Time.
| Demographic characteristic | April 2020 survey (April 1-14); likely to get a vaccine, % (95% CI) | December 2020 survey (November 25–December 8) | Change over time: April 2020 survey to December 2020 survey, % (95% CI)b | ||
|---|---|---|---|---|---|
| Sample size | Likely to get a vaccine, % (95% CI) | Adjusted risk ratio (95% CI)a | |||
| Overall | 74.1 (72.4-75.8) | 5660 | 56.2 (54.4-58.1) | −17.9 (−20.0 to −15.8) | |
| Sex | |||||
| Men | 79.1 (76.8-81.5) | 2362 | 62.3 (59.6-65.0) | 1 [Reference] | −16.9 (−20.0 to −13.8) |
| Women | 69.5 (67.1-71.9) | 3298 | 50.6 (48.1-53.1) | 0.9 (0.8-0.9) | −18.9 (−21.8 to −16.0) |
| Age group, y | |||||
| 18-49 | 69.1 (66.5-71.7) | 2431 | 50.9 (48.2-53.7) | 1 [Reference] | −18.2 (−21.4 to −15.0) |
| 50-64 | 76.7 (73.8-79.6) | 1755 | 57.0 (53.7-60.4) | 1.2 (1.1-1.3) | −19.7 (−23.5 to −15.8) |
| ≥65 | 83.8 (81.0-86.6) | 1473 | 69.1 (65.7-72.5) | 1.4 (1.3-1.5) | −14.7 (−18.6 to −10.8) |
| Race/ethnicityc | |||||
| White | 77.8 (75.9-79.6) | 3851 | 58.6 (56.5-60.8) | 1 [Reference] | −19.1 (−21.5 to −16.8) |
| Black | 50.7 (44.7-56.7) | 413 | 37.6 (32.0-43.2) | 0.7 (0.6-0.8) | −13.1 (−19.9 to −6.3) |
| Hispanic | 73.1 (67.8-78.3) | 810 | 52.7 (47.1-58.2) | 1.0 (0.9-1.1) | −20.4 (−27.2 to −13.6) |
| Asian | 90.9 (86.1-95.7) | 290 | 80.6 (73.9-87.2) | 1.3 (1.2-1.4) | −10.3 (−18.5 to −2.2) |
| Education | |||||
| High school completion or less | 67.0 (63.7-70.2) | 1146 | 47.6 (44.3-50.9) | 1 [Reference] | −19.4 (−23.4 to −15.4) |
| Some college | 70.5 (67.4-73.6) | 2056 | 50.2 (46.9-53.5) | 1.1 (1.0-1.2) | −20.3 (−24.3 to −16.4) |
| Bachelor’s degree or higher | 85.0 (82.8-87.1) | 2457 | 70.3 (67.5-73.0) | 1.5 (1.3-1.6) | −14.7 (−17.6 to −11.9) |
Abbreviation: COVID-19, coronavirus disease 2019.
Mutually adjusted for all the demographic factors in the table.
Ordinary least squares with cluster-robust standard errors was used to assess change over the time frame, which accounted for correlation of repeated measures and survey sampling weights.
Online panel members self-reported race/ethnicity from a set of survey questions. Data are not presented for American Indian/Alaska Native, Native Hawaiian and Other Pacific Islander, and other non-Hispanic multiracial groups due to small sample sizes.
Between April 1-14 and November 25–December 8, the percentage who stated they were somewhat or very likely to get vaccinated declined from 74% to 56% (difference: 18 percentage points [95% CI, 16-20]) (Figure). Significant declines over time in the likelihood of seeking vaccination were observed for both women and men and in all age, racial/ethnic, and educational subgroups.
Figure. Percentage of US Adults Who Say They Are Likely to Get a COVID-19 Vaccine.
Data were not collected for the periods of June 10-23 and July 8-21. Percentages and 95% CIs are plotted. COVID-19 indicates coronavirus disease 2019.
Discussion
In this nationally representative survey, self-reported likelihood of getting a COVID-19 vaccine declined from 74% in early April to 56% in early December 2020, despite the early November press releases of high vaccine efficacy for 2 vaccines in phase 3 trials, although prior to Emergency Use Authorization. Low likelihood of getting a COVID-19 vaccine among Black individuals and those with lower educational backgrounds is especially concerning because of their disproportionately higher burden from COVID-19 disease.
Study strengths include analyses of change over time within the same group of individuals from a representative online survey panel. Limitations include use of only English- and Spanish-language surveys, self-reported metrics that may not correlate with future behavior, and small sample sizes for certain minority populations.
Educational campaigns to raise the public’s willingness to consider COVID-19 vaccination are needed.
Section Editor: Jody W. Zylke, MD, Deputy Editor.
References
- 1.Reiter PL, Pennell ML, Katz ML. Acceptability of a COVID-19 vaccine among adults in the United States: how many people would get vaccinated? Vaccine. 2020;38(42):6500-6507. doi: 10.1016/j.vaccine.2020.08.043 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fisher KA, Bloomstone SJ, Walder J, Crawford S, Fouayzi H, Mazor KM. Attitudes toward a potential SARS-CoV-2 vaccine: a survey of US adults. Ann Intern Med. 2020;173(12):964-973. doi: 10.7326/M20-3569 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Malik AA, McFadden SM, Elharake J, Omer SB. Determinants of COVID-19 vaccine acceptance in the US. EClinicalMedicine. 2020;26:100495. doi: 10.1016/j.eclinm.2020.100495 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Head KJ, Kasting ML, Sturm LA, Hartsock JA, Zimet GD. A national survey assessing SARS-CoV-2 vaccination intentions: implications for future public health communication efforts. Science Communication. 2020;42(5):698-723. doi: 10.1177/1075547020960463 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kapteyn A, Angrisani M, Bennett D, et al. . Tracking the effect of the COVID-19 pandemic on American households. Surv Res Methods-Ger. 2020;14(2):179-185. [Google Scholar]