COVID-19 vaccine hesitancy January-March 2021 among 18-64 year old US adults by employment and occupation ========================================================================================================= * Wendy C King * Max Rubinstein * Alex Reinhart * Robin J. Mejia ## Abstract **Introduction** COVID-19 vaccine hesitancy increased among US adults April-December, 2020, and threatens efforts to end the pandemic. Among US adults 18-64 years, we report prevalence of and reasons for vaccine hesitancy, overall and by employment and occupation, during the COVID-19 vaccine rollout. **Methods** The Delphi Group at Carnegie Mellon University conducted a COVID-19 survey administered by Facebook. In January, February and March, 2021, 791,716, 710,529, and 732,308 Facebook users, respectively, reported age 18-64 years and answered a vaccine acceptance question. Weights matched the sample to the age, gender, and state profile of the US population. Percentages and risk ratios (RR) for vaccine hesitancy were estimated using a weighted Poisson regression; 95% confidence intervals (CI) were calculated using robust standard errors. **Results** Vaccine hesitancy decreased among adults 18-64 years from January (27.5% [95%CI, 27.3-27.6]) to March (22.1% [95%CI, 21.9-22.2]). Vaccine hesitancy varied widely by occupational category: 9.6%, (95%CI, 8.5-10.7) in life/physical/social sciences to 46.4% (95%CI, 45.1-47.7) in construction/extraction. Almost half (47.9%, 95%, 47.6-48.3) of hesitant participants indicated concern about side effects, and over a third didn’t believe they needed the vaccine, didn’t trust the government, were waiting to see if it was safe, and didn’t trust COVID-19 vaccines (versus 14.5% [95%CI, 14.3-14.8] who didn’t like vaccines in general). **Conclusions** In this nationally representative survey of adults 18-64 years, vaccine hesitancy decreased to 22.1% by March, 2021. Still, hesitancy, which varies widely by occupation, remains a barrier to pandemic control. Reasons for hesitancy indicate messaging about safety and addressing trust are paramount. ## Introduction Vaccine hesitancy is a barrier to ending the COVID-19 pandemic.1 Despite the COVID-19 death toll,2 COVID-19 vaccine hesitancy increased among US adults April-December, 2020.3 In early December, only 56.2% reported they were likely to get vaccinated1. The likelihood was similar among healthcare workers (HCW) (55.3%)4, and lower among working-age adults (<65 years) versus older adults1. Workplace outbreaks, spread of infection from HCW to patients and from workers to customers are public health threats.5,6 The extent to which vaccine hesitancy changed during the first three months of the US COVID-19 vaccine rollout, and if and how current vaccine hesitancy varies by employment is unknown. Among a large representative sample of US adults 18-64 years, we report vaccine hesitancy by month, January-March, 2021, and for March, cumulative COVID-19 vaccine uptake and current prevalence of and reasons for vaccine hesitancy, overall and by employment status and occupational category. ## Methods Since April, 2020, the Delphi Group at Carnegie Mellon University (CMU) has been conducting an ongoing national COVID-19 survey in collaboration with the Facebook Data for Good group. Each month the survey is offered to a random sample, stratified by geographic region, of ≈100 million US residents from the Facebook Active User Base who use one of the supported languages (English [American and British], Spanish [Spain and Latin American], French, Brazilian Portuguese, Vietnamese, and simplified Chinese) via a link at the top of their Facebook News Feed. The CMU Institutional Review Board approved the survey protocol and instrument. ### Study sample Facebook users may be offered the survey more than once at different intervals depending on which strata they fall in, from once a month to once every six months. To show trends over time in vaccine hesitancy, we used data from January 6 (date the question was initially asked in its current form) to March 31, aggregated by month. While is possible there are repeat respondents across months, we cannot link respondents longitudinally, so data is treated as repeat cross-sectional surveys. In the primary analysis, we only use March data. In March, 2021, 104,768,154 Facebook users were offered the survey of whom 1,291,957 completed at least two survey questions. Respondents were excluded if they did not report age 18-64 (n=558,972) or did not answer the vaccine acceptance question (n=677), leaving 732,308 participants. Applying the same criteria, the January and February samples had 791,716 and 710,529 participants, respectively. ### Measures The survey questions and response sets utilized in this report are provided in an **appendix** (supplemental material). Participants age 18-64 were categorized as vaccine hesitant if they answered that they probably or definitely would not choose to get vaccinated if offered a vaccine to prevent COVID-19 today (versus probably or definitely would choose to get vaccinated or were vaccinated). Participants were categorized by employment status in the past 4 weeks, and if employed, by occupational group and profession. ### Statistical analysis Percentage vaccine hesitant by month (Jan-March 2021) was calculated using Facebook-provided weights, which account for the sampling design, non-response, and differences between the demographics of survey respondents and the US adult population.7 Among the March 2021 sample, weighted percentages for worked outside the home, vaccinated, and vaccine hesitant were calculated, overall and by employment variables. Additionally, risk ratios (RR) for vaccine hesitancy by employment status and occupation were estimated using a weighted Poisson regression. Finally, weighted percentages for reasons for hesitancy were calculated among vaccine hesitant participants overall and among occupations with the highest vaccine hesitancy. For all parameters, 95% confidence intervals (CI) were calculated using robust standard errors.8 Analyses were conducted in R (Version 4.0.2, R Core Team, Vienna, Austria). ## Results March participants 18-64 years (N=732,308) had a median age range of 35-44 years, 45.0% were male, 51.1% female, 3.8% other/unknown gender; 18.9% were Hispanic, 63.8% were non-Hispanic white, 6.8% non-Hispanic black, 3.2% non-Hispanic Asian, and 7.3% mixed/other/unknown race. Two-thirds (66.4%, 95%CI, 66.3-66.5) worked for pay; half (50.8%, 95%CI, 50.6-50.9) worked outside the home. Demographics were similar in January and February (data not shown). Vaccine hesitancy decreased among adults 18-64 years each month from January (27.5% [95%CI, 27.3-27.6]) to February (25.7 [95%CI, 25.6, 25.8]) to March (22.1% [95%CI, 21.9-22.2]). The time trend was similar among those who worked outside the home (January: 29.5% [95% CI, 29.3-29.7] to March: 24.8% [95%CI, 24.6-24.9]) and those who worked from home (January: 15.1% [95% CI, 14.9-15.3] to March: 11.4% [95%CI, 11.2-11.6]) (**Figure 1**). ![Figure 1.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2021/04/23/2021.04.20.21255821/F1.medium.gif) [Figure 1.](http://medrxiv.org/content/early/2021/04/23/2021.04.20.21255821/F1) Figure 1. Vaccine hesitancy overall and by employment status, by month. March cumulative vaccination uptake and current hesitancy by employment status and occupational category are reported in **Table 1**. HCW (78.3% [95%CI, 77.8-78.7] practitioners/technicians; 67.7% [95%CI, 67.0-68.3] support) and educators (68.5%, 95%CI, 68.0-69.0) led vaccine uptake, which was 38.0% (95%CI, 37.9-38.1) overall. View this table: [Table 1.](http://medrxiv.org/content/early/2021/04/23/2021.04.20.21255821/T1) Table 1. SARS-CoV-2 cumulative vaccination uptake and current vaccine hesitancy, overall and by employment status and category, in March, 2021, among 18-64 year old US adults. Vaccine hesitancy varied widely by occupational category: 9.6%, (95%CI, 8.5-10.7) in life/physical/social sciences and 9.7% (95%CI,9.4-10.1) in educators to 42.6% (41.5-47.7) in installation/maintenance/repair and 46.4% (95%CI, 45.1-47.7) in construction/extraction. Several occupational categories had lower hesitancy than HCW (practitioners/technicians and support; **Table 1**). Among HCW, Pharmacists had the lowest hesitancy (range 8.5%, 95%CI, 6.8-10.2). Physicians (12.2% [95%CI, 11.2-12.3]) and registered nurses/nurse practitioners had relatively low hesitancy (11.7% [95%CI, 10.6-13.9]) versus licensed practical/vocational nurses (19.0% [95%CI, 17.6-20.4]). Home health/personal-care aids, medical assistants, emergency medical technicians/paramedics, and nursing assistants/psychiatric aides had the highest hesitancy (20.5% [95%CI, 19.2-21.8] to 23.1% [95%CI, 21.2-25.0]). Reasons for vaccine hesitancy are reported in **Table 2** overall and by occupational categories with at least one-third hesitant. Almost half (47.9%, 95%, 47.6-48.3) of hesitant participants indicated concern about side effects, and over a third didn’t believe they needed the vaccine, didn’t trust the government, were waiting to see if it was safe, and didn’t trust COVID-19 vaccines (versus 14.5% [95%CI, 14.3-14.8] who didn’t like vaccines in general). More than one fifth indicated allergic reaction risk, unknown efficacy and other people need more as reasons. Worry of side effects, an allergic reaction and a plan to wait and see if safe were reported less frequency among high-hesitancy occupational groups (versus all hesitant participants), whereas lack of trust in the COVID-19 vaccine and the government, disbelief of need and dislike of vaccines were more frequent (**Table 2**). View this table: [Table 2.](http://medrxiv.org/content/early/2021/04/23/2021.04.20.21255821/T2) Table 2. Reasons 18-64 year old US adults reported vaccine hesitancy overall, and by occupational categories with the highest percentage hesitant. ## Discussion This data highlights the disparity in vaccine hesitancy by occupation. Among HCW, several professions with high patient contact (e.g., nursing assistants/psychiatric aides) reported hesitancy >20%. Occupational categories with the highest hesitancy (construction/extraction, installation/maintenance/repair, farming/fishing/forestry, transportation/material moving, and production), include some that have suffered workplace outbreaks.5 Reasons for hesitancy indicate a need for messaging about safety and addressing trust. ### Limitations Our large sample size allowed for precise vaccine hesitancy estimates by occupation. However, three cross-sectional samples were used to evaluate time trends, and the sample representativeness may have been affected by the recruitment method and low response rate, addressed with weighting.7 Compared to the [dataset] American Community Survey 2015-2019 5-year 2 Data Release9, demographics of the weighted sample are similar to the US population, but white race and having a college degree are slightly over-represented. Thus, overall hesitancy prevalence estimates might be underestimated.3 This should have minimal effects on time trends or comparisons between occupational categories. ## Conclusion In this nationally representative survey of adults 18-64 years, vaccine hesitancy decreased from 27.5% to 22.1% January-March, 2021. Still, hesitancy among the workforce (21.8%), which varies widely by occupation, remains a barrier to pandemic control. ## Supporting information Appendix with Survey [[supplements/255821_file02.pdf]](pending:yes) ## Data Availability If you are interested in using the survey data for your research, you can start the process by submitting a form requesting a data use agreement (DUA) from Facebook. The data is not available from the authors. [https://dataforgood.fb.com/docs/covid-19-symptom-survey-request-for-data-access/](https://dataforgood.fb.com/docs/covid-19-symptom-survey-request-for-data-access/) ## Conflict of interest statement Drs. King, Mejia and Mr. Rubenstein have no conflict of interest to report. Dr. Reinhart received salary support from an unrestricted monetary gift Facebook. ### Funding/Support This material is based upon work supported by Facebook (unrestricted gift) and a research grant from the Centers for Disease Control and Prevention (U01IP001121). Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of Facebook or the Centers for Disease Control and Prevention. ### Role of the Funder Facebook was involved in the design and conduct of the study. The CDC provided funding only. Neither Facebook nor the Centers for Disease Control and Prevention had a role in the collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. ## Acknowledgements The authors would like to thank the Delphi Group at Carnegie Mellon University for input and support on the survey instrument. Wichada La Motte-Kerr, MPH, of Delphi contributed to the development and deployment of the survey and received compensation for her contributions to the study. We thank Sarah LaRocca, PhD and Katherine Morris, PhD of Facebook for contributions to the survey instrument design. ## Footnotes * mrubinst{at}andrew.cmu.edu * areinhar{at}stat.cmu.edu * rmejia{at}andrew.cmu.edu * Received April 20, 2021. * Revision received April 20, 2021. * Accepted April 23, 2021. * © 2021, Posted by Cold Spring Harbor Laboratory The copyright holder for this pre-print is the author. All rights reserved. The material may not be redistributed, re-used or adapted without the author's permission. ## References 1. 1.Schaffer DeRoo S, Pudalov NJ, Fu LY. Planning for a COVID-19 vaccination program. JAMA. 2020;323(24):2458–2459. doi: 10.1001/jama.2020.8711 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/jama.2020.8711&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2021%2F04%2F23%2F2021.04.20.21255821.atom) 2. 2.Centers for Disease Control and Prevention. Trends in number of COVID-19 cases and deaths in the United States reported to CDC, by state/territory. Atlanta, GA: CDC. CDC COVID Data Tracker. Accessed April 19, 2022. 3. 3.Szilagyi PG, Thomas K, Shah MD, et al. National trends in the US public’s likelihood of getting a COVID-19 vaccine-April 1 to December 8, 2020. JAMA. 2020;325(4):396–398. doi: 10.1001/jama.2020.26419. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/jama.2020.26419&link_type=DOI) 4. 4.Meyer MN, Gjorgjieva T, Rosica D. Trends in health care workers intentions to receive a COVID-19 vaccine and reasons for hesitancy. JAMA Netw Open. 2021;4(3):e215344. doi: 10.1001/jamanetworkopen.2021.5344. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/jamanetworkopen.2021.5344&link_type=DOI) 5. 5.Althouse BM, Wenger EA, Miller JC, et al. Superspreading events in the transmission dynamics of SARS-CoV-2: Opportunities for interventions and control. PLoS Biol. 2020;18(11):e3000897. doi: 10.1371/journal.pbio.3000897. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/journal.pbio.3000897&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=33180773&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2021%2F04%2F23%2F2021.04.20.21255821.atom) 6. 6.Gold JAW, Gettings JR, Kimball A, et al. Clusters of SARS-CoV-2 infection among elementary school educators and students in one school district - Georgia, December 2020-January 2021. MMWR Morb Mortal Wkly Rep. 2021;70(8):289–292. doi: 10.15585/mmwr.mm7008e4. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.15585/mmwr.mm7008e4&link_type=DOI) 7. 7.Barkay N, Cobb C, Eilat R, et al. Weights and methodology brief for the COVID-19 symptom survey by University of Maryland and Carnegie Mellon University, in partnership with Facebook. 6 Oct 2020, arXiv preprint [https://arxiv.org/abs/2009.14675](https://arxiv.org/abs/2009.14675) 8. 8.Freedman, D. A. On the so-called “Huber sandwich estimator” and “robust standard errors”. The American Statistician. 2006;60(4):299–302. doi: 10.1198/000313006X152207 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1198/000313006X152207&link_type=DOI) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000241706900001&link_type=ISI) 9. 9.United States Census Bureau. American Community Survey 5-year 2019 Data Release. Version: December 10, 2020. American Community Survey 5-Year Data (2009-2019) (census.gov). Accessed April 7, 2021.