To the Editor —Healthcare workers (HCWs) are at high-risk for severe acute respiratory coronavirus virus 2 (SARS-CoV-2) infection. Reference Wilkins, Hirschhorn and Gray1,Reference Wilkins, Gray and Wallia2 They are influencers in vaccination for the public, and some have demonstrated hesitancy in receiving the coronavirus disease 2019 (COVID-19) vaccination. Reference Evans, DeYoung and Gray3 As of June 2022, 17% of HCWs in the United States had not been vaccinated. Reference Gu, Taylor, Pollack, Schneider and Zaller4 In addition, little is known of HCW receipt of booster vaccinations. In the face of SARS-CoV-2 becoming endemic, it is important to understand facilitators and barriers to booster uptake. Thus, we conducted an analysis describing HCW perceptions on and intentions to get the booster vaccine using data from a prospective cohort of HCWs August 2021 to April 2022 at Northwestern Medicine. Reference Wilkins, Hirschhorn and Gray1–Reference Evans, DeYoung and Gray3
Our study’s methodology has been previously described. Reference Wilkins, Hirschhorn and Gray1–Reference Evans, DeYoung and Gray3 Briefly, HCWs from 10 hospitals, 18 immediate-care centers, and 325 outpatient practices in the Chicago area and suburbs originally provided consent in May 2020–June 2020 to participate in a cohort study assessing the risk of COVID-19. Individuals completed baseline and follow-up surveys capturing demographics and risk factors including occupational tasks and COVID-19 vaccination. This analysis used questionnaires completed between August 2021 and October 2021, and a follow-up survey conducted in April 2022.
We used descriptive statistics for participants’ self-reported demographics (age, race or ethnicity, living situation), occupation, and exposure risks for COVID-19. Due to the small sample size, the responses to the question on intentions to get a COVID-19 booster were combined (Yes: ‘Yes, I have already gotten the shot’, ‘Yes, I plan to get the shot’; No: ‘No’, or ‘Unsure’) to form a binary outcome of intention or receipt of the booster vaccine. The χ Reference Wilkins, Gray and Wallia2 analyses (unadjusted associations) and logistic regression (adjusted associations) were performed between respondent factors and intention or receipt of the COVID-19 booster vaccine. Previous research on COVID-19 risk and sample size drove classifications for occupation and race and ethnicity. Reference Evans, DeYoung and Gray3 The final model was selected based on variable significance and the literature showing associations with vaccination. Reference Wilkins, Hirschhorn and Gray1–Reference Evans, DeYoung and Gray3 Statistical significance was determined using the adjusted odds ratios (AORs) or P < .05.
Overall, 2,600 (72.8%) of 3,571 enrolled completed the baseline survey. Respondents were older and more likely to identify as female and White race than nonrespondents and were more likely to work in non–patient-facing administrative positions and less likely to work as physicians or registered nurses. After excluding 152 respondents due to incomplete data, the final analytical sample was 2,448. Overall, 82.8% of HCW respondents indicated they would get or had already received the booster vaccine, and 17.2% were not willing to get or unsure about getting the booster vaccine. Unadjusted analyses showed booster vaccine intentions or receipt were highest in physicians (93.5%), Asian HCWs (92.7%), men (88.4%), and HCWs who had been in patient rooms (85.0%).
Logistic regression showed physicians continued to have 2.8 times higher odds of intending to get or having received the booster compared to administrative staff (Table 1). Compared to Asian HCWs, all other race and ethnicity groups had lower odds of intention to get or having received the booster. Respondents aged 30–49 years had lower odds of intention to get or receipt of a booster compared to those aged <30 years.
a Due to smaller sample size and similar responses, ‘No’ and ‘Unsure’ responses were grouped together, and ‘Yes’ responses were grouped together to form a binary outcome of intention or receipt of the booster compared to no or being unsure about getting the booster.
b Bolded items were statistically significant at P < .05 in unadjusted χ Reference Wilkins, Gray and Wallia2 analyses.
c The final parsimonious logistic regression model only included factors that remained significant in the model (95% confidence interval does not include 1). These factors were age, race/ethnicity, and occupation.
d AI/AN/NH/PI/Other/NA/Multiracial:s American Indian/Alaska Native/Native Hawaiian/Other Pacific Islander/other/did not answer.
e Other occupation: clinical education staff, high risk respiratory therapist, laboratory personnel, medical assistants, mental health consoler, PT/OT/speech pathologist, pharmacy, phlebotomist, radiology-radiograph technician, sonographer, patient care technician, and support services.
The following reasons were most frequently selected (>10% endorsed) for willingness to get or having gotten a booster vaccine: data showing additional protection (43.7%), not seeing a downside to getting another shot (24.8%), and recommendation by provider (19.1%). The following reasons were most frequently given for not wanting a booster vaccine (>10% endorsed): needing more data on the vaccine (32.4%), feeling that they do not need the booster (26.0%), fear of possible side effects (20.0%), and concern that other countries have not gotten enough shots (11.9%). Respondents were asked about willingness to get annual COVID-19 vaccine; most (86.4%) gave a positive response, but 13.6% were unsure or not willing. Most of those unwilling to get an annual COVID-19 shot (n = 335), were also not willing to get a booster (77.6%). The follow-up survey in April 2022 showed that of those 1,778 who completed this survey, 85.0% had received at least 1 booster vaccine. Of those 313 who reported no intentions to receive a vaccine booster in August 2021–October 2021 survey, fewer than half had received at least 1 booster vaccine (42.5%).
Most HCWs (82.8%) in this sample planned or had received the COVID-19 vaccine booster August 2021–October 2021, and 85% had received the booster by April 2022. This willingness or receipt was higher than the uptake rate for US adults (51.5%) as of August 2022, 5 for US healthcare personnel (67.1%) in the 2021–2022 season, Reference Razzaghi, Srivastav, de Perio, Laney and Black6 and for Chicago adults (47.3%) as of November 21, 2022. 7 This higher rate of actual or planned booster uptake in our cohort is encouraging given the toll of COVID-19 among HCWs and the need for HCWs to be trusted purveyors of health information. High booster uptake rates in physicians and specific HCW populations are consistent with previous data showing higher vaccination rates in physicians than other HCWs including nurses and by age and race. Reference Evans, DeYoung and Gray3,Reference Razzaghi, Srivastav, de Perio, Laney and Black6,Reference Huang, Gilkey, Thompson, Grabert, Dailey and Brewer8
This study had several limitations. The survey was self-reported. There were differences in respondents versus nonrespondents. We were unable to distinguish between vaccine refusers (‘no’) and those who are hesitant (‘unsure’) due to low numbers endorsing these responses. The generalizability of our results may be limited because attitudes and political beliefs are associated with COVID-19 vaccination and vary by geographic location. Reference Dhanani and Franz9
Reflecting our results on most common reasons for booster hesitancy or refusal, more data and effective communication are key for continued efforts to improve COVID-19 vaccination overall and booster uptake. This is specifically important for those who still had not received a booster vaccine as of April 2022. Continued understanding of hesitancy in HCWs and providing clear communication as vaccine guidelines change is necessary to encourage uptake in the population.
Acknowledgments
The opinions, results, and conclusions reported are those of the authors. No endorsement by Northwestern Medicine or any of its funders or partners is intended or should be inferred.
Financial support
This work was supported by the Northwestern University Clinical and Translational Sciences Institute (grant no. UL1TR001422), the Northwestern Memorial Foundation, and the Peter G. Peterson Foundation Fund.
Conflicts of interest
All authors report no conflicts of interest relevant to this article.