Introduction
The Veterans Health Administration (VHA) has a unique responsibility to protect patients from COVID-19. Previous, non-VA based, studies have shown that being African American, Reference Gardner, Brown, Vargas-Correa, Weaver, Rubinstein and Gordon1 Hispanic, Reference Khubchandani, Sharma, Price, Wiblishauser, Sharma and Webb2 and younger Reference Fisher, Bloomstone, Walder, Crawford, Fouayzi and Mazor3 were associated with decreased vaccination. Vaccinated Veterans were more often male, older, and had more comorbidities associated with serious COVID-19 complications than unvaccinated Veterans. Reference Haderlein, Wong, Jones, Moy, Yuan and Washington4 While these previous reports provided important information about demographic characteristics associated with Veterans’ COVID-19 vaccination, attitudes and beliefs regarding COVID-19 vaccination were not explored. This project aimed to assess Veteran’s perceptions and beliefs surrounding COVID-19 vaccination and its association with COVID-19 vaccination status.
Methods
Two independent cross-sectional surveys (Wave 1, Wave 2) were conducted. These surveys were designated as quality improvement by the Edward Hines, Jr. VA Hospital Institutional Review Board. Wave 1 surveys were mailed and collected between March 4 – May 19, 2021. Wave 2 was mailed and collected between May 15 – August 13, 2021. Survey responses were combined across waves for this analysis as there was no difference between surveys in factors associated with vaccination.
Wave 1 Veterans were identified from VA facilities with ≤4% COVID-19 vaccination rates (from the VHA Support Service Center (VSSC) “COVID Vaccine Surveillance Dashboard”) as of February 1, 2021. Veterans with VA utilization within the previous 2 years, not in long-term care, with valid US mailing addresses, and complete information on race and ethnicity were included. A stratified random sample of 2,500 Veterans was created with 500 Veterans in each of the following racial/ethnic categories: Non-Hispanic White, Non-Hispanic Black, Hispanic, Native American, and all others.
Wave 2 Veterans were identified from VA hospitals with COVID-19 vaccination rates of ≤33% as of April 26, 2021. Inclusion criteria were the same as Wave 1. No Wave 1 Veterans were included in Wave 2. A final sample of 1,000 Veterans stratified into 500 males and 500 females (to ensure adequate representation of females and identify any possible sex differences), with the five race/ethnicity groups from Wave 1 evenly distributed within each sex was created.
The survey included questions on current COVID-19 vaccination status, willingness to receive a two-dose COVID-19 vaccine, perceived safety and effectiveness of available COVID-19 vaccines, general attitudes regarding vaccination, and reasons for or against COVID-19 vaccination. These items were drawn from co-investigators’ previous survey efforts Reference Han, Scharnetzki and Scherer5,Reference Scherer, Knaus, Zikmund-Fisher, Das and Fagerlin6 and prior frameworks for assessing vaccination attitudes and intentions. Reference Janz and Becker7,Reference Betsch, Schmid, Heinemeier, Korn, Holtmann and Böhm8 Returned surveys were entered and managed using REDCap (Research Electronic Data Capture). Age, sex, geographic region, and rurality were extracted from the VA Corporate Data Warehouse (CDW) and merged with survey data.
The main outcome was vaccination status. If participants indicated at any point on the survey that they had received the COVID-19 vaccine, they were classified as vaccinated. All others were classified as “not vaccinated.” The association between demographic characteristics and survey responses was evaluated by vaccination status using chi-square/Fischer’s exact tests for categorical variables and t-tests for continuous variables. Multivariable binomial logistic regression was used to assess the independent association between these factors and vaccination status. Variables with a p-value less than 0.05 were included in the final model. Data were analyzed using Stata MP software, version 17 (StataCorp, College Station, TX, USA).
Results
There was no difference between the recruitment facilities and non-recruitment facilities by geographical region, facility complexity, or urban/ rural designation for Wave 1. There were more mid-level and Southern facilities represented among Wave 2 recruitment facilities. 1,080 (30.8%) Veterans responded: 818 in Wave 1; 262 in Wave 2. Respondents were male (80%) and non-Hispanic White (31%). The mean age was 64.2 years (SD = 14.1). Non-respondents were more likely to be under 50 years old (47.2% versus 15.7%, respectively; p < .001), female (27.5% versus 19.7%; p < .001), and Black non-Hispanic (21.1% versus 17.6%) or Hispanic (21.7% versus 16.3%; p < .001). Overall, 65% of Veterans reported receiving a COVID-19 vaccination (Wave 1: 60%, Wave 2: 81%). Unvaccinated respondents were more likely to be female (22.5% vs 18.2% (p = 0.094)) and under age 50 (28% vs 9% (p < 0.001)) (Table 1).
Table 1. Respondent demographics and comorbidities by survey wave and vaccination status (N = 1,080)
Respondents mostly believed that the COVID-19 vaccines were effective (62.8%) and safe (76.0%). However, unvaccinated respondents more often believed COVID-19 vaccines were ineffective (21.4% versus 2.3%, p < 0.001), unsafe (21.2% versus 0.9%, p < 0.001), and were more worried about vaccine side effects (32.0% versus 11.3%, p < 0.001) than vaccinated respondents (Supplemental Table 1). Unvaccinated respondents more frequently expressed disagreement about the general safety of vaccines compared to vaccinated respondents (14.6% versus 2.4%, p-value < 0.0001).
Unvaccinated respondents reported logistical concerns about receiving a two-dose vaccine more frequently than vaccinated respondents (Supplemental Table 1). Finding time to schedule a second appointment (11.9% vs 3.4%, p-value < 0.0001) having to schedule two appointments within 30 days (14.8% vs 5.0% p-value < 0.0001) and concerns that a second dose of the vaccine would increase the risk of side effects (27.5% vs 8.7% p-value < 0.0001) were the primary logistical concerns identified by unvaccinated Veterans.
In multivariable binomial regression modeling, factors associated with increased odds of being unvaccinated were being under 50 years of age (OR 7.68, 95%CI = 4.36, 13.51), negative (OR = 3.88, 95%CI = 1.52, 9.90), or neutral (OR = 1.54, 95%CI = 1.02, 2.30) feelings about vaccines in general, and disagreement with the statement “I am confident vaccines are safe” (OR = 2.30, 95%CI = 1.07, 4.93). Logistical difficulties including finding transportation (OR = 1.95, 95%CI = 1.01, 3.45), time (OR = 2.20, 95%CI = 1.18, 4.08), and need for two appointments within 30 days to get a second dose of a vaccine (OR = 2.64, 95%CI = 1.58, 4.44) were all associated with being unvaccinated. Unvaccinated respondents had higher odds of selecting “If my doctor recommended it” as a reason for getting a COVID-19 vaccine (OR = 1.44, 95%CI = 1.01, 2.04) (Table 2).
Table 2. Multivariable, adjusted binomial logistic regression models examining the association of demographics and attitudes and beliefs with not being vaccinated
Discussion
This survey of 1,080 Veterans found that 65% of respondents were vaccinated. Unvaccinated respondents were younger, had greater concerns about COVID-19 vaccine safety/side effects, and negative attitudes regarding vaccines in general. Our results are consistent with previous findings from the Veteran Reference Fisher, Bloomstone, Walder, Crawford, Fouayzi and Mazor3 and non-Veteran populations Reference Fisher, Bloomstone, Walder, Crawford, Fouayzi and Mazor3,Reference Kelly, Southwell and McCormack9 that younger age was associated with being unvaccinated against COVID-19. Unique to our study was the fact that logistical concerns around getting two vaccines were also associated with being unvaccinated. Potential solutions to this issue could include the expansion of programs designed to help Veterans with transportation to and from healthcare appointments including deploying community vaccination vans.
There are several strengths to this evaluation. We were able to survey Veterans from a geographically diverse section of the United States. Also, due to our sampling strategy, we were able to assess differences in race/ethnicity as well as sex and their association with vaccination. Our evaluation also had some limitations. We had lower response rates among those of racial and ethnic minorities which may have impacted our ability to detect vaccine intentions in those groups. Additionally, vaccinated individuals may have been more likely to respond than unvaccinated individuals.
This study identifies key attitudinal and belief factors associated with COVID-19 vaccination among Veterans. This prospect may have key implications for increasing uptake of the bivalent mRNA booster, which has only been received by 15.5% of the eligible population 10 . Finally, the role of practical barriers to receiving two-dose vaccines indicates that removing logistical challenges to vaccination will be key for facilitating uptake as the need for regular boosters or new vaccinations continues. Considering the critical role that vaccination will continue to play, not only for COVID-19 but for other communicable diseases, it is important to continue to make strides to improve the acceptance of vaccines.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/ash.2023.446
Financial support
This work was supported by The Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Quality Enhancement Research Initiative QUE 15-269 and RCS 20-192 (Evans, Charlesnika). Dr. Scherer was supported by a K01 award through the NIH National Insititue on Aging (1K01AG065440).
Competing interests
The authors have no conflicts of interest to disclose.
Disclaimer
The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs, the US government, or the National Institues of Health.
Open access
