Vaccination against viral pathogens can markedly improve public health [1, 2]. However, vaccine hesitancy, a delay or refusal of vaccination despite the availability of safe and effective vaccines, has recently increased[3–5] and has been particularly prevalent for COVID-19 vaccines [6–8].

COVID-19 vaccines became widely available in the United States (US) in early 2021 [9–14]. The vaccines are safe and highly effective in preventing COVID-19 associated severe disease[12, 14–20]. Despite widespread availability, approximately 18.6% of eligible persons in the US had not received a primary COVID-19 vaccine series, and 83% had not received the updated bivalent COVID-19 vaccine by May 2023 [5, 21]. Understanding why some people decline COVID-19 vaccination is crucial to increasing vaccination rates and maintaining a highly vaccinated population against respiratory viral illnesses, especially when serial vaccinations are necessary [22].

Well-described strategies to decrease hesitancy for seasonal influenza vaccination may inform approaches to overcome COVID-19 vaccine hesitancy, especially if risk factors for being unvaccinated and vaccine hesitancy for COVID-19 and influenza vaccines are similar[23]. These include community outreach programs and worksite interventions,[24] educating patients through various tailored methods,[25–27] training physicans to improve communication with patients,[25, 28] and expanding access to vaccine services.[29] To evaluate the similarities and differences in vaccine hesitancy between COVID-19 and influenza vaccines, we evaluated risk factors for being unvaccinated and patient perceptions related to these vaccines among US adults hospitalized with acute respiratory illnesses between 2019 and 2022.

A total of 16,272 adults hospitalized with an acute respiratory illness were prospectively enrolled. Self-reported influenza or COVID-19 vaccination status was unavailable for 2997(18.4%), resulting in an analytical population of 13,275 patients, including 12,292 (92.6%) patients in the COVID-19 vaccine assessment period (January 1, 2021 – January 25, 2022) and 983 (7.4%) patients in the influenza vaccine assessment period (October 10, 2019 – December 31, 2020) (Figure S1). Our analysis focused on understanding vaccine hesitancy during this critical period, when initial responses to COVID-19 vaccines were forming, and offers valuable insights into the early drivers of vaccine acceptance.

Of 12,292 inpatients during the COVID-19 vaccine assessment period, 5637 (45.9%) tested positive for SARS-CoV-2, 6655 (54.1%) tested negative for SARS-CoV-2, 6807 (55.4%) reported receiving at least one COVID-19 vaccine, and 5485 (44.6%) reported being unvaccinated for COVID-19 (Table 1). The median age was 60 years (IQR 47–71), 5939 (48%) were female, 2549 (21%) were Black, and 2033 (17%) were Hispanic. In addition, 4421 (36%) had at least 3 chronic medical conditions.

Of 983 inpatients during the influenza vaccine assessment period, 335 (34.1%) tested positive for influenza, 648 (65.9%) tested negative for influenza, 602 (61.2%) reported receiving the influenza vaccine, and 381 (38.8%) reported being unvaccinated for influenza (Table 1). The median age was 59 years (IQR 45–70), 476 (48%) were female, 219 (22%) were Black, 100 (10%) were Hispanic, and 384 (39%) had at least 3 chronic medical conditions.

In multivariable regression, patient characteristics significantly associated with not receiving the COVID-19 vaccine after adjusting for other covariates included younger age, female sex, increasing BMI, lack of insurance, absence of chronic comorbid medical conditions, no or rare influenza vaccination in prior years, higher SVI, living in the Midwest or Southern US, lack of college or higher education and not wearing a mask (Figure 1a, Table S5). Patient characteristics significantly associated with not receiving influenza vaccination after adjusting for other covariates included a higher SVI, being a current smoker, no or rare influenza vaccination in prior years, and no or one chronic comorbid medical condition (Figure 1b, Table S6).

Bivalent COVID-19 vaccines against the ancestral, Omicron BA.4/BA.5 and XBB.1.5 variants, have been shown to be safe and effective [20, 40, 41]. The uptake of the bivalent COVID-19 vaccine remains low, and as of May 2023, 18.6% of eligible persons in the US had not received a primary COVID-19 vaccine series, and 83% had not received the updated (bivalent) COVID-19 vaccine.[5, 21] Multiple factors have led to vaccine hesitancy, such as vaccine fatigue, perceived immunity after infection, lack of patient education on eligibility, vaccine availability, and perceived safety of the vaccination [8, 37, 38, 42]. Understanding vaccine hesitancy for the primary COVID-19 vaccine series and comparisons with influenza may help design interventions to improve vaccine uptake for future pandemics.

In this prospective analysis from the IVY Network, we identified similarities in vaccine hesitancy between the primary COVID-19 vaccine and the influenza vaccine. COVID-19 vaccine hesitancy was associated with younger age, female sex, absence of comorbid chronic conditions, no or rare prior influenza vaccination, and social factors such as lack of insurance, high school or lower education, and higher social vulnerability. Similarly, influenza vaccine hesitancy was associated with having one or fewer comorbid chronic conditions, no or rare influenza vaccination in prior years, and higher social vulnerability.

However there were notable differences between the two vaccines. Age, sex, BMI, and insurance status were associated with being unvaccinated for the COVID-19 vaccine but did not reach statistical significance for being unvaccinated for the influenza vaccine. Smoking was associated with influenza vaccine hesitancy but not with COVID-19 hesitancy. There were regional variations in COVID-19 vaccine hesitancy with lower acceptance in the Midwest and Southern US, while there were no significant regional variations for influenza vaccine.

A significant finding of our analysis is that higher proportions of patients vaccinated for influenza “every or most years” were vaccinated for both influenza and COVID-19 (Figures 1a and b). While early in the pandemic, logistic limitations were a significant barrier to vaccination, personal choice became the dominant factor over time (Figure 2). A discussion with a healthcare provider was identified as a more common reason for accepting the influenza vaccine than the COVID-19 vaccine, and in a small minority of cases cited as a reason for declining the COVID-19 vaccine more often than the influenza vaccine.

In our study, healthcare provider recommendations significantly influenced vaccine uptake, particularly for the influenza vaccine. However, since respondents could identify only one primary reason for vaccination, the role of healthcare provider recommendations may be understated. The higher proportion of individuals citing ‘personal protection’ as their primary reason for receiving the COVID-19 vaccine likely reflects an elevated perception of COVID-19-related risk rather than a diminished influence of healthcare providers. Additionally, while a small proportion of respondents reported that their healthcare provider advised against COVID-19 vaccination (2.2%), this finding may not be meaningful given the low numbers.

Our data have several implications for overcoming vaccine hesitancy. First, individuals frequently rely on healthcare providers as trusted and credible sources for vaccine information[3, 43]. In our study, healthcare provider discussion was the primary reason for 27% of respondents to accept the influenza vaccine and 8.3% of patients to accept the COVID-19 vaccine, and these differences in numbers were significant. Notably, healthcare providers advised a small proportion, 0.5%, of patients to decline the influenza vaccine and 2.2% to decline the COVID-19 vaccine, presumably due to providers’ concerns regarding the vaccine for those individuals. In addition, 6.1% of COVID-19 hesitant patients chose not to receive the vaccine due to worries about underlying medical conditions that their healthcare provider could have also addressed. Our data is similar to a study by Garza et al., which showed that trusting advice from a physician or healthcare organization increased the odds of receiving the COVID-19 vaccine[43]. Partnering with healthcare providers may significantly reduce vaccine hesitancy and improve acceptance for influenza and COVID-19. A factor which would be important in future pandemics.

Second, prior vaccination influences future behaviors.[43] In our cohort, a history of vaccination for influenza “every year or most years” was highly associated with both influenza and COVID-19 vaccination, suggesting the continued importance of influenza vaccination discussions. Our findings suggest that a history of influenza vaccination may influence future vaccine behaviors, including COVID-19 vaccination. Those vaccinated for influenza “every year or most years” were significantly more likely to receive both influenza and COVID-19 vaccines. This supports the notion that addressing influenza vaccine hesitancy may indirectly improve uptake of other vaccines, including the COVID-19 vaccine and any future pandemics. This is consistent with other studies demonstrating an association between influenza and COVID-19 hesitancy.[43, 44]

Third, in our sample, COVID-19 vaccine hesitancy was higher in younger individuals, females, those with lower educational attainment, and people living in areas with high social vulnerability. Increasing educational efforts in these high-risk groups may help address vaccine hesitancy, as seen in other studies [45].

Our data identified differences in COVID-19 vaccine acceptance and hesitancy based on insurance status and age, which has been noted in other studies [46–49]. A 2020 Medicare beneficiary survey showed that up to 40% of Medicare beneficiaries were hesitant about getting a COVID-19 vaccine, however, unlike our study population, the hesitancy was greater in racial/ethnic minorities [50].

Additionaly as seen in other studies, socio-political factors like misinformation and trust in the scientific community may also contribute to hesitancy [7]. Approximately 11.6% of our cohort did not receive the COVID-19 vaccine due to safety concerns, a figure that aligns with other studies. Vaccine hesitancy may be related to trust in the scientific community, with a higher level of trust being associated with higher vaccination rates.[51] Basagra et al. found racial differences between groups. In their sample, Asian and Caucasian groups had a greater trust in the government and scientific community than Native Americans and African Americans. In a survey of 1,068 stable adult ED patients without active COVID-19, 31.6% of ED patients reported vaccine hesitancy.[52] In a multivariate regression analysis, Black race (odds ratio [OR] 4.24, 95% confidence interval [CI], 2.62 – 6.85) and younger age (age 18–24 years OR 4.57, 95% CI, 2.66 – 7.86; age 25–35 years OR 5.71, 95% CI, 3.71 – 8.81) were independently associated with hesitancy, to a greater degree than the level of education (high school education or less OR 2.27, 95% CI, 1.23 – 4.19). In addition, hesitant patients were significantly less likely to trust governmental sources of vaccine information than non-hesitant patients (39.6% v. 78.9%, p < 0.001); a smaller difference was noted in the domain of trust toward friends and family (51.1% v. 61.0%, p = 0.004). Hesitant patients also cited vaccine safety concerns and a perception of insufficient supportive research.

In a survey of approximately 1000 adults in April 2020 for the COVID-19 vaccine, 57.6% of participants (n = 571) intended to be vaccinated, 31.6% (n = 313) were not sure, and 10.8% (n = 107) did not intend to be vaccinated.[53] Factors independently associated with vaccine hesitancy included younger age, African-American race, lower educational attainment, and not receiving the influenza vaccine in the prior year. In addition, other reasons for vaccine hesitancy included vaccine-specific concerns, a need for more information, antivaccine attitudes or beliefs, and a lack of trust.

Similar to our study, vaccine hesitancy in Canada was significantly associated with younger age (18–39 years), lower education, and non-liberal political leaning [54]. In this study, the vaccine hesitant were less likely to believe that a COVID-19 vaccine could end the pandemic or that the benefits of a COVID-19 vaccine outweighed the risks [54]. Like our cohort, individuals with vaccine hesitancy had a higher prevalence of being concerned about vaccine side effects, a lower prevalence of being influenced by peers or healthcare providers, and a lower prevalence of trust in government institutions. Similarly, in a survey of 1,000 people in Nevada, 30.4% exhibited vaccine hesitancy, which was higher in younger individuals and middle-aged females, African Americans, lower-income, and conservative individuals [55].

Our data are also consistent with our prior work, showing that amongst individuals hospitalized with severe acute respiratory illness from 2019–2020, 38.9% reported not receiving the influenza vaccine [56]. In the unvaccinated group, 51.1% reported not receiving the vaccine due to perceived ineffectiveness or risks of influenza vaccination. Like our current study, lack of medical insurance and, unlike our current data, Black or Hispanic race were associated with higher rates of vaccine hesitancy.

Our study had a few notable limitations. We used self-reported vaccination status, which may introduce reporting bias. However, self-reported vaccination status has been congruent with biological tests, suggesting that it can be a valid vaccination measure [57, 58]. We surveyed hospitalized patients who may not be fully representative of the general public, limiting the generalizability of our findings. We excluded subjects 67 subjects unwilling to discuss reasons for their COVID-19 vaccine hesitancy. This may limit the diversity of perspectives captured. In addition, participant responses may have been affected by social desirability bias [37, 59]. In our survey participants could only select one primary reason for vaccination, which likely limits the accuracy of capturing the full range of motivations and likely underestimates the influnace of healthcare providers. Our models adjust for various risk factors, but it is possible that other unmeasured factors influencing hesitancy were not included. Despite discordant sample size, both influenza and COVID-19 cohorts should have sufficient power to identify risk factors, though we focused on similarities and differences based on point estimates. Finally our data were collected during the early phase of the pandemic, a period characterized by high levels of public concern regarding COVID-19 and strong expectations surrounding the impact of vaccination. Since then, public opinions and vaccine behaviors have shifted significantly. This shift limits the direct applicability of our findings to the present, though our study provides valuable insights into the initial drivers of vaccine hesitancy.

Vaccine hesitancy remains a significant challenge in the US particularly among certain socioeconomic and demographic groups. Our analysis showed that individuals with a history of regular influenza vaccination were more likely to choose to be vaccinated for both influenza and COVID-19, and infrequent vaccination for influenza was associated with vaccine hesitancy for both influenza and COVID-19. Therefore, efforts to improve influenza vaccination may indirectly improve COVID-19 vaccination rates, including the new generation of COVID-19 vaccines. We observed regional variations in vaccine acceptance for COVID-19, vaccine hesitancy more prevalent among younger individuals, females, the uninsured, those with lower education levels and socially vulnerable populations. Targeted outreach and education efforts directed at may be crucial to improving vaccination rates. Finally, healthcare provider recommendations played an important role in patients decisions to accept or decline vaccination. Ensuring consistent, clear and proactive messaging from healthcare providers may be a key strategy in addressing vaccine hesitancy for both influenza and COVID-19.