Page 150 - GHES-3-2

Global Health Economics and
Sustainability
Vaccine hesitancy in the US, India, and China

these states, which were divided into red, red/blue or The age is another important criterion; the maximum
swing states, or blue states, depending on the results of age is 90 and the minimum age is 20 in the data set. We
presidential elections over the past decade. All missing calculated the age from the year of birth to the current year
values were omitted, resulting in a final dataset of 5,758 and saw that old individuals are the most hesitant, followed
respondents. by the mid-age group and young groups. When we combine

Table 2 describes the summary statistics of baseline both unsure and hesitant groups, both mid and old-age
characteristics among the three hesitancy groups. The Chi- groups are at 94.1% of vaccine hesitancy, while the young
square test of significance and p-values are reported. Gender, group is at 92.9%. These numbers are in the ballpark, so a
MH services, income, race, marital status, education, and clearer conclusion about the age and hesitancy cannot be
stress indicators such as anxiety, worry, interest, being down, made. There are other studies that support this conclusion.
region, visual impairment, mobility, and remembering were 4.2.1. Model 1: MLR
statistically significant (p < 0.05). Figure 1 shows the counts
of the vaccine hesitancy status of the households in the US. Considering previous literature and other vaccine hesitancy
The maximum proportion of the households was “Hesitant,” studies, we selected the above-mentioned factors that
followed by “Unsure” about taking a vaccine. would contribute to understanding the outcome of interest
(vaccine hesitancy). Model 1 considers the full model,
4.2. Primary results including all the samples and the 17 chosen variables.
Those who did not receive MH services exhibited higher We considered the variables to be significant if
percentages of vaccine hesitancy (strongly hesitant+ p < 0.05 and with OR >1 compared to the reference
hesitant = 61%) than those who received MH services group. Furthermore, we created a joint p-value called a
(strongly hesitant+ hesitant = 56%) (Figure 2A). Among combined p-value that gathers the information from both
the vaccine-hesitant groups, transgender individuals had comparisons and is utilized for testing the overall result.
the highest level of hesitancy, followed by males and females The cut-off for this combined p-value is also 0.05.
(Figure 2B). Individuals receiving MH (Figure 3B) services In Model 1, the MLR analysis is carried out, in which
appeared to be less vaccine-hesitant than those who did not we have considered the “Hesitant” and “Strongly hesitant”
receive any MH services. The West and Midwest regions categories in one group. This is the full model. The
were mostly hesitant toward vaccines, followed by the computational algorithm converged, thereby providing
northeast and south (Figure 4A). Graphically, it seemed the estimates of the coefficients. In Model 2, the BLR was
that middle-and low-income groups were more likely to applied to two groups. In Model 3, the dataset was split
be unsure and hesitant than high-income groups, while into training and testing datasets, and the prediction
White individuals were more hesitant, followed by other accuracies, sensitivities, and specificities were reported.
racial groups and Black individuals (Figure 5A and B). In all the models, “Not hesitant” was considered as the
Those who reported “not at all” anxious and “not at all”
losing interest were more likely to be vaccine-hesitant reference category. Results from Model 1 MLR relating
(Figure 8B). Those with some degree/being high school sociodemographic and health characteristics to the odds
graduates were also more likely to be vaccine-hesitant of belonging to three hesitancy classes (n = 5758). The OR
estimate, 95% confidence interval (CI), and the p-values
(Figure 6A). In addition, females receiving MH services corresponding to the Wald test are reported. Reference
were more likely to be hesitant/unsure about vaccines categories are in parenthesis in Table 3.
(Figure 3B). The distribution of vaccine hesitancy status
across the US is shown in Figure 7. Race is a major criterion Males were more likely to be hesitant or unsure about
in assessing the variability of responses toward vaccine vaccines than females. Individuals belonging to the
hesitancy. In our study, we found that 93.4% of Whites Northeast and Southern regions are more vaccine hesitant/
were hesitant or unsure about vaccines, whereas 87.2% unsure than those in the Midwest. Low- and middle-
of Blacks fell into the same categories. Notably, when we income groups of individuals were more likely to be
separated the percentages of hesitant and unsure groups, hesitant or unsure about vaccines than high-income groups.
we found that Blacks were more unsure about the vaccine Unmarried individuals were vaccine pro than married
than Whites. This discrepancy may vary from study to individuals. Asians are more likely to be in the “Not hesitant”
study; here, we are dealing with raw survey values rather groups than other races. Those who are not at all down or
than weighted values by population. Therefore, this may worried are more likely to be in the “Not hesitant” group
be one of the limitations and thus not representative of the than in the “Hesitant” group. High school students, those
broader population or align with conclusions from other having some education, no degree or graduate degree, were
studies. more likely to be in the “Hesitant” or “Unsure” group than

Volume 3 Issue 2 (2025) 142 https://doi.org/10.36922/ghes.2958