COVID-19 vaccination and race – A nationwide survey of vaccination status, intentions, and trust in the US general population

Abstract

BACKGROUND:

COVID-19 vaccine hesitancy for adults and children varies depending on societal factors, race, and trust ascribed to the source of vaccine information.

OBJECTIVE:

To assess COVID-19 vaccination rates and trust levels for vaccine information by race at 2 time points.

METHODS:

Online cross-sectional data from US adults were collected in February/March 2021 (T1) and November 2021 (T2). Questions included vaccination status, reasons for vaccine refusal, trust levels for vaccine information and the Wake Forest Physician Trust Scale. At T2, parents were asked about vaccination status of children aged 12-18 years and intent for children aged 5-11 years. Vaccination rates and trust levels for vaccine information were measured. Multivariable logistic regression was used to identify characteristics predictive of receiving COVID-19 vaccination.

RESULTS:

Vaccination rates were 20.2% and 70.8% at T1 and T2, respectively. At T1 and T2, higher proportions of White (23.2% and 72.0%) and Other race (14.4% and 75.2%) respondents were vaccinated relative to Black respondents (9.6% and 64.4%) (P < 0.05). In descending order, respondents’ doctors, family members, and pharmacists were the most trusted information sources. Black parents, relative to White and Other parents with unvaccinated children aged 12-18 years or who were not very likely to vaccinate younger children, reported lowest physician trust (P < 0.01).

At T1, being married, college educated, and older and having greater Wake Forest Physician Trust Scale scores and a higher number of comorbidities predicted a higher likelihood of being vaccinated. Being Black, having a median household income less than $100,000, and residing in the Northeast or Midwest, relative to the West, predicted a decreased likelihood of being vaccinated. At T2, race and comorbidities were no longer predictive of vaccination.

CONCLUSIONS:

Racial variation in vaccination status decreased from T1 to T2. Physician trust predicted vaccination status and intent regardless of race. Respondents’ doctors, family members, and pharmacists are trusted sources of vaccine information, and targeting these influencers may reduce vaccination hesitancy.

Plain language summary

COVID-19 vaccine is a crucial way of combating the global pandemic. Although vaccine uptake is a personal choice, it highly impacts the society’s immunity. We surveyed US adults at the beginning and end of 2021 about their vaccine status and that of their children. We found that COVID-19 vaccine uptake increased over time and racial differences in vaccine uptake decreased. Physician trust was an important predictor of vaccination status and intent regardless of race.

Implications for managed care pharmacy

COVID-19 vaccination rates increased from approximately 20% in February/March 2021 to 71% in November 2021. At both times, proportions of vaccinated individuals were lower among Black individuals, as compared with individuals of other races, though racial disparities in uptake diminished over time. Physician trust predicted vaccination status and parental intent to vaccinate their children. Initiatives to reduce vaccine hesitancy should involve doctors, family members, and pharmacists—the most common trusted sources of vaccine information.

The COVID-19 pandemic has affected the lives of millions of people around the world. The Centers for Disease Control and Prevention (CDC) recorded more than 980,000 COVID-19-related deaths in the United States since the start of the pandemic.¹ Various strategies have been implemented in the United States to reduce the spread of COVID-19, but vaccination is considered the most effective medical approach to decrease the risk and severity of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection.^2-4 However, its effectiveness may be limited by vaccine hesitancy.⁵ At the time of the study, 3 COVID-19 vaccines were authorized for emergency use by the US Food and Drug Administration (FDA): Pfizer-BioNTech, Moderna (both have since received full approval), and Johnson & Johnson’s Janssen (J&J/Janssen).^6-8 As of April 9, 2022, more than 565 million vaccine doses have been administered in the United States, and 218.3 million (65.8%) people are fully vaccinated and 98.6 million have received a first booster dose.⁹ Nonetheless, vaccine acceptance is based on trust, access, need, and value. Vaccine hesitancy is influenced by societal factors, race, and trust in the source of vaccine information, as well as understanding of the vaccination process.^10-11 It remains unclear which factors are most influential when consenting to COVID-19 vaccination.

Medical mistrust is multifaceted and leads to underutilization of health services, less involvement in biomedical research, and low adherence to medical services.¹² Medical mistrust is believed to be one of the prominent reasons for lingering health disparities in the United States, especially among African American individuals,¹³ and is likely influential in vaccine hesitancy. A recent study found that 1 in 5 adults were vaccine hesitant, with highest hesitancy among people with low trust in vaccines (56%), who identify as African American (50%), and with annual incomes of less than $25,000K (31%) among other factors.¹⁴ These findings support the potentially negative influences of vaccine mistrust and sociodemographic correlates on COVID-19 vaccine uptake.

COVID-19 vaccine hesitancy among adults who are also parents or guardians may also be a barrier to vaccinating children. It has been increasingly documented that the number of parents who have avoided vaccinating their children base their decision on faulty scientific information.¹⁵ In 2021, the CDC recommended everyone older than 5 years get vaccinated to help protect against the spread SARS-CoV-2. Early reports showed significant levels of vaccine hesitancy among parents for vaccinating children, especially so for younger children.^16-18 The Kaiser Family Foundation’s ongoing COVID-19 Vaccine Monitor project reported that between November 2021 and January 2022 the share of parents whose children received at least 1 dose of COVID-19 vaccine has increased. Sixty percent of parents of children aged 12-17 years (up from 49% in November 2021) reported their children received at least 1 dose, whereas one-third of parents of children aged 5-11 years did so (up from 16% in November 2021).¹⁹ Vaccination uptake among children has significantly slowed,²⁰ and extant research points to parents’ vaccine hesitancy and safety concerns as major contributors to the low and slow uptake in children.^16,17,21 Additionally, lack of trust in the government and a belief that the vaccine was not needed were prominent among parents of unvaccinated children.^18,22

Here, we present the findings of a study investigating the levels and correlates of vaccination and intent against COVID-19 across the United States among adults and children. Though the literature is expanding around COVID-19 vaccination, there is limited information about the relationship between the vaccination and trust factors. We assessed COVID-19 vaccination rates by race and societal factors and identified trusted sources of vaccine information at 2 time points. Children’s vaccination status and parental intention to vaccinate and their trust in the vaccine development process were also examined. Our national findings build on the burgeoning literature on COVID-19 vaccine uptake in the United States and its potential predictors.

Methods

Data were collected through an online cross-sectional survey of US adults fielded in February/March of 2021 (time point 1 [T1]) and November 2021 (time point 2 [T2]). Participants were unique at each time point (ie, participants who completed the survey at T1 did not complete the survey at T2). A national online survey was conducted for its reach, convenience, and low cost. Eligible respondents were aged 18 years or older, resided in the United States, and confirmed their voluntarily agreement to participate. The survey was open to the general population.

Participants were recruited through online research panels. Multiple quality control processes integrated throughout data collection, including digital fingerprinting technologies, verified that data comprised unique and validated non-fraudulent respondents. The survey was deemed exempt from Institutional Review Board approval, as all responses were anonymized and aggregated and could not be related back to participants.

The survey was conducted at 2 timepoints (T1 and T2) to identify changes overtime. At T2, participants who were a parent or guardian of children were asked about their children vaccination status, since the vaccine had been approved for children by T2. Another difference between T1 and T2 was that participants at T1 were asked about their action plan regarding receiving COVID-19 vaccine, whereas at T2 they were asked about booster vaccination and their action plan towards boosters.

COVARIATES

All participants completed questions on demographics, comorbid conditions, body mass index (BMI), smoking status, region, and the level of trust attributed to different sources of COVID-19 information. Participants completed a comorbidity checklist used to calculate the Charlson Comorbidity Index [CCI] score (range 0-27). A higher CCI score indicates greater comorbid burden.²³ Race categories were American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Pacific Islander, White, or Other. For analysis, race was classified as the Black population (those who selected Black or African American), White population (those who selected White), and Other population (those who selected all other race categories). Income was median household income imputed from the participants’ zip code and US Census data.

OUTCOMES

COVID-19 vaccination was self-reported as having received at least 1 dose of an approved COVID-19 vaccine. At T1, vaccines by Pfizer and Moderna were authorized for emergency use by the FDA, and by T2, J&J/Janssen vaccine was available.

Trust in physicians was measured using the Wake Forest Physician Trust Scale (WFPTS), a 5-item, 1-dimensional scale²⁴ consisting of questions concerning trust in their doctor [scored 1 (lowest trust) to 5 (highest trust)]. Physician trust, using the WFPTS, has been shown to predict patient behavior in other health domains (eg, colorectal screening).^25,26 Responses were scored, summed, and rescaled from 0 (lowest trust) to 100 (highest trust) for ease of interpretation.

Respondents also rated their trust in COVID-19 vaccination information from 0 (no trust at all) to 10 (complete trust) for the following sources: family members, friends, nonmedical leaders in your community, your local pharmacist, your doctors, governor of your state, former President Donald Trump, President Joseph Biden, Dr Anthony Fauci, and pharmaceutical companies.

Parents/guardians were asked if their children aged 12-18 years were vaccinated and, for those indicating their children were not vaccinated, the reasons for not receiving the vaccination. In addition, they were asked about the likelihood of vaccinating their children aged 5-11 years. Their levels of trust in the vaccine development process for children in general and in the governmental approval process to ensure the safety of a COVID-19 vaccine for children were assessed in addition to their physician trust (described above).

STATISTICAL ANALYSIS

Baseline characteristics at T1 and T2 were summarized using descriptive statistics in SPSS version 28. Chi-square and Fisher exact tests compared the distribution of categorical variables and an analysis of variance was used for continuous variables. A multivariable logistic model determined predictors of receiving the vaccination, controlling for age, sex, smoking status, race, region, marital status, education, WFPTS score, CCI and income. Regression coefficients (or their transformation [eg, odds ratios]) with 95% CIs and associated P values are reported, as well as the overall concordance statistic (C-statistic).

Results

Of the 2,404 potential participants at T1, a total of 2,098 (87.3%) completed the survey. Of the 1,359 potential participants at T2, a total of 1,174 (86.4%) completed the survey. The overall response rate across T1 and T2 was 87.0%. At T1, 47.8% (n = 1,002) were female, with a mean age of 49.72 ± 17.02 years, and at T2, 58.9% (n = 691) were female, with a mean age of 47.98 ± 17.28 years (Table 1). A comparison of participant characteristics to the US adult population from the Census is provided in Supplementary Table 1 ^{(229.2KB, pdf)} (available in online article).

TABLE 1.

Demographics of the Total Survey Population by Racial Group at Time Points 1 (February/March 2021) and 2 (November 2021)

	February/March 2021				November 2021
	Total population (N = 2,098)	Black population (N = 344)	White population (N = 1,566)	Other population (N = 188)	Total population (N = 1,174)	Black population (N = 233)	White population (N = 828)	Other population (N = 113)
Sex, %
Female	47.8	45.1	49.2	41.0	58.9	65.7	59.2	42.5
Male	52.2	54.9	50.8	59.0	41.1	34.3	40.80	57.5
Age, mean (SD)	49.72 (17.02)	49.85 (16.56)	50.76 (17.01)	40.82 (15.37)	47.98 (17.28)	40.26 (14.71)	51.22 (17.13)	40.11 (15.74)
Race and ethnicity, %
African American or Black	16.4	100.0	0.0	0.0	19.8	100.0	0.0	0.0
White	74.6	0.0	100.0	0.0	70.5	0.0	100.0	0.0
Other	9.0	0.0	0.0	100.0	9.6	0.0	0.0	100.0
Hispanic, Latino, or Spanish origin	5.5	5.5	3.6	21.8	9.1	8.6	6.3	31.0
Region, %
Northeast	16.4	13.4	17.2	14.9	18.9	17.6	19.9	14.2
Midwest	18.7	3.8	13.4	1.5	22.1	21.5	23.2	15.0
South	40.2	49.1	39.5	29.3	37.5	45.5	37.0	24.8
West	24.7	14.2	25.2	39.4	21.6	15.5	19.9	46.0
CCI score, mean (SD)	0.49 (0.98)	0.53 (1.12)	0.51 (0.99)	0.20 (0.60)	0.50 (1.22)	0.45 (1.22)	0.55 (1.23)	0.31 (1.15)
BMI, mean (SD)	27.70 (7.02)	29.60 (7.78)	27.38 (6.91)	26.80 (5.82)	28.22 (7.47)	29.45 (8.75)	28.08 (7.19)	26.73 (6.20)
Education more than high school, %	79.8	76.2	81.2	75.0	71.0	65.2	73.1	67.3
Employed ≥ 32 hours per week, %	34.3	30.2	34.8	37.8	34.6	38.6	32.4	42.5
Smoking status (yes), %	27.2	27.6	27.8	21.3	27.0	29.6	27.5	17.7
Marital status (married), %	50.3	30.8	56.6	33.5	44.0	23.2	50.8	37.2

BMI calculated as weight in pounds divided by height in inches squared.

BMI = body mass index; CCI = Charlson Comorbidity Index.

COVID-19 VACCINATION IN ADULTS

At T1, 23.2% (n = 364), 14.4% (n = 27), and 9.6% (n = 33) of White, Other, and Black respondents reported receiving the COVID-19 vaccination, respectively (P < 0.01) (Supplementary Table 2 ^{(229.2KB, pdf)} ). Of those not vaccinated, 26.0% (n = 312), 30.4% (49), and 35.0% (n = 109) of White, Other, and Black respondents did not plan on receiving it, respectively (P = 0.01). The most common reasons for not receiving the vaccination were not trusting the vaccination (65.1%, n = 306), followed by being unsure of the long-term safety of the vaccination (32.6%, n = 153) (Supplementary Table 3 ^{(229.2KB, pdf)} ). More White respondents indicated they had taken some steps to get vaccinated compared with Black and Other race respondents (P < 0.05), whereas more Other respondents indicated they had not signed up for the vaccination compared with Black and White respondents (P = 0.013) (Supplementary Table 2 ^{(229.2KB, pdf)} ).

At T2, the COVID-19 vaccination rate increased to 72.0% (n = 596), 75.2% (n = 85), and 64.4% (n = 150) of White, Other, and Black respondents, respectively (P = 0.043). (Supplementary Table 4 ^{(229.2KB, pdf)} ). Across groups, more than 80% of those vaccinated had received 2 of 2 available COVID-19 vaccinations. Of those not vaccinated, the most common reasons were not trusting the vaccination (47.8%, n = 164), followed by being unsure of the long-term safety (37.9%, n = 130). (Supplementary Table 4 ^{(229.2KB, pdf)} ).

Physician Trust. We compared WFPTS (physician trust) scores in vaccinated and unvaccinated respondents overall and by race (Table 2). The overall WFPTS score was lower at T2. At T1 and T2, WFPTS scores were lower in unvaccinated than in vaccinated respondents (P < 0.001). WFPTS scores by race were also lower at T2 and lower in unvaccinated than in vaccinated respondents, with the exception of the Other population (P = 0.199). At both T1 and T2, the Other population had the lowest WFPTS scores compared with the White and Black race groups. However, at T2, mean WFPTS scores were higher in vaccinated and unvaccinated groups in the Other population (P = 0.007), showing similar trends as other race groups.

TABLE 2.

Mean WFPTS Score by Vaccination Status for Each Racial Group at Time Points 1 (February/March 2021) and 2 (November 2021)

	February/March 2021								November 2021
	Total population (N = 2,098)		Black population (N = 344)		White population (N = 1,566)		Other population (N = 188)		Total population (N = 1,174)		Black population (N = 233)		White population (N = 828)		Other population (N = 113)
	WFPTS	P value	WFPTS	P value	WFPTS	P value	WFPTS	P value	WFPTS	P value	WFPTS	P value	WFPTS	P value	WFPTS	P value
Vaccinated	70.54 (15.83)	< 0.001	72.12 (13.41)	0.002	70.80 (15.77)	< 0.001	65.19 (18.78)	0.199	68.80 (16.15)	< 0.001	67.83 (18.54)	< 0.001	69.21 (15.64)	< 0.001	67.65 (15.1)	0.01
Unvaccinated	63.51 (18.00)		62.09 (18.18)		64.36 (17.63)		59.91 (19.82)		59.08 (19.47)		55.12 (22.18)		60.54 (18.89)		58.75 (13.7)

WFPTS range: 0 (Lowest trust score) to 100 (highest trust score). Scores displayed as mean (SD).

WFPTS = Wake Forest Physician Trust Scale.

Trusted Sources of COVID-19 Information. As shown in Table 3, respondents trust their own doctors the most at T1 and T2 (mean = 7.90 ± 2.36 and mean = 7.86 ± 2.53, respectively). Respondents also indicated numerically higher levels of trust in information from family members and local pharmacists at T1 and T2 relative to other sources. In general, White respondents exhibited significantly higher trust scores than Black and Other race respondents at both T1 and T2, except in 3 categories: trust in President Biden and Dr Fauci were highest in Black respondents at T1 and T2, and pharmaceutical companies were highest for Other respondents at T2. The lowest scores across all race groups were reported for trust in former President Trump across T1 and T2, except for White respondents at T2, where President Biden received the lowest score.

TABLE 3.

Mean “How Much Do You Trust” Score for COVID-19 Vaccination Information Sources Across Racial Groups at Time Points 1 (February/March 2021) and 2 (November 2021)

	February/March 2021					November 2021
	Total population (N = 2,098)	Black population (N = 344)	White population (N = 1,566)	Other population (N = 188)	P value	Total population (N = 1,174)	Black population (N = 233)	White population (N = 828)	Other population (N = 113)	P value
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)		Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Family members	7.41 (2.77)	6.92 (3.06)	7.57 (2.67)	6.99 (2.84)	< 0.001	7.57 (2.78)	7.23 (2.97)	7.71 (2.66)	7.20 (3.15)	0.020
Friends in your community	6.35 (2.60)	5.49 (2.78)	6.61 (2.50)	5.78 (2.69)	< 0.001	6.24 (2.75)	5.50 (2.99)	6.45 (2.65)	6.25 (2.70)	< 0.001
Nonmedical leaders in your community	5.13 (2.78)	4.93 (2.83)	5.22 (2.75)	4.74 (2.83)	0.030	5.24 (2.80)	5.22 (2.88)	5.29 (2.78)	4.87 (2.83)	0.320
Your local pharmacist	6.93 (2.45)	6.48 (2.74)	7.13 (2.32)	6.13 (2.66)	< 0.001	7.09 (2.63)	6.70 (2.95)	7.25 (2.53)	6.75 (2.58)	0.006
Your doctors	7.90 (2.36)	7.42 (2.78)	8.11 (2.19)	7.12 (2.60)	< 0.001	7.86 (2.53)	7.35 (2.81)	8.03 (2.43)	7.63 (2.50)	0.001
The governor of your state	5.58 (3.04)	5.32 (2.99)	5.72 (3.05)	4.94 (2.89)	0.001	5.27 (3.12)	5.13 (3.15)	5.32 (3.13)	5.22 (3.04)	0.710
Former President Donald Trump	3.98 (3.37)	2.19 (2.36)	4.41 (3.45)	3.68 (3.15)	< 0.001	4.28 (3.43)	3.03 (2.89)	4.73 (3.50)	3.48 (3.17)	<0.001
President Joseph Biden	5.81 (3.37)	6.70 (2.99)	5.66 (3.44)	5.45 (3.27)	< 0.001	4.83 (3.40)	5.50 (3.38)	4.58 (3.39)	5.26 (3.27)	<0.001
Dr Anthony Fauci	6.18 (3.31)	6.54 (3.29)	6.19 (3.29)	5.43 (3.36)	0.001	5.24 (3.45)	5.57 (3.40)	5.12 (3.51)	5.45 (3.15)	0.170
Pharmaceutical companies	5.55 (2.84)	5.01 (2.89)	5.74 (2.81)	4.92 (2.88)	< 0.001	4.83 (2.99)	4.91 (2.97)	4.79 (2.99)	4.94 (3.01)	0.780

Vaccine information source trust score range: 0 (Do not trust at all) to 10 (trust a great deal).

COVID-19 Vaccination in Children. Of the 1,174 participants at T2, 31.3% (n = 368) were parents/guardians: 23.3% (n = 274) of children aged 12-18 years, 19.0% (n = 223) of children aged 5-11-years, and 11.0% (n = 129) of children in both age groups. Parent/guardian respondents were 62.5% female, with a mean age of 40.26 ± 10.98 years, and reported being White 64.1% (n = 236), Black 25.5% (n = 94), and Other races 10.3% (n = 38).

Full vaccination of all children aged 12-18 years was not statistically different across race groups (White: 42.9% [n = 75], Black: 31.9% [n = 23], Other: 48.1% [n = 13]; P = 0.269). Higher percentages of participants reported none of their children had been vaccinated compared with some or all children (Black: 63.9% [n=46], White: 49.7% [n = 87], Other: 48.1% [n = 13]). Parents with children aged 12-18 years who were vaccinated (regardless of race) had higher WFPTS scores compared with parents with some or no vaccinated children (P = 0.001) (Table 4). The most common reasons for not vaccinating children aged 12-18 across all races were lack of trust in the vaccine and uncertainty about its longterm safety (Supplementary Table 5 ^{(229.2KB, pdf)} ).

TABLE 4.

Mean Wake Forest Physician Trust Scale Scores for Parent/Guardian Intent to Have Their Child/Children Vaccinated Across Racial Groups at Time Point 2 (November 2021)

	Total population (N = 1,174)			Black population (N = 233)			White population (N = 828)			Other population (N = 113)			P value
	N	Mean	(SD)	N	Mean	(SD)	N	Mean	(SD)	N	Mean	(SD)
Have your children aged 12-18 years been vaccinated? Response options: Yes, all of them; some of them; no, none of them.
Respondents answering question (have child/children aged 12-18 years), N	274			72			175			27
None or some	163	61.17	(20.78)	49	58.88	(21.32)	100	61.35	(20.60)	14	67.86	(20.07)	0.001
All	111	70.54	(16.97)	23	65.00	(20.62)	75	72.40	(16.32)	13	69.62	(11.63)
How likely are you to get your children aged 5-11 years vaccinated? Response options: Very likely; somewhat likely; unsure; somewhat unlikely, very unlikely.
Respondents answering question (have child/children 5-11 years), N	223			53			148			22
Somewhat likely to unsure	144	63.47	(20.63)	36	58.47	(22.00)	95	65.79	(20.59)	13	60.38	(14.50)	0.003
Very likely	79	73.42	(15.78)	17	74.71	(17.18)	53	73.49	(15.74)	9	70.56	(14.67)

Wake Forest Physician Trust Scale range: 0 (Lowest trust score) to 100 (highest trust score).

The reported likelihood of vaccinating children aged 5-11 years was not significantly different across races (P = 0.403); however, parents “very likely” to vaccinate their younger children (regardless of race) also reported higher WFPTS scores compared with parents “somewhat likely,” “somewhat to very unlikely,” or “unsure” to vaccinate (P = 0.003) (Table 4). Parents very likely to vaccinate their children aged 5-11 years also reported the highest WFPTS scores.

Of the total T2 population, regardless of having children, Black respondents relative to White and Other race respondents reported the least trust in the vaccine development process in general and in the governmental approval process to ensure the COVID-19 vaccine is safe for children, though not statistically different. Across racial groups, the greatest proportion reported they “mostly trust” the vaccine development process in general and the governmental approval process to ensure the COVID-19 vaccine is safe for children (Supplementary Table 6 ^{(229.2KB, pdf)} ).

Predictors of COVID-19 Vaccination. A multivariable logistic regression analysis explored the relationships between societal factors, physician trust, and vaccination status at T1 and T2 (Table 5). At T1, married status, a college education or higher, a higher WFPTS (physician trust) score, and a higher CCI score were associated with a higher likelihood of vaccination. Being Black relative to not being Black, having a median household income below $100,000 per year relative to income above $100,000 per year, and residing in the Northeast or Midwest relative to the West were associated with a decreased likelihood. Sex, age, smoking, and residing in the South relative to the West were not associated with the likelihood of vaccination. At T2, among a more highly vaccinated population, some predictors changed. Older age, married status, a college education or higher, and a higher WFPTS (physician trust) score were associated with a higher likelihood of vaccination, whereas being a smoker and having a median household income below $100,000 per year were associated with a decreased likelihood. Sex, race, geographic residence, and CCI score were not associated with the likelihood of vaccination at T2. Both models exhibited a C-statistic greater than 0.7, indicating good model fit (T1 C-statistic: 0.726; T2 C-statistic: 0.732).

TABLE 5.

Multivariable Logistic Regression Analysis to Identify Predictors of Vaccination Uptake in Adults at Time Points 1 (February/March 2021) and 2 (November 2021)

	February/March 2021				November 2021
	Odds ratio	95% CI		P value	Odds ratio	95% CI		P value
		Lower	Upper			Lower	Upper
Age	1.027	1.019	1.035	< 0.001	1.018	1.009	1.026	< 0.001
Female sex	1.121	0.875	1.435	0.370	1.225	0.917	1.636	0.170
Smoking status (yes)	0.943	0.717	1.241	0.680	0.640	0.477	0.860	0.003
Race (Black)	0.377	0.253	0.561	< 0.001	1.072	0.762	1.507	0.690
Northeast	0.569	0.393	0.822	0.003	1.551	0.970	2.478	0.070
Midwest	0.584	0.408	0.836	0.003	0.819	0.544	1.232	0.340
South	0.838	0.637	1.102	0.210	0.750	0.519	1.084	0.130
Marital status (married)	1.557	1.220	1.987	< 0.001	1.362	1.010	1.836	0.040
Education (college or higher)	1.686	1.322	2.150	< 0.001	2.350	1.698	3.252	< 0.001
WFPTS score	1.018	1.011	1.026	< 0.001	1.026	1.018	1.034	< 0.001
CCI score	1.255	1.131	1.394	< 0.001	0.972	0.870	1.087	0.620
Income (median < $100K)	0.678	0.636	0.722	< 0.001	0.845	0.793	0.899	< 0.001

CCI = Charlson Comorbidity Index; WFPTS = Wake Forest Physician Trust Scale.

Discussion

Our data provide an insight into the COVID-19 vaccination status of the US general population in 2021 and revealed significant racial disparities in adult vaccination status in early 2021 (T1), when around 20% of the survey population had been vaccinated. However, vaccination uptake increased by late 2021 (T2) to around 70%, and racial disparities were no longer significantly different, though vaccination rates among Black individuals remained lower at T2, as it did at T1. Our findings indicated that a high percentage of unvaccinated respondents had no plans to get vaccinated in early 2021 and that vaccine hesitancy was highest among Black individuals. However, current CDC data support the upward trajectory in COVID-19 vaccine uptake of our 2021 findings. April 2022 CDC data indicated that 75.8% of adults are now vaccinated (2 COVID-19 vaccine shots) and 48.8% have received their first booster dose, though some differences still exist among all racial and ethnic groups.⁹ Increased rates of vaccine uptake among individuals of ethnic and racial minorities were a positive occurrence given that they have increased risks of hospitalization and death from COVID-19 compared with White individuals.²⁷ Extant research shows that the late increase of vaccine uptake among ethnic and racial minorities was not unexpected given that Black and Hispanic individuals planned to delay receiving the vaccine.²⁸ These higher numbers were likely a reflection of increased awareness and access in ethnic and racial minority and poor communities, increased comfort with receiving the vaccine by the end of 2021, as well as the implementation of employer vaccine mandates.

Primary predictors of being vaccinated also changed as time passed. Although some sociodemographic factors, such as education, income, and age, along with physician trust, were predictive at both time points, race as a significant predictor of vaccine uptake disappeared at the later time period. Similarly, a recent RAND national study found that trust in the scientific community was strongly associated with receiving at least 1 dose of COVID-19 vaccine,²⁹ and our findings show that vaccine mistrust among the unvaccinated participants diminished over time. Still, not trusting the vaccine and being unsure about its long-term safety persisted as barriers to vaccine uptake across racial groups in late 2021. In addition, respondents’ doctors remained the most trusted source of COVID-19 information. These findings corroborate earlier work that reported significant vaccine hesitancy in US adults, with higher reluctance and mistrust among Black individuals and other ethnic and racial minority individuals, as well as lower socioeconomic populations.^30-33 Trust in the scientific community has been rated highest among Asian individuals compared with other races.²⁹ The disproportionate financial and emotional impact of the pandemic on poor and ethnic and racial minority communities likely exacerbated their initial unwillingness and inability to receive the vaccine.³⁴ Moreover, our findings show that trust level was lower in late 2021 compared with earlier in all race groups. Competing priorities such as financial and housing situations dictated by socioeconomic circumstances may have played a more significant role in driving vaccine uptake early on compared with trust, as was supported by differences in early and late predictors of vaccine uptake. Nevertheless, our later findings support the notion that any strategies to improve vaccination rates or reduce vaccine hesitancy from this point forward should involve increasing trust in the scientific community and vaccine, as well as utilizing physicians to accomplish it.

At the time of our November survey, vaccination status of children aged 12-18 years was not statistically different across race; however, parents/guardians who fully vaccinated all their children or were very likely to vaccinate their younger children (when eligible) had higher physician trust scores compared with their respective counterparts. Moreover, Black parents/guardians with unvaccinated children exhibited the least trust in doctors as well as the vaccine development process. These findings corroborate current evidence that shows the importance of parental trust for attaining a high vaccination rate in children. Parents’ concerns continue to center around vaccine side effects and long-term safety.^16,17,21,35

The success of vaccination programs is highly dependent on rates of acceptance and coverage· especially in communities of color and low socioeconomic status.³⁶ Factors such as misinformation about vaccine safety· speed of vaccine development· and the lack of trust in the scientific community and political leaders can all affect the public’s willingness to receive vaccination against COVID-19·^37,38 and our findings support the significance of these factors.

LIMITATIONS

This work has several important limitations. First, participants were limited to those with computer access who volunteered to take online surveys. The results are generalizable to a similar group of individuals; however, data collection was not a representative sample· and weighting was not used to make the results fully representative. Previous data indicate variability within the Other group on willingness to receive vaccines³³; however, small sample sizes did not allow for analyses of the races or ethnicities comprising the Other group. Our data provide a snapshot of the general population’s vaccination status at 2 time points in 2021. Vaccination rates increased with time· but it is unknown if other factors not measured are at work during the rapidly changing COVID-19 environment. Survey data were self-reported and subject to recall bias, although the accuracy of self-reported vaccination status has been previously suppported.^39,40 Participants were excluded if they could not complete the survey electronically which could bias the findings against adults who are older, of lower socioeconomic status· or not English literate.

Conclusions

Racial variation in vaccine hesitancy diminished in 2021· whereas physician trust remained a significant predictor of COVID-19 vaccine uptake. Physician trust is associated with vaccination status and intent· regardless of race. Lower WFPTS (physician trust) scores and trust in the vaccine development process were exhibited by unvaccinated respondents and associated with unvaccinated children· especially among the Black population. Respondents’ doctors· family members· and pharmacists are trusted sources of vaccine information· and targeting these influencers may reduce vaccination hesitancy.