Geographic, Occupational, and Sociodemographic Variations in Uptake of COVID-19 Booster Doses Among Fully Vaccinated US Adults, December 1, 2021, to January 10, 2022

IMPORTANCE COVID-19 booster vaccine can strengthen waning immunity and widen the range of immunity against new variants. OBJECTIVE To describe geographic, occupational, and sociodemographic variations in uptake of COVID-19 booster doses among fully vaccinated US adults. DESIGN, SETTING, AND PARTICIPANTS This cross-sectional survey study used data from the Household Pulse Survey conducted from December 1, 2021, to January 10, 2022. Household Pulse Survey is an online, probability-based survey conducted by the US Census Bureau and is designed to yield estimates nationally, by state, and across selected metropolitan areas. MAIN OUTCOMES AND MEASURES Receipt of a booster dose was defined as taking 2 or more doses of COVID-19 vaccines with the first one being the Johnson and Johnson (Janssen) vaccine, or taking 3 or more doses of any of the other COVID-19 vaccines. Weighted prevalence estimates (percentages) were computed overall and among subgroups. Adjusted prevalence ratios (APRs) were calculated in a multivariable Poisson regression model to explore correlates of receiving a booster dose among those fully vaccinated.


Introduction
COVID-19 vaccines are a crucial preventive measure against the pandemic. 13][4] More than 553 million doses of COVID-19 vaccines have been administered in the US, covering an estimated 65.2% of the US population. 5rly on, there was a recognition that certain special populations, such as older adults or immunocompromised patients, needed additional COVID-19 vaccine doses after completion of an initial COVID-19 vaccination series because of their lower immune response. 6,7][10][11] Immunologic and epidemiologic studies performed on potency of COVID-19 vaccines showed reduction in antibody levels activated by the vaccines and increased incidence of breakthrough cases among fully vaccinated individuals. 1,12In addition, new variants emerged with the potential to evade vaccine immunity. 11,13Increased transmission of the Delta variant beginning in the spring of 2021 led to an increase in numbers of breakthrough COVID-19 infections. 1 According to US Centers for Disease Control and Prevention (CDC) recommendations, all individuals aged 12 years or older should receive a COVID-19 mRNA vaccine booster dose 5 months after completion of an initial vaccination series. 9The administration of booster vaccines strengthens waning immunity and widens the range of immunity against new variants detected. 14though studies 11,[14][15][16] have been performed to evaluate the need for and attitudes toward booster doses, to the best of our knowledge, no representative study has been conducted assessing the prevalence of booster dose uptake and factors associated with uptake.Understanding geographic, occupational, and sociodemographic differences in receipt of booster doses among the fully vaccinated population can help inform targeted efforts to increase coverage.The aim of this survey study was to explore prevalence and correlates of self-reported receipt of a booster vaccine dose among the fully vaccinated population.To explore place-based differences in booster vaccine uptake, prevalence estimates were computed and compared at the level of the major metropolitan statistical areas (MSAs), all 50 US states, and Washington, DC, as well as the US overall.

Data Source
Data were obtained from the Household Pulse Survey (HPS), an ongoing, biweekly, national and state-specific, COVID-19-related, cross-sectional survey of US adults aged 18 years or older. 17The type of internet access.For this study, we analyzed pooled data from 2 survey cycles spanning the period December 1, 2021, to January 10, 2022.This was a unique period that marked the shift from Delta to Omicron as the dominant variant of COVID-19 diagnoses in the US.Between December 13, 2021, and January 10, 2022, the share of SARS-CoV-2 sequences that were the Omicron variant increased dramatically from 3.74% to 93.07%. 5Ethical review and informed consent were not sought because the secondary, publicly available data set was deidentified and lacked private information, in accordance with 45 CFR §46.This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for observational studies and complies with ethical standards set by the American Association for Public Opinion Research (AAPOR) reporting guideline. 18,19

Receipt of a Full and Booster COVID-19 Vaccine Dose and COVID-19 Diagnosis
The survey asked participants whether they had ever "received a COVID-19 vaccine" (yes or no); "how many dose(s) of a COVID-19 vaccine" they had received (1, 2, 3, or 4 or more vaccinations) and the brand of first vaccination ("Pfizer-BioNtech", "Moderna", "Johnson and Johnson (Janssen)", "One of the brands that requires two initial shots, but not sure which brand", "None of these brands", or "Don't know").These variables were used to create indicators for receipt of a full dose (ie, completing the initial vaccination schedule) and a booster dose.Participants were said to have received a full COVID-19 vaccine dose (ie, be fully vaccinated) if they reported taking at least 1 dose of the Johnson and Johnson (Janssen) COVID-19 vaccine or at least 2 doses of any other COVID-19 vaccines.Receipt of a booster dose was defined as taking at least 2 doses of COVID-19 vaccines with the first one being the Johnson and Johnson (Janssen) vaccine, which is only a 1-dose regimen, or taking at least 3 doses of any of the other vaccines that require an initial 2-dose regimen.Responses of "Don't know" were set to missing to reduce misclassification bias.Self-reported diagnosis of COVID-19 was defined as a response of yes to the question, "Has a doctor or other healthcare provider ever told you that you have COVID-19?"

Sociodemographic Factors
We were interested in how receipt of a booster dose varied by occupational setting, especially among those working in jobs with frequent contact with individuals of unknown COVID-19 status.The survey asked, "In the last 7 days, which best describes the primary location/setting where you worked or volunteered outside your home?"A variety of settings was assessed (eg, hospital, K-12 school, or correctional facility) and only 1 response could be selected.Participants could also select if they did not work or volunteer "outside [their] home."Other sociodemographic factors asked about in the survey included marital status, number of children in household, US Census region, race and ethnicity, gender, age, education, and type of housing (eg, multiunit dwelling).Race and ethnicity were assessed in this study to determine whether disparities existed in booster dose coverage.

Statistical Analysis
Data (percentages) were weighted to yield representative estimates.The weights were provided in the online data set to account for the complex survey sampling.The final HPS weights are designed to produce biweekly estimates for the total persons aged 18 years or older living within housing units.These weights were created by adjusting the household-level sampling base weights by various factors to account for nonresponse, adults per household, and coverage.
The sampling base weights for each incoming sample in each of the sample areas are calculated as the total eligible housing units in the sampling frame divided by the number of eligible housing units selected for interviews each week.Therefore, the base weights for all sampled housing units sum to the total number of housing units for which contact information is known.Prevalence estimates were computed nationally, by state, and by the major MSAs.We expressed COVID-19 booster dose coverage both as prevalence in the overall population (

Descriptive Statistics
For our analyses, the pooled sample size was 135 821 participants.2).In the New York, New York, MSA, the original epicenter of the pandemic, 91.7% of all adults reported being fully vaccinated.Among those who had ever been vaccinated (partially or completely), the reported brand of the first vaccine dose was as follows: Pfizer-BioNtech (52.1%),Moderna (39.2%), Johnson and Johnson (Janssen) (7.0%), one of the brands that requires 2 initial shots, but not sure which brand (0.6%), none of these brands (ie, other, 0.3%), and missing brand information (0.2%).Differences in first administered dose by work setting are shown in the Figure.

Geographic Variations in Booster Coverage
Of all adults who were fully vaccinated, 48.5% in the US overall had been boosted (Table 1).
Expressed as a percentage of the total US adult population, this was 40.2%.By state, booster coverage among the fully vaccinated ranged from 39.1% in Mississippi (or 29.4% of the total adult population), to 66.5% in Vermont (or 60.3% of the total adult population).By MSA, booster coverage among the fully vaccinated ranged from 40.4% in the Phoenix-Mesa-Chandler, Arizona, MSA to 58.7% in the San Francisco-Oakland-Berkeley, California, MSA (Table 2).In the New York, New York, MSA, 46.0% of all fully vaccinated adults had received a booster dose, similar to the 47.0% reported in New York State.

Sociodemographic Variations in Booster Coverage
Sociodemographic characteristics associated with booster coverage among those fully vaccinated are presented in   One of the brands that requires 2 initial shots, but not sure which brand Occupational setting was defined as the primary location or setting where respondents worked or volunteered outside their home.Service), 22 whereas our estimate was derived only from the 50 US states and Washington, DC.
Moreover, our sample covers only the noninstitutionalized civilian population, whereas the CDC data captures all eligible individuals.In our study, boosted individuals were more likely than nonboosted individuals to be male, Asian, more educated, earning higher income, older, live in the Northeast, and work in hospitals or deathcare facilities.We found wide variations in booster coverage even among workers in the same industry.For example, within health care facilities nationwide, booster coverage was almost 25 percentage points higher among those working in hospitals compared with those working in a pharmacy, even with similar rates of initial vaccination completion within these settings.
This wide variability in uptake of booster doses may reflect greater volition in individuals' decision to receive the vaccine, in contrast to the initial vaccination series for which mandates existed in certain settings. 23Mass vaccination campaigns that incentivize people to receive and publicize their booster vaccinations may cue similar action among individuals in similar social networks according to research showing that anecdotal evidence is more influential than statistical evidence in situations where emotional engagement is high (eg, personal or health issues). 24rriers to uptake of booster doses are likely to be different from those against the initial doses.
Viewed through the lens of the health belief model components (ie, health motivation, perceived susceptibility, perceived severity, perceived benefits, perceived harms, and cues to action), [25][26][27] it might be surmised that hesitancy to initial vaccination was largely associated with perceived harm, including fear of unknown adverse effects of COVID-19 vaccines, and/or that some vaccines use the relatively new mRNA platform. 28,29On the other hand, the perception of having low susceptibility to COVID-19, or not having a severe illness in the instance of a breakthrough infection, may be factors that may explain relatively low uptake of booster doses.This is suggested by our finding that fully vaccinated individuals who were also COVID-19 survivors had 30% less likelihood of getting a booster dose compared with those with no history of SARS-CoV-2 infection, a finding that might be linked with perceived additional protection from infection-induced immunity.Uninsured individuals were significantly less likely to receive a booster dose compared with those with insurance.Also, those reporting experiencing very difficult financial situation had half the booster coverage of those reporting no difficulty financially.Booster coverage also decreased with decreasing education and was double among those with a doctoral, professional degree, or master's degree than those with less than a high school education.Taken together, these findings suggest socioeconomic barriers may limit uptake of the booster doses.2][33] Providing mobile vaccination clinics within low-income neighborhoods on a walk-in basis and launching multilingual awareness campaigns may benefit public health.

Limitations
This study has limitations.First, data were self-reported and may be subject to social and cognitive biases, which may lead to misreporting.Second, the cross-sectional design of the study means that only associations can be inferred.An important implication of the cross-sectional design is that we are unable to tease apart age, period, and cohort effects.For example, although our results suggested that individuals who ever received a diagnosis of COVID-19 were among the least likely to receive a booster dose, we are unable to assess the timing of diagnosis vs vaccination (eg, whether the diagnosis occurred between the completed vaccination series and the boosting).Thus, we cannot make inferences on whether people got boosted in response to having been infected or got infected despite having been boosted.Third, the results of this study may not be generalizable to individuals outside the sampling frame, including persons in the military, in prisons, or other institutionalized settings.Despite adjustment for differential nonresponse bias, the web-based immunocompromised should receive a third primary dose. 9Because we had no indicator to identify individuals who were immunocompromised, we could not discriminate between a third dose that was a primary dose (among those who were immunocompromised) or a booster dose (among those who were immunocompetent).The magnitude of this bias is, however, likely small as only about 2.7% of the US population is immunosuppressed. 34

Conclusions
Of all adults who were fully vaccinated, the percentage who had been boosted was 48.5% in the US overall, with variations seen by geographic, demographic, and occupational characteristics.Boosted individuals are more likely to be male, Asian, more educated, earning higher income, older, live in the Northeast, and work in hospitals.Intensified efforts to increase coverage of booster doses may benefit public health by averting excess premature morbidity and mortality from COVID-19.

Figure . First
Figure.First Administered COVID-19 Vaccine Brand Among Those Who Ever Took the Vaccine, by Occupational Setting, US, December 1, 2021, to January 10, 2022

Table 3
for the US overall.Nationally, the proportion of boosted adults was highest among non-Hispanic Asian individuals (54.1%), those aged 65 years or older (71.4%); those with a doctoral, professional, or master's degree (68.1%); those who were married with no children in the household (61.2%); those with an annual household income of $200 000 or more (69.3%);those enrolled in Medicare (70.9%); and those working in hospitals (60.5%) or in deathcare facilities such as funeral homes (60.5%).Conversely, only one-third or less of individuals in certain subgroups reported being boosted, including those who ever received a diagnosis of COVID-19; those enrolled

Table 1 .
National and State-Specific Prevalence Estimates of Full Vaccination Status and the Proportion Who Had Received a Booster Dose in the Overall Population and Among Those Fully Vaccinated, US, December 1, 2021, to January 10, 2022 (continued)

Table 2 .
Prevalence Estimates of Full Vaccination Status and the Proportion Who Had Received a Booster Dose in the Overall Population and Among Those Fully Vaccinated, by Selected Metropolitan Statistical Areas, US, December 1, 2021, to January 10, 2022 JAMA Network Open.2022;5(8):e2227680. doi:10.1001/jamanetworkopen.2022.27680(Reprinted) August 19, 2022 6/15 Downloaded From: https://jamanetwork.com/ on 09/24/2023 Discussion This survey study's key findings were that less than half (48.5%) of those who had been fully vaccinated nationwide had received a booster dose, with marked variations seen across geographic locations, occupation types, and other sociodemographic characteristics.Our estimates are similar to national surveillance data from the CDC; of the 189.78 million US adults aged 18 years or older who were fully vaccinated by mid-January 2022, a total of 77.73 million had received a booster dose, yielding a booster coverage rate of 41% among the fully vaccinated. 22The difference between the CDC estimate vs ours may reflect the fact that the CDC vaccination surveillance data includes records from territories (Guam, American Samoa, Republic of Palau, Federated States of Micronesia, Northern Mariana Islands, Marshall Islands, and Virgin Islands) and federal entities (eg, Indian Health

Table 3 .
Prevalence Estimates of Full Vaccination Status and the Proportion Who Had Received a Booster Dose in the Overall Population and Among Those Fully Vaccinated, by Selected Sociodemographic Characteristics, US, December 1, 2021, to January 10, 2022

Table 3 .
Prevalence Estimates of Full Vaccination Status and the Proportion Who Had Received a Booster Dose in the Overall Population and Among Those Fully Vaccinated, by Selected Sociodemographic Characteristics, US, December 1, 2021, to January 10, 2022 (continued) 30

Table 3 .
Prevalence Estimates of Full Vaccination Status and the Proportion Who Had Received a Booster Dose in the Overall Population and Among Those Fully Vaccinated, by Selected Sociodemographic Characteristics, US, December 1, 2021, to January 10, 2022 (continued)

Table 4 .
APRs for Factors Associated With Receipt of a COVID-19 Booster Dose Among Those Fully Vaccinated, US, December 1, 2021, toJanuary 10, 2022 (continued)Variations in Uptake of COVID-19 Booster Doses Among Fully Vaccinated US Adults survey may have resulted in some selection bias to the exclusion of individuals of low socioeconomic status.Fourth, small sample sizes within some subgroups yielded imprecise estimates.Fifth, among the subpopulation of immunocompromised individuals, we may have overestimated the proportion of boosted individuals.According to CDC recommendations, individuals who are a In the survey, other referred to "Any other race alone, or race in combination."JAMA Network Open | Public Health JAMA Network Open.2022;5(8):e2227680.doi:10.1001/jamanetworkopen.2022.27680(Reprinted) August 19, 2022 12/15 Downloaded From: https://jamanetwork.com/ on 09/24/2023