key: cord-0849619-nbmwmdmj
authors: Kuhn, R.; Henwood, B.; Lawton, A.; Kleva, M.; Murali, K.; Gelberg, L.; King, C.
title: COVID-19 vaccine access and attitudes among people experiencing homelessness from pilot mobile phone survey in Los Angeles, CA
date: 2021-03-26
journal: nan
DOI: 10.1101/2021.03.23.21254146
sha: 3bfa9e77f9d5b597bcb9b9ea2133de3f58139e67
doc_id: 849619
cord_uid: nbmwmdmj

Background: People experiencing homelessness (PEH) are at high risk for COVID-19 complications and fatality, and have been prioritized for vaccination in many areas. Yet little is known about vaccine acceptance in this population. The objective of this study was to determine the level of vaccine hesitancy among PEH in Los Angeles, CA and to understand the covariates of hesitancy in relation to COVID-19 risk, threat perception, self-protection and information sources. Methods and findings: A novel mobile survey platform was deployed to recruit PEH from a federally qualified health center (FQHC) in Los Angeles to participate in a monthly rapid response study of COVID-19 attitudes, behaviors, and risks. Of 90 PEH surveyed, 43 (48%) expressed some level of vaccine hesitancy based either on actual vaccine offers (17/90 = 19%) or a hypothetical offer (73/90 = 81%). In bivariate analysis, those with high COVID-19 threat perception were less likely to be vaccine hesitant (OR=0.34, P=.03), while those who frequently practiced COVID-19 protective behaviors were more likely to be vaccine hesitant (OR=2.21, P=.08). In a multivariate model, those with high threat perception (OR=0.25, P=.02) were less likely to be hesitant, while those engaging in COVID-19 protective behaviors were more hesitant (OR=3.63, P=.02). Those who trusted official sources were less hesitant (OR=0.37, P=.08) while those who trusted friends and family for COVID-19 information (OR=2.70, P=.07) were more likely to be hesitant. Conclusions: Findings suggest that targeted educational and social influence interventions are needed to address high levels of vaccine hesitancy among PEH.

People experiencing homelessness (PEH), who have high rates of comorbid conditions 56 more typical of individuals 15-20 years older than their chronological age [1] [2] [3] 

This study was designed as a pilot for a larger platform to address the challenge of 76 gathering ongoing, longitudinal data from PEH through monthly online surveys. A university-77 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Once enrolled in the study, surveys were delivered through a HIPAA-compliant, cloud-94 based data collection platform that was designed to suit the capabilities of the study population, 95 with extensive consultation with a lived experience advisory group and testing with unhoused 96 clients. Informed consent was conducted via the survey questionnaire, requiring affirmative 97 consent before proceeding with the survey and providing complete informed consent 98 documentation at the start of each survey. A 5-minute baseline demographic and risk factor 99 survey was conducted December 2020 through January 2021. Monthly surveys lasted 15 100 minutes on average and included questions on COVID-19 risk perception, protective behaviors 101 and information sources along with physical and mental well-being. The third monthly survey 102 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Participants received financial incentives of $5 for the baseline and for each monthly survey. 104

The study protocols were approved by the 1st author's university IRB. 105 Dependent variables: Vaccine uptake was measured with a two-part question that first asked 106 whether a respondent had been offered a vaccine, followed by a hesitancy question based on 107 actual or hypothetical behavior. For those who had been offered a vaccine, individuals who did 108 not accept the vaccine were coded as vaccine hesitant. Among those who had not been offered 109 the vaccine, respondents were asked if they would take the vaccine if they were offered it, with 110 possible responses of "yes," "no" or "prefer not to answer." Those who responded "no" or "prefer 111 not to answer" were coded as vaccine hesitant. classified as responding "agree/strongly agree" to at least 3/4 questions. COVID-19 self-120 protective behavior was measured using a four-item index of how frequently the respondent 121 wore a mask, washed their hands, stayed 6 feet from others, and avoided touching their face. 122

Anxiety/depression was measured using the Patient Health Questionnaire-4 (PHQ-4), with 123 moderate-severe psychological distress classified using the documented scoring system.

Statistical analysis: After describing the univariate distribution for all dependent and 125 independent variables, we conduct bivariate analysis of vaccine hesitancy in terms of all 126 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 26, 2021. ; https://doi.org/10.1101/2021.03.23.21254146 doi: medRxiv preprint independent variables using two-tailed chi-square tests of differences in proportions and two-127 tailed t-tests of differences in means. We then estimated a multivariate model including all 128 factors shown to be significant in bivariate analysis. All statistical analyses were performed in 129 Stata 16. Due to the relatively small sample size, we report significance at both the 5% and 10% 130 levels. 131

132

The mean age of the sample was 48.7 and 59% of respondents were female (Table 1) . 133

The sample was predominantly White (49%), followed by Hispanic/Latino (18%), other (18%), 134

and Black/African American (9%). Most respondents were unsheltered (44%). More than half 135 (52%) of respondents were coded as having moderate/severe psychological distress according 136 to the PHQ-4 screening. Thirty three percent of respondents perceived COVID-19 as a high 137 threat, and 42% reported high COVID-19 protective behavior. More than half reported trust in 138 some official source (62%) or mass media (56%), while 42% reporting trusting personal 139 information sources such as friends, family or social media. 140 141 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 1 shows that, of the 90 respondents in the sample, 17 (19%) have been offered 152 the vaccine, 10 of whom accepted. Among the 73 not offered the vaccine, 37 (51%) said they 153 would take it if offered, 23 said they would not (32%), and 13 declined to answer (17%). Given 154 these results, 43 (48%) expressed vaccine hesitancy, as defined above. Among those who 155 rejected an offer of the vaccine or stated that they would not get the vaccine if offered (n=30), 156 the most common reasons cited for vaccine hesitancy or refusal were fear of side effects (37%), 157 wanting to have more information (30%), and rejection of all vaccines (27%) (Figure 2) . 158 159 Fig 1. COVID-19 vaccine hesitancy by prior vaccine access. Respondents who were offered 160 a vaccine (n=17) were asked whether or not they received the vaccine; those who received the 161 vaccine (n=10) were classified as not vaccine hesitant and those who did not receive the 162 vaccine were classified as vaccine hesitant (n=7). Respondents who had not been offered the 163 vaccine (n=73) were asked if they would take the vaccine. Those who said they would take the 164 vaccine (n=37) were classified as not hesitant and those who said they wouldn't (n=23) or 165 declined to answer (n=13) were classified as hesitant. 166 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. considerably higher than most online or phone-based polls, we know that those who answered 227 the survey were more likely to be female (59% vs. 35%) and less likely to be African-American 228 (9% vs. 24%) than that clinic's homeless patient base as a whole. Given the lack of differences 229 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. in hesitancy across any demographic groups and the small sample size, we did not report 230 weighted results. Finally, we note that these interviews were conducted prior to PEH receiving 231 universal vaccine eligibility on March 15, 2021, and that some hesitancy may more accurately 232 reflect indifference or frustration at the difficulty of obtaining the vaccine. 233

In spite of these limitations, our findings point to challenges in widespread vaccine 234 scaleup that are not so different than those faced in the general population. It is important to 235 know that those people who need the vaccine most -those who fear COVID-19 but are less 236 likely to protect themselves through social distancing measures -are those most highly willing to 237 be vaccinated. But achieving widespread vaccine acceptance may be far more challenging 238 among individuals who are more proactive with protective behaviors but who may be more 239 skeptical of the COVID-19 vaccine and who may have low trust in official information sources. 240

Preliminary results from a small survey of PEH in Los Angeles reveal a high rate of 242 vaccine hesitancy in this population, with higher levels of hesitancy observed among those with 243 low threat perception, those engaging in self-protective behaviors, and those with higher trust in 244 personal sources of information versus official sources. Our data suggest the need for targeted 245 educational and social influence interventions to increase vaccine uptake among PEH, who are 246 at greater risk of suffering from severe COVID-19 than the general population. Additional data 247 collected on a larger, more representative sample is necessary to determine differences in 248 vaccine attitudes across demographic variables like race. 249

Author contributions: Dr. Kuhn had full access to all of the data in the study and takes 250 responsibility for the integrity of the data and the accuracy of the data analysis. Concept and 251 design: Kuhn, Henwood, Gelberg, King. Acquisition of data: Kuhn, Henwood, Lawton, Murali, 252 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Matthew Ware; and Carrie Kowalski. We would also like to thank the staff at Akido Labs who 265 played an instrumental role in designing the survey and rapidly adding new questions, 266 specifically Aishwarya Badanidiyoor. Finally, we are tremendously grateful to the support 267 provided by our lived expertise group, a panel of individuals with past or current homelessness 268 experience, for informing design of our study and research platform. 269 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

Geriatric Conditions in a 271 Population-Based Sample of Older Homeless Adults

Hospitalized younger: a comparison 274 of a national sample of homeless and housed inpatient veterans. J Health Care Poor 275 Underserved

Estimated Emergency and 277 Observational/Quarantine Capacity Need for the US Homeless Population Related to 278 COVID-19 Exposure by County; Projected Hospitalizations, Intensive Care Units and 279 Mortality

Elevated mortality among 281 people experiencing homelessness with COVID-19. Epidemiology

Epidemiology of COVID-19 among people experiencing 284 homelessness: early evidence from Boston

Assessment of 286 SARS-CoV-2 Infection Prevalence in Homeless Shelters -Four

Prevalence of SARS-CoV-2 infection in 290 residents of a large homeless shelter in Boston

Understanding 292 Drivers of COVID-19 Vaccine Hesitancy Among Blacks

Readiness to Get COVID-19 Vaccine Steadies at 65%

Correlates and 297 disparities of intention to vaccinate against COVID-19

Vaccine Hesitancy: Clarifying 300 a Theoretical Framework for an Ambiguous Notion

Of natural bodies and antibodies: Parents' vaccine refusal and the dichotomies 303 of natural and artificial

The health of homeless people in high-income countries: 305 Descriptive epidemiology, health consequences, and clinical and policy recommendations

Barriers, 308 beliefs, and practices regarding hygiene and vaccination among the homeless during a 309 hepatitis A outbreak in Detroit

Scale: Development and Initial Validation

An ultra-brief screening scale for anxiety and 314 depression: the PHQ-4

Piloting the Use of Artificial Intelligence to Enhance HIV Prevention Interventions for Youth 317 Experiencing Homelessness

No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity