Table 2.

Characteristics of included studies.

Reference	Country	Study Design	Sample/Settings	Intervention	Measurement Scale/Instrument	Results	Quality Assessment
Abdul-Mutakabbir et al. (2021) [35]	United States	Not stated	General public—Mass vaccination clinic (n = 24,868); Mobile vaccination clinic (n = 1542); San Bernardino County (n = 2,180,085)	A three-tiered approach intervention: engagement with faith leaders in the academic community; a culturally inclusive and representative group of healthcare professionals from Loma Linda University to deliver the COVID-19 informational webinars; the completion of remote vaccination clinics.	Vaccination counts	Over the course of the intervention (February 1–30 April 2021), 24,808 persons received their vaccination at the routine mass vaccination clinic. 3.7% of these individuals were Black. At the remote clinics, 1542 individuals received a dose of the vaccine, with 44% of these individuals identifying as Black.	7
Albarracin et al. (2021) [36]	United States	Multicomponent study: 1 survey & 3 experiments (3-cell, within-subjects experimental design; 3 × 2 within-subjects factorial design; between-subjects design)	Total sample, randomly sampled US population including White, Black, & Hispanic American respondents (n = 1621); Survey (n = 299); Experiment 1 (n = 359); Experiment 2 (n = 357); Experiment 3 (n = 606). Randomly sampled US population	Survey: participants were asked two questions; Experiment 1: participants were randomly rotated through 3 different conditions; Experiment 2: participants were randomly rotated through 3 different conditions; Experiment 3: participants were randomized to either condition 1 or condition 3.	Yes/No scale, Likert scales	Survey: for Black participants, there was an increased likelihood of vaccination with the requirement questions than with the control question (p = 0.001); 80% of Black respondents reported the intent to vaccinate within the required condition; 56% of Black respondents reported the intent to vaccinate within the control condition. Experiments 2, 3, 4: data not disaggregated	9
Andrasik et al. (2021) [37]	United States	Community-based participatory research	Scientists and community leaders from four BIPOC communities (n = 40–60). Clinical trial sample size: variable throughout the course of study.	Meaningful involvement of the community; Stakeholder engagement and trust-building; Faith initiative: faith-based advisory; Communications & community influence	Pre-screening survey, enrollment counts	Across all 4 clinical trials, 47% of the enrolled participants were BIPOC, 15% of whom identified as Black or African American. The enrollment rates of various BIPOC communities aligned with their proportions within the broader US population. BIPOC enrollment was slow initially, but steadily grew and eventually overtook White enrollment. COVID-19 Prevention Network (CoVPN) sites that actively engaged with BIPOC communities had more success in recruiting BIPOC participants compared to non-CoVPN sites.	7
Feifer et al. (2021) [38]	United States	Not stated	Black and Hispanic healthcare workers/employees (n = 27,000)	Facts and informational sessions provided by experts; social media educational campaign; resources on the COVID-19 vaccine in Spanish and English; provision of culturally sensitive information through small group discussions and one-on-one conversations	Not stated	At the first time point, rates of vaccination were lowest among Black employees (45.5%) and highest among Asian employees (74.5%). The greatest increase in vaccination rates were among Alaskan Native employees (8.2%), Hispanic employees (6.1%), and Black employees (5.4%). There was a statistically significant increase in the likelihood of Black employees receiving the COVID-19 vaccination compared to White employees (p = 0.004).	6
Fox A. & Choi, Y. (2021) [39]	United States	Survey experiment	New York residents (n = 1353 (443 non-Hispanic Blacks, 429 non-Hispanic Whites, 481 Hispanics).	A messaging campaign that acknowledged historical racism through a newspaper prime that prioritizes minorities	Questionnaire	Participants’ intention to get vaccinated against COVID-19 was not influenced by the messaging campaign.	5
Hirshberg et al. (2021) [40]	United States	Not stated	High-risk obstetrical patients (n = 87)	Counseling on COVID-19 vaccination for pregnant and breastfeeding patients; onsite vaccine availability	Standardized COVID-19 vaccination discussion tool, state database	Patients seen before onsite vaccination was available: 1/32 (3%) patients received the vaccination offsite post-counselling. Patients who were seen after onsite vaccination: 2/55 (3%) received the vaccination onsite; 4/55 (7%) received the vaccination offsite. The availability of onsite vaccination was not significantly associated with an increase in vaccination (p = 0.22). Among the 55 patients who received counseling during the availability of onsite vaccination, 25 were counseled during the pilot program but did not receive a vaccination onsite or offsite.	7
Reddy et al. (2021) [41]	South Africa	Not stated	Healthcare workers (n = 7400)	Creation of staff database for appointment management; Vaccinators completed an online training program for healthcare workers; Vaccines were distributed by pharmacy teams and occupational health nurses. Reconciliation of doses was undertaken after every 24 doses dispensed. Staff arriving to be vaccinated had to display identification of booking confirmation, screened for symptoms, provided consent, then were observed for 15 min in a separate area after vaccination.	Dose counts	There was an upward trend in the number of vaccine doses administered in the week after the roll-out was initiated. Number of vaccines administered each subsequent day of the vaccination campaign (from 17 February 2021 to 26 February 2021, excluding 20 &21 February) = 32, 240, 460, 788, 823, 1018, 1138, 1160.	5
Reddy et al. (2021) [42]	South Africa	Not stated	South Africa’s population (specific sample size not stated)	COVID-19 vaccination program simulations over 360 days under various scenarios	Incremental cost-effectiveness ratio (ICER) = difference in healthcare costs divided by difference in years-of-life saved (YLS)	There were varying economic and clinical benefits to various vaccination strategies in the two scenarios—lack of vaccination program led to greatest amount of infections, deaths, costs; a 40% vaccination rate led to the lowest healthcare costs on both scenarios and decrease in deaths in both scenarios; a 67% vaccination rate also decreased deaths, however healthcare costs increased; an 80% vaccine supply and 80% vaccine acceptance decreased deaths and increased healthcare costs in both scenarios. Highest pace of vaccination at 300,000 daily led to optimal clinical outcomes and lowest costs when compared with lower paces in both scenarios. In the scenarios where the effective reproductive number = 1.4, 67% vaccine supply was less efficient (in terms of incremental cost-effectiveness ratio: the difference between healthcare costs divided by the difference between years-of-life saved, compared with other approaches to supply and pace) than the 80% vaccine supply. In the scenarios a two-wave epidemic, 67% vaccine supply was more efficient than the 80% vaccine supply; 20% vaccine supply was less efficient that higher levels of vaccine supply, yet decreased deaths by 72–76% and decreased healthcare costs by 15–32% when compared with no vaccination.	4
Robertson et al. (2021) [28]	United States	Randomized survey experiment (between-subjects design)	American adults (n = 1000)	Participants were assigned to four conditions for financial incentives ($1000, $1500, $2000, or no incentive).	Online survey form	Large financial incentives were found to be counterproductive in Black populations. 53% of Black Americans in the group receiving no incentives reported that they would most likely take the vaccine. The $1500 incentive resulted in an increase to 68% in the proportion of Black participants who reported that they would most likely take the vaccine. At the $2000 incentive, only 39% of Black respondents reported willingness to take the vaccine (a value 13.6% below the acceptance rate in the control group receiving no financial incentive).	7
Serper et al. (2021) [43]	United States	Not stated	Solid organ transplant recipients (n = 103)	Outreach & motivational interviewing by transplant center staff. Follow-up calls were scheduled to evaluate if participants proceeded to receive the vaccine.	Vaccination counts	People who identified as Black were more likely to schedule a vaccination appointment during the call with transplant center staff or to schedule a vaccine appointment on their own time compared to all other races (odd ratio = 1.6, 95%CI 1.02–2. 6; p = 0.042) after they received information from staff.	7
Ugwuoke et al. (2021) [44]	Nigeria	Quasi-experimental design	Victims of conflict in Nigeria (n = 470)	Artistic illustration communication intervention delivered concurrently with counseling on the acceptance of COVID-19 vaccination after it was available. This included experts from various disciplines: fine and applied arts, guidance and counseling, mass communication. It took place over the course of 10 days, with each session lasting 1 h. The control group did not receive this intervention. There was collaboration between the research team and the officials at the camp to ensure that the two groups did not interact over the course of the study.	Questionnaire	Participants who engaged with visual communication and counseling had greater average reports of self-efficacy than those who did not on the post-test questionnaire (3.3 vs. 1.4); they also had greater average reports of task efficacy on the post-test questionnaire (3.6 vs. 1.4). There were greater positive reports in terms of intent to vaccine in those who had exposure to visual arts on COVID-19 vaccination on the post-test questionnaire (3.8 vs. 1.3). Participants who had exposure to visual messages about COVID-19 vaccination reported higher levels of intention to vaccinate compared to the control group (self-efficacy 1.4 vs. 3.3, task efficacy 1.6 vs. 3.6, vaccination intention 1.1 vs. 3.8).	8
Wagner et al. (2021) [45]	United States	Randomized survey experiment	Detroit residents (n = 1117; 76.5% Black)	Hypothetical vaccine profile experiment	Questionnaire	More participants (77.1%) presented with a higher (95%) hypothetical effective vaccine had more acceptance rate compared to participants presented with vaccine that has lower (50%) effectiveness. Vaccine safety and some vaccine profiles had no significant impact on participants’ vaccine acceptance.	9
Williams-Gunpot (2021) [46]	United States	Cross-sectional survey	African American adults (n = 188)	E-health educational intervention that consisted of a pre and post knowledge test. The educational intervention (Our COVID-19 Knowledge Test) is a true-false tool programmed to have all true answers. The test consisted of 44 questions.	‘Our COVID-19 Knowledge’ Test (OCKT-44); COVID-19 Knowledge Scale; COVID-19 Prevention Self-Efficacy Scale; Intention to Vaccinate for COVID-19 (IVC-1) tool; Diffusion of Innovation of Our COVID-19 Knowledge Test (DOI-OCKT-1) Tool; COVID-19 Knowledge and Self-Efficacy for Risk Reduction Behaviors	Knowledge of COVID-19 was increased in the post-test (mean score 4.57 vs. 4.85, p = 0.000); knowledge of COVID-19 self-efficacy related to risk reduction (including vaccination) was also higher in the post-test (mean score 5.17 vs. 5.33, p = 0.000). 127 participants reported being vaccinated or having the intent to vaccinate. For the open-ended question, participants had feelings of hope about vaccination, noting that while there is still a chance of contracting COVID-19 with the vaccine, only a small percentage of vaccinated individuals get COVID-19.	8
					(C-K-SE-FRRB-4) tool
Yemer et al. (2021) [47]	Ethiopia	Not stated	Leaders disseminating messages; sample size not stated	Various communication strategies and messages were analyzed. Messages such as health information about COVID-19 vaccination, vaccination campaigns, advocacy, and vaccine hesitancy. SMS messages were sent, religious leaders were involved to increase the spread of positive COVID-19 vaccine messages, workshops were developed to increase awareness of COVID-19 vaccines.	Not stated	Messages (TV, radio, newspaper, social media) about COVID-19 vaccine hesitancy reported on side effects (4/4), accuracy/low quality (2/4), safety testing (3/4), doubts about vaccines (4/4), and others (1/4). The use of various communication strategies contributed to greater vaccine advocacy and vaccine campaigns contributed to increased vaccine acceptance.	3