Published on in Vol 9 (2023)

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/40587, first published .
COVID-19 Vaccination Preferences Among Non-Chinese Migrants in Hong Kong: Discrete Choice Experiment

COVID-19 Vaccination Preferences Among Non-Chinese Migrants in Hong Kong: Discrete Choice Experiment

COVID-19 Vaccination Preferences Among Non-Chinese Migrants in Hong Kong: Discrete Choice Experiment

Original Paper

1The Jockey Club School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

2Centre for Health Systems and Policy Research, The Jockey Club School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong

3Faculty of Law, The Chinese University of Hong Kong, Shatin, Hong Kong

*these authors contributed equally

Corresponding Author:

Dong Dong, PhD

Centre for Health Systems and Policy Research

The Jockey Club School of Public Health and Primary Care

The Chinese University of Hong Kong

4/F School of Public Health Building

Prince of Wales Hospital

Shatin

Hong Kong

Phone: 852 22528461

Email: dongdong@cuhk.edu.hk


Background: Studies have shown increasing COVID-19 vaccination hesitancy among migrant populations in certain settings compared to the general population. Hong Kong has a growing migrant population with diverse ethnic backgrounds. Apart from individual-level factors, little is known about the migrants’ preference related to COVID-19 vaccines.

Objective: This study aims to investigate which COVID-19 vaccine–related attributes combined with individual factors may lead to vaccine acceptance or refusal among the migrant population in Hong Kong.

Methods: An online discrete choice experiment (DCE) was conducted among adults, including Chinese people, non-Chinese Asian migrants (South, Southeast and Northeast Asians), and non-Asian migrants (Europeans, Americans, and Africans) in Hong Kong from February 26 to April 26, 2021. The participants were recruited using quota sampling and sent a link to a web survey. The vaccination attributes included in 8 choice sets in each of the 4 blocks were vaccine brand, safety and efficacy, vaccine uptake by people around, professionals’ recommendation, vaccination venue, and quarantine exemption for vaccinated travelers. A nested logistic model (NLM) and a latent-class logit (LCL) model were used for statistical analysis.

Results: A total of 208 (response rate 62.1%) migrant participants were included. Among the migrants, those with longer local residential years (n=31, 27.7%, for ≥10 years, n=7, 20.6%, for 7-9 years, n=2, 6.7%, for 4-6 years, and n=3, 9.7%, for ≤3 years; P=.03), lower education level (n=28, 28.3%, vs n=15, 13.9%, P=.01), and lower income (n=33, 25.2%, vs n=10, 13.2%, P=.04) were more likely to refuse COVID-19 vaccination irrespective of vaccination attributes. The BioNTech vaccine compared with Sinovac (adjusted odds ratio [AOR]=1.75, 95% CI 1.14-2.68), vaccine with 90% (AOR=1.44, 95% CI 1.09-1.91) and 70% efficacy (AOR=1.21, 95% CI 1.03-1.44) compared with 50% efficacy, vaccine with fewer serious adverse events (1/100,000 compared with 1/10,000; AOR=1.12, 95% CI 1.00-1.24), and quarantine exemption for cross-border travelers (AOR=1.14, 95% CI 1.01-1.30) were the vaccine attributes that could increase the likelihood of vaccination among migrants. For individual-level factors, full-time homemakers (AOR=0.44, 95% CI 0.29-0.66), those with chronic conditions (AOR=0.61, 95% CI 0.41-0.91) and more children, and those who frequently received vaccine-related information from the workplace (AOR=0.42, 95% CI 0.31-0.57) were found to be reluctant to accept the vaccine. Those with a higher income (AOR=1.79, 95% CI 1.26-2.52), those knowing anyone infected with COVID-19 (AOR=1.73, 95% CI 1.25-2.38), those having greater perceived susceptibility of COVID-19 infection (AOR=3.42, 95% CI 2.52-4.64), those who received the influenza vaccine (AOR=2.15, 95% CI 1.45-3.19), and those who frequently received information from social media (AOR=1.52, 95% CI 1.12-2.05) were more likely to accept the vaccine.

Conclusions: This study implies that migrants have COVID-19 vaccination preference heterogeneity and that more targeted and tailored approaches are needed to promote vaccine acceptance for different subgroups of the migrant population in Hong Kong. Vaccination promotion strategies are needed for low-education and low-income migrant groups, migrants with chronic diseases, the working migrant population, homemakers, and parents.

JMIR Public Health Surveill 2023;9:e40587

doi:10.2196/40587

Keywords



COVID-19 vaccination has been implemented worldwide as a crucial public health tool to reduce morbidity and mortality and eventually put an end to social distancing measures and the pandemic as a whole [1]. As of December 29, 2021, about 8.68 billion vaccine doses had been administered worldwide [2], first targeting priority groups and gradually extending vaccination to the general population.

For this campaign to be successful, public acceptance and the willingness to undergo vaccination are needed across the whole population [1,3]; yet research indicates that migrant populations in certain settings are particularly hesitant toward COVID-19 vaccines [4,5]. However, little is known about the migrants’ preference related to COVID-19 vaccines [6]; this undermines health equity, since as vulnerable populations, migrants are at a higher risk of COVID-19 infection, hospitalization, and mortality compared to the general population [7].

Previous studies have mainly focused on individual-level factors. They have reported that migrants’ socioeconomic conditions and gender have an impact on COVID-19 vaccine acceptance [3,8,9]. Ethnicity-related factors can influence vaccine hesitancy as well [4,9-11]. Moreover, the literature on migrant health suggests that migration history can differently affect access to primary medicine and vaccination uptake [12].

Finally, low vaccine coverage among migrants may be due to limited knowledge of the overall health care system [4]; in addition, during the pandemic, barriers within health care systems usually identified as possible factors affecting migrants’ lower access to preventive care and vaccinations have increased [5,9]. Trust in the health care system or in vaccines may affect vaccine uptake as well [4,9,11,13].

Hong Kong is an international city with a growing migrant and ethnic population, which reached 8% of the overall population in 2016, representing a sharp increase of about 70% compared to 2006 [14]. Despite its facade of an international and inclusive city, Hong Kong holds a hegemonic Chinese culture where exclusion is widespread [15]; this may affect migrant population access and attitudes toward vaccines as well.

Although research has focused on individual variables, little is known about how vaccine-related attributes may influence migrants’ attitudes toward COVID-19 vaccination. Similar to previous research conducted on the general population [16,17], this study aims to investigate COVID-19 vaccine preferences among the migrant population by identifying which individual factors combined with vaccine-related attributes may lead to vaccine acceptance or refusal. Although most of the literature focuses on the general population, this analysis will provide a more nuanced understanding of migrants’ overall attitudes toward vaccines and guide public health efforts to promote COVID-19 vaccination among more diverse groups.


Study Sample and Data Collection

An online discrete choice experiment (DCE) was conducted among all adults, including Chinese people and non-Chinese migrants in Hong Kong, China, from February 26 to April 26, 2021, the first 2 months after the commencement of the citywide COVID-19 vaccination program of 2 major vaccines. Chinese or non-Chinese residents of Hong Kong aged 18 years or above were eligible to participate in the survey, while those who had a history of being diagnosed with COVID-19 or had received any COVID-19 vaccine were excluded. The participants were recruited using quota sampling for people with different ethnicities according to Hong Kong census data [18], including Chinese, non-Chinese Asians (South Asians, Southeast Asians and Northeast Asians), and non-Asians (Europeans, Americans, and Africans). A web link to the survey was sent to potential participants through the network established in a previous survey, through local nongovernmental organizations that provide social services to socially disadvantaged individuals and migrants, and through persons familiar with the migrant communities. The distribution of ethnicities in the sample was monitored online during the survey to adjust the sampling strategies for the remaining participants, and ceiling limits (quotas) were set for the number of participants with different ethnicities based on the ethnical distribution in the population. The survey was originally designed in English and then forward- and back-translated to other languages and thus was also available in Bahasa Indonesia, Nepali, Urdu, Thai, and traditional Chinese.

Experimental Design

A DCE requires participants to make choices from a series of choice sets described by a number of attributes. The attributes describing the COVID-19 vaccination plans included vaccine brand, efficacy, safety, uptake of vaccine by people around, recommendations from professionals, venue for vaccination, and exemption of quarantine for vaccinated travelers (Table 1), which were generated from prior individual interviews with 45 Hong Kong residents with diverse demographical characteristics and health conditions [19]. In this qualitative study, the confidence in and concerns about the benefits and side effects of the vaccines, vaccine origins and brands, recommendations from health care professionals, social influences of family members and friends, expectations of ease of travel restrictions, and logistics arrangement of vaccination (eg, locations of receiving the vaccines) were frequently reported as factors influencing the participants’ willingness to accept the vaccines [19]. The levels of the attributes were determined based on the existing studies on vaccine effectiveness and safety [20-23] and the vaccination practice in Hong Kong [24]. For vaccine brands, all 3 (Sinovac, BioNtech, and AstraZeneca) were planned to be used for vaccination in Hong Kong at the time of the study, although only the former 2 of them were actually launched for the public eventually.

The full factorial design of the 7 attributes involved 648 combinations and 209,628 pairwise choice sets, so it was impossible to adopt the full factorial design in the DCE. The choice sets were designed using the D-optimality algorithm, and a total of 32 choice sets were generated. To further reduce the cognitive burden on participants, these choice sets were divided into 4 blocks, with 8 choice sets in each. The participants were randomly assigned to 1 of the 4 blocks: participants only provided their response to the choice sets in the block they were randomized to. In each choice set, participants were asked to choose 1 of the 3 choices (“vaccination plans”), 2 of which were choices to get a COVID-19 vaccine with different attributes, while the third choice was an opt-out option for accepting neither vaccination plan (“no vaccination”). An example of a choice set is shown in Table 2. The questionnaire was validated in the Chinese language. In total, 12 eligible adults were invited to a pilot survey for refinement of the translation in other languages.

Table 1. Attributes and levels for the DCEa.
AttributeLevels
Brand
  • Sinovac
  • BioNTech
  • AstraZeneca
Probability of COVID-19 infection (efficacy)
  • Reduce 50% infections
  • Reduce 70% infections
  • Reduce 90% infections
Probability of serious adverse event (safety)
  • 1/10,000
  • 1/100,000
Vaccine uptake by people around
  • Nobody
  • Friends/colleagues
  • Family members
Recommendations from professionals
  • Recommended by general physicians
  • Recommended by the government expert panel
Venue for vaccination
  • Community hall
  • Health care facilities
  • Housing estate/workplace
Quarantine arrangement for vaccinated travelers
  • At least 14-day compulsory quarantine
  • Exempted from the 14-day quarantine

aDCE: discrete choice experiment.

Table 2. Example of choice sets.
AttributeVaccination plan 1Vaccination plan 2Do not receive any vaccination
BrandSinovacBioNTechNone
Probability of COVID-19 infectionReduce 50% infectionsReduce 90% infectionsNo reduction
Probability of serious adverse event1/100,0001/10,000No serious adverse event
Vaccine uptake by people aroundFamily members received the vaccineNo known people received the vaccineN/Aa
Recommendations from professionalsGovernment expert advisory panelGeneral physiciansN/A
Venue for vaccinationHousing estate/workplaceCommunity hallN/A
Quarantine arrangement for vaccinated travelerAt least 14-day compulsory quarantineExempted from the 14-day quarantineAt least 14-day compulsory quarantine
Which vaccination plan would you choose?Plan 1: □Plan 2: □Neither plan: □

aN/A: not applicable.

Measurement

In addition to measuring the participants’ choice of COVID-19 vaccine via the DCE, the questionnaire also collected data on (1) experience, knowledge, and behaviors during the pandemic, such as whether the participant knew anyone infected with COVID-19, and the perceived susceptibility and severity of COVID-19 infection; (2) experience and perceptions related to vaccination, including previous uptake of influenza vaccination and information sources on COVID-19 vaccination; and (3) demographic and socioeconomic status of the participants, including their ethnicity, years of residence in Hong Kong, income level, education level, and employment status.

Ethical Considerations

The study was approved by the Survey and Behavioral Research Ethics Committee of the anonymous university (ref. no. SBRE-20-540). At the start of the survey, an electronic informed consent form was provided to the participants with details of the study purpose, data anonymity, and confidentiality. Those who agreed to join signed it electronically before moving to the questionnaire. Later, a supermarket coupon of HK $100 (US $12.74) was sent to each participant through mail. All the responses were anonymized and contained no personal information about any participant. The data were stored online and secured with passwords.

Statistical Analysis

To find out the preference of all participants for vaccination and its heterogeneity across ethnic groups, a latent-class logit (LCL) model was applied to the entire sample. LCL allocated participants in the survey sample with a similar preference for the vaccination attributes into the same latent class and provided estimates of the preference for each of the classes. The number of latent classes was determined based on the Akaike information criterion (AIC) and Bayesian information criterion (BIC) of the LCL model with the different number of classes. The difference in the preference between the Chinese and non-Chinese participants was explored. Subsequently, the analysis focused on the preference pattern within the non-Chinese migrants due to a substantial difference found between Chinese people and non-Chinese migrants, where the evidence for the latter group was limited.

For migrants, COVID-19 “vaccine refusal” irrespective of vaccine attributes was considered when a participant consistently chose “no vaccination” throughout all 8 choice sets. The vaccine refusal rate across different socialdemographic characteristics and the perception and experience during the COVID-19 pandemic were summarized using cross-tabulation and the chi-square test. To find out the influence of vaccination attributes, a nested logistic model (NLM) was adopted to simulate the choices over the vaccination plans in the DCE survey, which involved 2 decisions, namely (1) whether to accept a COVID-19 vaccine and (2) which COVID-19 vaccine to accept. NLM allowed estimation of the influences of both vaccination attributes and individual-level factors on the vaccine acceptance. The dependent variable was the binary choice (0=“not choose”, 1=“choose”) made for each of the alternatives in the DCE choice sets, and the independent variables were vaccination attributes and individual-level factors. The individual vaccine acceptance probability was estimated based on NLM outcomes [25,26]. In addition, sensitivity analysis was performed using the LCL model among migrant participants to find out whether there was any preference heterogeneity.


Sample Characteristics and COVID-19 Vaccine Refusal

In total, 2892 Chinese people and non-Chinese migrants were invited, of which 2392 (82.7%) were eligible. Among them, 434 (18.1%) refused to participate in the survey and 261 (10.9%) did not complete the questionnaire; hence, 2032 (84.9%) valid responses were received.

Figure 1 shows the participant flow of the migrant group. A total of 462 people received the invitation, and 335 (72.5%) were eligible to the survey. Among them, 109 (32.5%) did not agree to participate and 18 (5.4%) did not complete the questionnaire; therefore, 208 (response rate 62.1%) participants were recruited. The characteristics of the participants can be found in Table 3. Of the 208 participants, 143 (68.8%) were female, and 67 (32.2%) were aged 18-29 years, 103 (49.5%) were aged 30-44 years, and 38 (18.3%) were aged 45 years or more. Most of them (n=180, 86.5%) reported their ethnicity as Asian. In addition, 108 (51.9%) of them attained a bachelor’s degree or above. Almost half of them (n=100, 48.1%) had a full-time job, and 33 (15.9%) were full-time homemakers. Furthermore, 76 (36.5%) of them had more than HK$ 30,000 (US $3821.68) as their monthly household income, which was the approximate median monthly household income in Hong Kong in 2020. Only 19 (9.1%) had chronic conditions. For years of residence, 112 (53.8%) had lived in Hong Kong for 10 years or more, while 31 (14.9%) of them had lived here for 3 years or less.

Regarding COVID-19 vaccine refusal irrespective of vaccination attributes (Table 3), participants with longer local residential years, lower education level, and lower income were more likely to refuse COVID-19 vaccination. With regard to ethnicity, Asian participants were more likely to refuse vaccination than non-Asian participants, although this difference was marginal. Full-time homemakers/housewives, those who had chronic conditions, those who did not know anyone having COVID-19, and those who did not receive influenza vaccines were more likely to refuse vaccination irrespective of the attributes, although the differences were not statistically significant. On the contrary, those with greater perceived susceptibility of COVID-19 infection were less likely to refuse vaccination.

Figure 1. Participant selection flowchart.
Table 3. Sample characteristics and COVID-19 vaccine refusal irrespective of vaccine attributes.
CharacteristicsNo refusal, n (%)Refusal irrespective of attributes, n (%)Migrants (N=208), n (%)Total participant (N=2032), n (%)
Age (years), P=.26

18-2956 (83.6)11 (16.4)67 (32.2)832 (40.9)

30-4483 (80.6)20 (19.4)103 (49.5)706 (34.7)

45-5925 (69.4)11 (30.6)36 (17.3)361 (17.8)

≥601 (50.0)1 (50.0)2 (1.0)133 (6.6)
Sex, P=.84

Male51 (78.5)14 (21.5)65 (31.2)769 (37.8)

Female114 (79.7)29 (20.3)143 (68.8)1263 (62.2)
Ethnicitya, P=.06

Asian139 (77.2)41 (22.8)180 (86.5)2004 (98.6)

Non-Asian26 (92.9)2 (7.1)28 (13.5)28 (1.4)
Years of residence in Hong Kong, P=.03

≤3 years28 (90.3)3 (9.7)31 (14.9)N/Ab

4-6 years28 (93.3)2 (6.7)30 (14.4)N/A

7-9 years27 (79.4)7 (20.6)34 (16.3)N/A

≥10 years81 (72.3)31 (27.7)112 (53.8)N/A

MissingN/AN/A1 (0.4)N/A
Education, P=.01

Below bachelor’s degree71 (71.7)28 (28.3)99 (47.8)923 (45.5)

Bachelor’s degree or above93 (86.1)15 (13.9)108 (52.2)1108 (54.6)
Employment, P=.11

Full-time81 (81.0)19 (19.0)100 (48.1)1353 (68.6)

Part time28 (80.0)7 (20.0)35 (16.8)

Unemployed12 (80.0)3 (20.0)15 (7.2)83 (4.2)

Students/interns23 (92.0)2 (8.0)25 (12.0)271 (13.7)

Full-time homemakers/housewives21 (63.6)12 (36.4)33 (15.9)175 (8.9)

RetiredN/AN/AN/A91 (4.6)
Monthly household income (HK $/US $)c, P=.04

<30,000/<3821.6898 (74.8)33 (25.2)131 (63.3)953 (46.9)

≥30,000/≥3821.6866 (86.8)10 (13.2)76 (36.7)1078 (53.1)
Chronic condition, P=.07

No153 (81.0)36 (19.1)189 (90.9)1802 (88.7)

Yes12 (63.2)7 (36.8)19 (9.1)230 (11.3)
Know anyone diagnosed with COVID-19, P=.07

No115 (76.2)36 (23.8)151 (72.6)1676 (91.9)

Yes50 (87.7)7 (12.3)57 (27.4)356 (19.5)
Perceived “likely/very likely” to be infected, P<.001

No86 (69.9)37 (30.1)123 (59.1)1163 (63.8)

Yes79 (92.9)6 (7.1)85 (40.9)869 (47.6)
Perceived “slightly severe/very severe” if get infected COVID-19, P=.82

No80 (80.0)20 (20.0)100 (48.1)978 (53.6)

Yes85 (78.7)23 (21.3)108 (51.9)1054 (57.8)
Previous influenza vaccination, P=.44

No138 (78.4)38 (21.6)176 (84.6)1636 (89.7)

Yes27 (84.4)5 (15.6)32 (15.4)396 (21.7)
Often received vaccine information from social media, P=.60

No77 (77.8)22 (22.2)99 (47.6)956 (52.4)

Yes88 (80.7)21 (19.3)109 (52.4)1076 (59.0)
Often received vaccine information from family/friends, P=.06

No114 (83.2)23 (16.8)137 (65.9)1237 (67.8)

Yes51 (71.8)20 (28.2)71 (34.1)795 (43.6)
Often received vaccine information from the workplace, P=.33

No94 (77.1)28 (23.0)122 (58.7)1339 (73.4)

Yes71 (82.6)15 (17.4)86 (41.3)693 (38.0)
Often received vaccine information from the government\'s official source, P=.45

No97 (77.6)28 (22.4)125 (60.1)1292 (70.8)

Yes68 (81.9)15 (18.1)83 (39.9)740 (40.6)

Total165 (79.3)43 (20.7)208 (100.0)1824 (100.0)

aEthnicity groups included Asians (South Asians, Southeast Asians, and Northeast Asians) and non-Asians (Europeans, Americans, and Africans).

bN/A: not applicable.

cHK $1=US $0.13.

Influence of COVID-19 Vaccination Attributes on Vaccine Acceptance Among the Entire Sample

A model with 6 latent classes was applied to test the preference heterogeneity between Chinese people and non-Chinese migrants (Table 4). Using class 6 as a reference, the migrants were much more likely to be assigned to class 1 based on their preference, followed by class 2 and class 4. In class 1, vaccine brand was the most important attribute that affected their choices, where BioNtech was the most preferable brand, followed by vaccine efficacy, venue of vaccination, vaccine uptake by others, and quarantine arrangement for vaccinated travelers. In class 2, vaccine brand, efficacy, safety, and quarantine arrangement for vaccinated travelers shared a similar level of perceived importance based on the respondents’ choices, while vaccination at health care facilities and vaccination at the housing estate/workplace were preferable options as well. In class 4, higher vaccine efficacy was more important than the other attributes that lead to vaccine acceptance. The different preference pattern across classes 1, 2, and 4 implied that preference heterogeneity may be found within the migrant group, which is reported in the following sections. Compared to classes 3, 5, and 6, respondents in classes 1, 2, and 4 were more likely to be affected by the venue of vaccination. In class 3, the participants were efficacy oriented, preferred BioNtech and AstraZeneca, and were more likely to be influenced by general physicians with regard to vaccination decisions. In class 5, the participants preferred Sinovac, were more safety oriented, and presented a lower tendency to refuse vaccination than the other classes. Class 6 had the largest class share, where the participants shared similar preferences with participants in class 3 but were not influenced by recommendations from different health care professionals and vaccine uptake by their family members.

Table 4. Preference heterogeneity for COVID-19 vaccination plans in both Chinese people and non-Chinese migrants.
AttributeClass 1, coefficienta (95% CI)Class 2, coefficienta (95% CI)Class 3, coefficienta (95% CI)Class 4, coefficienta (95% CI)Class 5, coefficienta (95% CI)Class 6, coefficienta (95% CI)
Brand (reference=Sinovac)

BioNtech3.51b (2.97 to 4.06)0.49b (0.33 to 0.65)1.10b (0.87 to 1.32)1.43b (1.05 to 1.81)–3.59b (–4.12 to –3.07)0.57b (0.01 to 1.12)

AstraZeneca–0.29 (–0.73 to 0.16)0.69b (0.52 to 0.85)1.29b (1.06 to 1.53)1.41b (0.94 to 1.89)–4.04b (–4.58 to –3.50)0.59b (0.03 to 1.15)
Efficacy (reference=50%)

Reduce 70% infections1.15b (0.83 to 1.48)–0.03 (–0.16 to 0.10)0.93b (0.69 to 1.16)2.10b (1.73 to 2.48)–0.01 (–0.44 to 0.41)0.20 (–0.38 to 0.78)

Reduce 90% infections0.56b (0.05 to 1.07)0.74b (0.58 to 0.91)2.32b (2.06 to 2.59)3.43b (2.93 to 3.94)0.30 (–0.14 to 0.75)1.12b (0.58 to 1.65)
Serious adverse event (reference=1/10,000 people)

1/100,000 people–0.20 (–0.49 to 0.08)0.62b (0.51 to 0.72)0.64b (0.49 to 0.80)–0.33b (–0.55 to –0.11)0.33b (0.01 to 0.64)0.72b (0.26 to 1.18)
Vaccine uptake by others (reference=no known people take the vaccine)

Friends/colleagues received0.58b (0.03 to 1.13)–0.05 (–0.18 to 0.08)0.33b (0.14 to 0.52)1.17b (0.84 to 1.51)0.64b (0.24 to 1.04)0.73b (0.19 to 1.28)

Family members received0.63b (0.17 to 1.09)0.08 (–0.07 to 0.23)0.42b (0.21 to 0.63)0.97b (0.59 to 1.34)0.48b (0.03 to 0.94)0.41 (–0.21 to 1.03)
Recommendations from experts (reference=from general physicians)

From expert advisory panel of the government–0.13 (–0.58 to 0.31)0.04 (–0.06 to 0.14)–0.21b (–0.36 to –0.06)0.82b (0.53 to 1.11)0.10 (–0.21 to 0.41)–0.13 (–0.58 to 0.33)
Venue for vaccination (reference=community center)

Health care facilities–1.00b (–1.51 to –0.50)0.23b (0.09 to 0.36)–0.09 (–0.28 to 0.09)–0.63b (–0.96 to –0.29)–0.08 (–0.45 to 0.28)0.05 (–0.47 to 0.57)

Housing estate/workplace0.08 (–0.33 to 0.50)0.25b (0.12 to 0.39)–0.04 (–0.22 to 0.15)–0.40b (–0.69 to –0.10)0.24 (–0.26 to 0.74)–0.13 (–0.69 to 0.44)
Quarantine arrangement for vaccinated traveler (reference=compulsory quarantine required)

14-day compulsory quarantine can be exempted–0.54b (–0.98 to –0.10)0.67b (0.55 to 0.79)0.44b (0.28 to 0.61)0.19 (–0.03 to 0.42)–0.11 (–0.46 to 0.25)0.80b (0.35 to 1.25)
Opt out (reference=not opt out)

Opt out/no vaccination5.07b (2.35 to 7.80)1.70b (0.87 to 2.54)10.07b (8.72 to 11.42)9.95b (7.43 to 12.48)–8.24b (–10.27 to –6.21)10.67b (7.49 to 13.85)
Class size10.7% (N/Ac)12.2% (N/A)16.4% (N/A)21.5% (N/A)5.8% (N/A)33.4% (N/A)
Membership

Age: 30-44 years1.19b (0.74 to 1.64)0.31 (0.00 to 0.63)0.28 (–0.04 to 0.60)0.58b (0.19 to 0.97)2.10b (1.21 to 2.99)Reference

Age: 45-60 years1.57b (1.02 to 2.11)1.07b (0.68 to 1.46)–0.04 (–0.58 to 0.50)1.39b (0.93 to 1.86)3.24b (2.35 to 4.14)Reference

Age: >60 years1.80b (1.04 to 2.56)1.05b (0.43 to 1.66)–0.15 (–1.03 to 0.74)1.31b (0.60 to 2.02)3.78b (2.80 to 4.76)Reference

Bachelor’s degree or higher0.01 (–0.39 to 0.41)–0.08 (–0.38 to 0.22)0.77b (0.43 to 1.11)0.26 (–0.11 to 0.62)–0.58b (–1.08 to –0.07)Reference

Monthly income: >HK $30,000/US $3821.68d0.02 (–0.38 to 0.42)–0.02 (–0.32 to 0.28)–0.34b (–0.66 to –0.03)–0.03 (–0.39 to 0.33)–0.30 (–0.76 to 0.16)Reference

Migrants1.48b (0.99 to 1.97)0.86b (0.41 to 1.32)–0.65 (–1.37 to 0.07)0.86b (0.34 to 1.38)–1.55 (–3.32 to 0.23)Reference

aCoefficient of attributes on utility estimated by the latent-class logit (LCL) model. The values of coefficients are weightings of the attribute levels, showing the magnitude of contributions from the attribute levels to the overall utility of the COVID-19 vaccine option. The higher the value is, the greater the contribution the corresponding attribute level makes to the overall utility, and hence the more important and preferable the attribute is.

bP<.05.

cN/A: not applicable.

dHK $1=US $0.13.

Influence of COVID-19 Vaccination Attributes on Vaccine Acceptance Among Migrants

Given that substantial preference heterogeneity was found between Chinese people and non-Chinese migrants, the preference of the migrants was further modeled separately. Among the COVID-19 vaccination attributes, brand, efficacy, safety (probability of serious adverse event), and the quarantine arrangement of vaccinated people could affect vaccine acceptance (Tables 5 and 6). Independent from efficacy and safety characteristics, participants were more likely to accept the BioNtech vaccine compared with Sinovac. The vaccine efficacy also made a substantial difference. A vaccine with 90% and 70% efficacy would increase the likelihood of vaccination by 44% and 21%, respectively. Fewer serious adverse events (1/100,000 vs 1/10,000) could also improve the likelihood of vaccination. Apart from brand, efficacy, and safety, if the quarantine for cross-border travelers can be exempted for those who received the vaccine, the likelihood of vaccine acceptance could increase by around 14%.

Table 5. Influence of COVID-19 vaccination attributes on vaccine acceptance in the NLMa.
AttributeVaccine acceptance, AORb (95% CI)
Brand (reference=Sinovac)

BioNtech1.75c (1.14-2.68)

AstraZeneca1.11 (0.97-1.26)
Efficacy (reference=50%)

70%1.21c (1.03-1.44)

90%1.44c (1.09-1.91)
Serious adverse event (reference=1/10,000)

1/100,0001.12c (1.00-1.24)
Vaccine uptake by others (reference=no known people take the vaccine)

Friends/colleagues received1.11 (0.98-1.25)

Family members received1.12 (0.97-1.28)
Recommendations from professionals (reference=general physicians)

Government expert advisory panel1.08 (0.99-1.17)
Venue for vaccination (reference=community center)

Health care facilities1.00 (0.91-1.10)

Housing estate/workplace0.98 (0.90-1.07)
Quarantine arrangement for vaccinated traveler (reference=14-day quarantine)
14-day compulsory quarantine can be exempted1.14c (1.01-1.30)

aNLM: nested logistic model.

bAOR: adjusted odds ratio.

cP<.05.

Table 6. Influence of individual-level factors on vaccine acceptance in the NLMa.
AttributeVaccine acceptance, AORb (95% CI)
Age (years; reference=18-29 years)

30-442.42c (1.60-3.64)

≥452.20d (1.34-3.60)
Female (reference=male)1.17 (0.87-1.57)
Ethnicity (reference=Asian)

Non-Asian2.28 (0.49-10.64)
Years of local residence (reference=less than 3 years)

4-60.97 (0.51-1.83)

7-90.24c (0.13-0.43)

≥100.41c (0.25-0.67)
Education (reference=below bachelor’s degree)

Bachelor’s degree or above0.76 (0.53-1.09)
Employment (reference=full-time)

Part time1.12 (0.77-1.63)

Unemployed1.08 (0.64-1.81)

Students/interns1.41 (0.81-2.43)

Full-time homemakers0.44c (0.29-0.66)
Number of children (reference=0)

10.52d (0.33-0.82)

20.53d (0.33-0.85)

≥30.36c (0.22-0.61)
Monthly household income>HK $30,000 (US $3821.68)1.79d (1.26-2.52)
With any chronic condition0.61d (0.41-0.91)
Know anyone diagnosed with COVID-191.73d (1.25-2.38)
Perceived “likely/very likely” to be infected3.42c (2.52-4.64)
Perceived “slightly severe/very severe” if get infected COVID-190.90 (0.69-1.18)
Previous influenza vaccination2.15c (1.45-3.19)
Often received vaccine information from social media1.52d (1.12-2.05)
Often received vaccine information from the workplace0.42c (0.31-0.57)
Often received vaccine information from family/friends1.08 (0.80-1.46)
Often received vaccine information from the government\'s official source1.27 (0.95-1.70)

aNLM: nested logistic model.

bAOR: adjusted odds ratio.

cP<.001.

dP<.05.

For individual-level factors, participants who had been in Hong Kong for more than 7 years were less likely to accept the COVID-19 vaccine compared with those who had been in Hong Kong for less than 3 years. Full-time homemakers, those with chronic conditions and more children, and those who frequently received vaccine-related information from the workplace were found to be reluctant to accept the vaccine. On the contrary, those with a higher income, those knowing anyone infected with COVID-19, those having greater perceived susceptibility of COVID-19 infection, those who received the influenza vaccine, and those who frequently received information from social media were more likely to accept the vaccine.

Based on the model outcomes, the individual probability of the acceptance of vaccines with different characteristics can be estimated [25,26]. According to the estimation, an Asian migrant aged 18-29 years has a 73.6% chance to accept the BioNtech vaccine with 90% efficacy and a 1/100,000 serious adverse event probability when the vaccine is available in a community center, recommended by the government, and received by family members and when the quarantine exemption is in place. The acceptance probability was estimated to be 48.6% for the Sinovac vaccine with 50% efficacy. A non-Asian aged 18-29 years was estimated to have an 87.1% and a 69.8% chance to accept BioNtech and Sinovac vaccines, respectively. The acceptance probability changes with increasing age. For a migrant aged 30-44 years, the acceptance probabilities were estimated to be 88.9% (BioNtech) and 73.0% (Sinovac) among Asian migrants and 95.1% (BioNtech) and 86.9% (Sinovac) among non-Asian migrants. The estimated probabilities were similar for those aged above 45 years, which were 88.2% (BioNtech) and 71.8% (Sinovac) among Asian migrants and 94.8% (BioNtech) and 86.1% (Sinovac) among non-Asian migrants.

Sensitivity Analysis Account for Preference Heterogeneity

Substantial preference heterogeneity was found using a 4-class LCL model (Table 7). The participants had a 50.2% probability of belonging to class 1, where they attached greater importance to brand, efficacy, and exemption of quarantine for vaccinated travelers than the other attributes, as found in the NLM model for the entire study sample, and it was considered the reference group for class membership as it had the largest class size. Compared to class 1 (the reference group), participants in class 2 (16.1% probability) were more likely to live in Hong Kong for 7-9 years and they valued safety and the venue for vaccination, in addition to the 3 attributes mentioned in class 1. They were also likely to refuse vaccination (ie, choosing “no vaccination”). Class 2 and class 4 shared similar characteristics, they were likely to have longer local living experience compared to class 1, while class 4 was slightly more likely to comprise full-time homemakers. In class 4 (22.7% probability), the participants did not have clear preferences for these vaccination attributes.

Preferences in class 3 were quite different from those in classes 1 and 2. In class 3 (11.0% probability), participants preferred AstraZeneca over Sinovac and BioNtech and vaccine uptake by family and friends/colleagues as well as recommendations from experts from the government panel could also improve their COVID-19 vaccine acceptance, while exemption of quarantine did not have an impact on it. Participants in class 3 were more likely to be non-Asians and much less likely to be full-time homemakers.

Table 7. Sensitivity analysis for preference heterogeneity for COVID-19 vaccination plans.
AttributeClass 1, coefficienta (95% CI)Class 2, coefficienta (95% CI)Class 3, coefficienta (95% CI)Class 4, coefficienta (95% CI)
Preference for brand (reference=Sinovac)

BioNtech1.61b (1.35 to 1.88)2.13b (1.57 to 2.69)–0.87 (–1.81 to 0.06)–0.52 (–3.28 to 2.24)

AstraZeneca0.36c (0.08 to 0.65)–0.04 (–0.67 to 0.59)1.21c (0.08 to 2.33)0.94 (–1.69 to 3.57)
Efficacy (reference=50%)

Reduce 70% infections0.85b (0.57 to 1.12)0.53c (0.01 to 1.06)–1.52c (–2.78 to –0.27)0.12 (–3.75 to 4.00)

Reduce 90% infections1.15b (0.86 to 1.44)0.62c (0.07 to 1.18)0.63 (–0.13 to 1.39)2.67 (–0.26 to 5.60)
Serious adverse event (reference=1/10,000 people)

1/100,000 people0.06 (–0.14 to 0.25)0.68c (0.26 to 1.09)1.54c (0.48 to 2.61)–2.53 (–5.74 to 0.68)
Vaccine uptake by others (reference=no known people take the vaccine)

Friends/colleagues received0.13 (–0.15 to 0.41)0.45 (–0.07 to 0.98)1.11c (0.20 to 2.02)1.55 (–1.85 to 4.94)

Family members received0.22 (–0.09 to 0.53)0.08 (–0.48 to 0.63)1.63c (0.33 to 2.92)2.85 (–1.20 to 6.90)
Recommendations from experts (reference=general physicians)

From expert advisory panel of the government0.19 (–0.01 to 0.39)–0.39 (–0.83 to 0.04)1.13c (0.36 to 1.90)–0.37 (–2.29 to 1.55)
Venue for vaccination (reference=community center)

Health care facilities–0.16 (–0.46 to 0.13)0.64c (0.12 to 1.17)0.63 (–0.14 to 1.40)0.95 (–1.36 to 3.26)

Housing estate/workplace–0.03 (–0.30 to 0.25)0.09 (–0.45 to 0.62)0.38 (–0.37 to 1.12)–0.16 (–3.08 to 2.76)
Quarantine arrangement for vaccinated traveler (reference=compulsory quarantine required)

14-day compulsory quarantine can be exempted0.33c (0.09 to 0.58)0.73c (0.26 to 1.19)–0.27 (–1.07 to 0.54)0.58 (–1.97 to 3.13)
Opt out (reference=not opt out)

Opt out/no vaccination–5.96 (–274.19 to 262.27)7.00b (4.16 to 9.85)5.93c (0.67 to 11.19)13.43 (–4.11 to 30.97)
Class size50.2% (N/Ad)16.1% (N/A)11.0% (N/A)22.7% (N/A)
Membership

Age: 30-44 yearsReference–0.93 (–1.92 to 0.06)–1.23 (–2.46 to 0.00)–0.35 (–1.27 to 0.56)

Age: ≥45 yearsReference–0.73 (–2.04 to 0.59)–0.74 (–2.17 to 0.69)0.15 (–0.97 to 1.26)

Non-AsianReference–0.78 (–2.42 to 0.85)1.20c (0.05 to 2.36)–0.89 (–2.26 to 0.48)

Residential years: 4-6 yearsReference0.55 (–1.26 to 2.36)0.79 (–1.26 to 2.84)0.36 (–1.44 to 2.15)

Residential years: 7-9 yearsReference1.71c (0.08 to 3.35)0.21 (–2.11 to 2.54)2.03c (0.48 to 3.57)

Residential years: ≥10 yearsReference1.35 (–0.07 to 2.76)1.36 (–0.28 to 3.00)1.69c (0.33 to 3.05)

Full-time homemakerReference0.48 (–0.75 to 1.70)–80.06b (–102.11 to –58.01)0.96c (0.00 to 1.91)

aCoefficient of attributes on utility estimated by the latent-class logit (LCL) model. The values of coefficients are weightings of the attribute levels, showing the magnitude of contributions from the attribute levels to the overall utility of the COVID-19 vaccine option. The higher the value is, the greater the contribution the corresponding attribute level makes to the overall utility, and hence the more important and preferable the attribute is.

bP<.001.

cP<.05.

dN/A: not applicable.


Principal Findings

This study investigated both vaccine-related attributes and individual-level factors to fully understand migrants’ COVID-19 vaccine preferences in Hong Kong. It also provided an insight into the heterogeneity of the vaccine preference of the migrant population by dividing them into 4 classes.

Attributes

Our DCE study found that brand, efficacy, and quarantine arrangement for vaccinated people to be the most relevant vaccine-related attributes for vaccine acceptance among the migrant population in Hong Kong. The heterogeneity analysis complemented these findings, as half of the participants belonging to class 1 made their preferences based on these attributes. In addition, participants in class 3 preferred AstraZeneca over Sinovac and BioNtech. This indicated that migrants are inclined to use the brand as a proxy of unobserved characteristics in decision-making for vaccine acceptance. Previous research conducted on the US general population similarly showed brand and, specifically, the country where the vaccine was developed as predictors of COVID-19 vaccine acceptance [17,27]. With regard to efficacy, our study showed that the higher the efficacy, the higher the willingness to get vaccinated against COVID-19; this is in line with what emerged from studies in the general and ethnic populations in the United States [27] and the United Kingdom [17], where most of these studies have been conducted. Furthermore, specific migrant and ethnic groups may be persuaded if vaccines are proved to reduce the risk of being infected with COVID-19, as shown by research conducted in the United Kingdom [11]. The exemption of quarantine for vaccinated travelers was a predictor of COVID-19 vaccine acceptancy; this result is understandable within the Hong Kong context, where, upon returning, unvaccinated residents undergo self-paid 21-day compulsory quarantine in designated hotels compared to only 7 days for those vaccinated [28]. Other studies did not mention the quarantine arrangement specifically but more generally confirmed the vaccine would be accepted if it enabled social and family life to return to normal [11]. In addition to Sinovac and efficacy, our study found that a lower likelihood of serious adverse events is an attribute for vaccine acceptance, particularly for class 2 and class 3 participants. Similarly, the literature showed that concerns about vaccine safety and side effects are a predictor of vaccine hesitancy in both general and migrant populations in the United Kingdom [6,10,11].

The vaccine uptake by friends, colleagues, and family members; information from family and friends; and recommendations from the government and health care professionals were insignificant for the entire population sample but significantly affected only class 3 participants’ vaccine acceptance. These participants were mostly non-Asian. The existing literature suggests that such recommendations may be welcomed only in the presence of trust toward the government and the health care system [5,6]. In addition, American Hispanics in the United States, for instance, rely on their community network to make decisions about COVID-19 vaccines [29]. Similarly, migrants in the United Kingdom tend to look for information from peers, especially when the government and health care system are not trusted [5]. To promote COVID-19 vaccinations and improve their health care access and quality, efforts should be made to establish more trustworthy relationships between the local government and the health care system with all the migrant populations in Hong Kong.

In this study, based on vaccine attributes, we did not find significant preference differences between the migrant population and the general population. Instead, there was a substantial heterogeneity representing diversity in the preferences of the non-Chinese migrant population in Hong Kong. This implied that while designing COVID-19 vaccination promotion programs for migrants, this diversity should be considered and more targeted strategies should be used for different migrant subgroups. The venue of vaccination was the only significant factor for vaccine acceptance among migrants when compared to the general population, which is consistent with the literature that suggests an inconvenient location would influence hesitancy [30] and minimal travel or comfortable places would reduce hesitancy among undocumented migrants, asylum seekers, and refugees [6]. However, in the sensitivity analysis for preference heterogeneity among the migrant population, only a small proportion of participants in class 2 of our study had an attached preference with the venue of vaccination. These participants also shared characteristics with class 4 participants who did not have a clear preference for vaccine attributes. This means the choices made by these people were quite random, so they were probably overwhelmed by the choice tasks.

Individual-Level Factors

Our results also showed that irrespective of the vaccination attributes, longer residential years, lower education and income, and less perceived COVID-19 infection susceptibility were associated with increased COVID-19 vaccine refusal. These individual-level factors were also related to decreased COVID-19 vaccine acceptance among those whose preferences were linked to vaccine attributes. Participants with more than 7 years of migration history in Hong Kong were least likely to get vaccinated compared to those with a more recent migration history. This aligns with much of the literature on migrant health, where a longer history of migration and host country language proficiency are associated with acculturation in the host society and health care use similar to nationals [12]; vaccine hesitancy among Hong Kong local residents is in fact similarly widespread [31]. Therefore, effective policies and strategies that can boost vaccination acceptance among the broader public will aid in achieving high vaccination uptake among migrants as well. Previous research has also identified the educational level as a predictor of vaccine acceptance [9,11]. In line with most research, a higher income was also associated with higher vaccine acceptance, while a lower income was associated with vaccine resistance [10,27]. Similar to what emerged from the previous literature, knowing someone diagnosed with COVID-19 was a factor for vaccine acceptance [17] and so was high risk perception; conversely, low risk perception was associated with higher hesitancy [30].

In addition, our study also identified other individual-level factors that predict COVID-19 vaccine acceptance among individuals who preferred different vaccine programs. A previous influenza vaccination experience was relevant for COVID-19 vaccine acceptance; this is in line with studies conducted in migrant and general populations in other settings [8,10,17].

Our study also indicated that those participants who often received information from social media were more likely to accept COVID-19 vaccination. Regarding the role of this source of information, the literature is mixed, with varying degrees of results; some studies suggest social media is used to solve confusion from contradicting information [5], while other research suggests it is a source of misinformation among migrants [4,6]. Our study also found that those who received information from the workplace were less likely to get vaccinated. This might be because, on the one hand, these people are employed and healthy and therefore have low risk perception, leading to low vaccine acceptance. On the other hand, working people are exposed to different information sources, including local ones; discussions with their colleagues may also increase their vaccination hesitancy, as there is a decreasing trend in the willingness toward COVID-19 vaccination among general working people in Hong Kong [32]. The promotion of vaccine-related information through this source should be encouraged.

Older age was identified as a factor of vaccine acceptance; specifically, participants aged 30-44 years would be the most likely to get vaccinated, while the youngest cohort (18-29 years old) would be the least likely to join the vaccination campaign. A study in the United Kingdom confirmed the young group (16-24 years old) to be the most hesitant [11]; this hesitancy may be motivated by low risk perception compared to other age groups [30]. COVID-19 vaccination promotion campaigns should target young age group specifically among the migrant population in Hong Kong.

In line with the international literature, the Hong Kong Centre for Health Protection encourages individuals with chronic conditions to get vaccinated against COVID-19, since they are at a higher risk of morbidity and mortality due to COVID-19 infection [33]; however, our study highlighted that people with chronic diseases are less likely to accept vaccines. This result suggests the need to plan COVID-19 vaccination promotion campaigns targeting this high-risk group specifically.

Being a homemaker and having children were indicators of COVID-19 vaccine hesitancy; in the United Kingdom, those living with children are similarly less likely to accept the vaccine [10]. This might be because migrants have a lack of family and social support and, being parents, they are main carers of their children. During COVID-19, the care burden of parents, particularly homemakers and mothers, has increased due to lockdown, school closure, and home schooling. They have busy routines and no time for making appointments and going for vaccination. Moreover, there might be concerns about who will take care of their children in case they have vaccine-related side effects. So, along with providing information about vaccine safety, an outreach approach can be considered for this particular migrant subgroup. Qualitative research is needed to further address the reasons behind this attitude.

In this study, gender was not a predictor of COVID-19 vaccine hesitancy; this is in contrast with previous research in other settings that identified women as being more hesitant compared to men both in the general and in the migrant population [8,10,11,17].

Limitations

Our study had some limitations that should be clarified when interpreting the results. First, the preference of respondents for COVID-19 vaccination is subject to change with time and could be different as the vaccination campaign progresses. It may also have been affected by reported adverse events following immunization for COVID-19, which were not captured in this survey. Second, although we incorporated “Asian”/“non-Asian” as a variable for ethnicity in the analysis (Tables 5 and 6), there were no significant difference in vaccine acceptance between these 2 groups, and due to the limited sample size of ethnicity subgroups and statistical power, we could not further subdivide the ethnicity in our analysis. Instead, the LCL model was used to identify the preference heterogeneity (Table 7), where non-Asian migrant respondents tended to attach more importance to the vaccination of family and friends and preferred AstraZeneca to the other 2 vaccine brands compared to Asian migrants. However, the ethnicity of the respondents could not be further divided in the analysis due to the limited subgroup sample size. Third, although the survey was piloted and refined prior to the formal investigation, there were around 20% respondents (class 4 in the LCL model) who may have been overwhelmed by the cognitive burden of the DCE: their COVID-19 vaccination choice appears to be random. By using the LCL model, we were able to separate them from other groups of respondents and avoid affecting the results.

Conclusion

In summary, this study provided insight into how different COVID-19 vaccine attributes, preferences, and individual-level characteristics of the migrant population could influence their COVID-19 vaccine acceptance. Although vaccine brand, efficacy, safety, and quarantine exemption were the most preferred attributes for the majority of the migrants, we reiterate that there is COVID-19 vaccination preference heterogeneity among migrants and that more targeted and tailored approaches are needed to promote vaccine acceptance for different subgroups of the migrant population in Hong Kong. The increased willingness toward COVID-19 vaccination in the general population will reduce vaccine hesitancy among migrants with long years of living experience in Hong Kong. Vaccination promotional programs for low-education and low-income migrant groups should be designed in different languages. The dissemination of COVID-19 vaccine information, particularly about efficacy and safety, through social media platforms can be continued, and information dissemination at the workplace for the working migrant population should also be considered. Detailed and tailored educational programs should be provided to people with chronic diseases to promote COVID-19 vaccination in this high-risk group. An outreach approach for people with children, particularly homemakers and mothers, can be used. Further follow-up studies, including qualitative interviews, can be conducted with these participants to explore in-depth the reasons behind their COVID-19 vaccination preferences and refusal.

Acknowledgments

SA, KW, EN, FFY, and DD managed conceptualization; SA, KW, EN, and DD, methodology; SA, EN, FFY, LZ, HCEF, YZ, and ZF, data collection; SA and KW, data analysis; DD and EL-YW, data curation; SA, KW, EN, and FFY, writing—original draft preparation; KW, visualization; DD, writing—review and editing; DD, EL-YW, and E-KY, supervision; and EN, LZ, and AW-LC, project administration. All authors have read and agreed to the published version of the manuscript.

Data Availability

All the data used in this study are available upon request from the corresponding author.

Conflicts of Interest

None declared.

  1. Williams L, Gallant A, Rasmussen S, Brown Nicholls LA, Cogan N, Deakin K, et al. Towards intervention development to increase the uptake of COVID-19 vaccination among those at high risk: outlining evidence-based and theoretically informed future intervention content. Br J Health Psychol 2020 Nov;25(4):1039-1054 [FREE Full text] [CrossRef] [Medline]
  2. WHO coronavirus (COVID-19) dashboard with vaccination data. World Health Organization. 2022.   URL: https://covid19.who.int/ [accessed 2022-03-22]
  3. Bell S, Clarke R, Mounier-Jack S, Walker J, Paterson P. Parents’ and guardians’ views on the acceptability of a future COVID-19 vaccine: a multi-methods study in England. Vaccine 2020 Nov;38(49):7789-7798 [FREE Full text] [CrossRef]
  4. Crawshaw AF, Deal A, Rustage K, Forster AS, Campos-Matos I, Vandrevala T, et al. What must be done to tackle vaccine hesitancy and barriers to COVID-19 vaccination in migrants? J Travel Med 2021 Jun 01;28(4):2021 [FREE Full text] [CrossRef] [Medline]
  5. Knights F, Carter J, Deal A, Crawshaw AF, Hayward SE, Jones L, et al. Impact of COVID-19 on migrants’ access to primary care and implications for vaccine roll-out: a national qualitative study. Br J Gen Pract 2021;71(709):e583-e595 [FREE Full text] [CrossRef]
  6. Deal A, Hayward S, Huda M, Knights F, Crawshaw AF, Carter J, ESCMID Study Group for Infections in Travellers and Migrants (ESGITM). Strategies and action points to ensure equitable uptake of COVID-19 vaccinations: a national qualitative interview study to explore the views of undocumented migrants, asylum seekers, and refugees. J Migr Health 2021 May 27;4:100050 [FREE Full text] [CrossRef] [Medline]
  7. COVID-19 immunization in refugees and migrants: principles and key considerations: interim guidance, 31 August 2021. World Health Organization. 2021 Aug 31.   URL: https:/​/www.​who.int/​publications/​i/​item/​covid-19-immunization-in-​refugees-and-migrants-principles-and-key-considerations-interim-guidance-31-august-2021 [accessed 2023-03-09]
  8. Alabdulla M, Reagu SM, Al-Khal A, Elzain M, Jones R. COVID-19 vaccine hesitancy and attitudes in Qatar: a national cross-sectional survey of a migrant-majority population. Influenza Other Respir Viruses 2021 May;15(3):361-370 [FREE Full text] [CrossRef] [Medline]
  9. Malik AA, McFadden SM, Elharake J, Omer SB. Determinants of COVID-19 vaccine acceptance in the US. EClinicalMedicine 2020 Sep;26:100495 [FREE Full text] [CrossRef] [Medline]
  10. Paul E, Steptoe A, Fancourt D. Attitudes towards vaccines and intention to vaccinate against COVID-19: implications for public health communications. Lancet Reg Health Eur 2021 Feb;1:100012 [FREE Full text] [CrossRef] [Medline]
  11. Robertson E, Reeve KS, Niedzwiedz CL, Moore J, Blake M, Green M, et al. Predictors of COVID-19 vaccine hesitancy in the UK household longitudinal study. Brain Behav Immun 2021 May;94:41-50 [FREE Full text] [CrossRef] [Medline]
  12. Rosano A. Access to Primary Care and Preventative Health Services of Migrants. Cham: Springer; 2018.
  13. Ferdinand KC, Nedunchezhian S, Reddy TK. The COVID-19 and influenza "Twindemic": barriers to influenza vaccination and potential acceptance of SARS-CoV2 vaccination in African Americans. J Natl Med Assoc 2020 Dec;112(6):681-687. [CrossRef] [Medline]
  14. Thematic report: ethnic minorities. Census and Statistics Department. 2017 Dec.   URL: https://www.statistics.gov.hk/pub/B11201002016XXXXB0100.pdf [accessed 2023-03-09]
  15. Crabtree S, Wong H. 'Ah Cha'! The racial discrimination of Pakistani minority communities in Hong Kong: an analysis of multiple, intersecting oppressions. Br J Soc Work 2012 Apr 02;43(5):945-963 [FREE Full text] [CrossRef]
  16. Dong D, Xu RH, Wong ELY, Hung MC, Feng D, Feng Z, et al. Public preference for COVID-19 vaccines in China: a discrete choice experiment. Health Expect 2020 Dec;23(6):1543-1578 [FREE Full text] [CrossRef] [Medline]
  17. Kreps S, Prasad S, Brownstein JS, Hswen Y, Garibaldi BT, Zhang B, et al. Factors associated with US adults' likelihood of accepting COVID-19 vaccination. JAMA Netw Open 2020 Oct 01;3(10):e2025594 [FREE Full text] [CrossRef] [Medline]
  18. Thematic report: ethnic minorities. Census and Statistics Department. 2017 Dec.   URL: https://www.statistics.gov.hk/pub/B11201002016XXXXB0100.pdf [accessed 2023-03-09]
  19. Wong CH, Zhong CC, Chung VC, Nilsen P, Wong EL, Yeoh EK. Barriers and facilitators to receiving the COVID-19 vaccination and development of theoretically-informed implementation strategies for the public: qualitative study in Hong Kong. Vaccines (Basel) 2022 May 12;10(5):1 [FREE Full text] [CrossRef] [Medline]
  20. 20 VM, Clemens S, Madhi S, Weckx L, Folegatti P, Aley P. Safetyefficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa,the UK. Lancet 2021;397:99-111 [FREE Full text] [CrossRef]
  21. Polack FP, Thomas S, Kitchin N, Absalon J, Gurtman A, Lockhart S, et al. Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N Engl J Med 2020;383:2603-2615 [FREE Full text] [CrossRef]
  22. Zhang Y, Zeng G, Pan H, Li C, Hu Y, Chu K, et al. Safety, tolerability, and immunogenicity of an inactivated SARS-CoV-2 vaccine in healthy adults aged 18–59 years: a randomised, double-blind, placebo-controlled, phase 1/2 clinical trial. Lancet Infect Dis 2021 Feb;21(2):181-192 [FREE Full text] [CrossRef]
  23. Wu Z, Hu Y, Xu M, Chen Z, Yang W, Jiang Z, et al. Safety, tolerability, and immunogenicity of an inactivated SARS-CoV-2 vaccine (CoronaVac) in healthy adults aged 60 years and older: a randomised, double-blind, placebo-controlled, phase 1/2 clinical trial. Lancet Infect Dis 2021 Jun;21(6):803-812 [FREE Full text] [CrossRef]
  24. COVID-19 vaccination programme. Government of the Hong Kong Special Administrative Region.   URL: https://www.covidvaccine.gov.hk/en/ [accessed 2023-02-13]
  25. Craig BM. United States COVID-19 vaccination preferences (CVP): 2020 hindsight. Patient 2021 May;14(3):309-318 [FREE Full text] [CrossRef] [Medline]
  26. Wang K, Wong E, Cheung A, Chung V, Wong C, Dong D, et al. Impact of information framing and vaccination characteristics on parental COVID-19 vaccine acceptance for children: a discrete choice experiment. Eur J Pediatr 2022 Nov;181(11):3839-3849 [FREE Full text] [CrossRef] [Medline]
  27. Pogue K, Jensen JL, Stancil CK, Ferguson DG, Hughes SJ, Mello EJ, et al. Influences on attitudes regarding potential COVID-19 vaccination in the United States. Vaccines (Basel) 2020 Oct 03;8(4):582 [FREE Full text] [CrossRef] [Medline]
  28. Inbound travel. Government of the Hong Kong Special Administrative Region.   URL: https://www.coronavirus.gov.hk/eng/inbound-travel.html [accessed 2023-03-09]
  29. Opel DJ, Lo B, Peek ME. Addressing mistrust about COVID-19 vaccines among patients of color. Ann Intern Med 2021 May;174(5):698-700 [FREE Full text] [CrossRef]
  30. Razai MS, Chaudhry U, Doerholt K, Bauld L, Majeed A. Covid-19 vaccination hesitancy. BMJ 2021 May 20;373:n1138 [FREE Full text] [CrossRef] [Medline]
  31. Chan PKS, Wong M, Wong ELY. Vaccine hesitancy and COVID-19 vaccination in Hong Kong. Hong Kong Med J 2021 Apr;27(2):90-91 [FREE Full text] [CrossRef] [Medline]
  32. Wang K, Wong E, Ho K, Cheung A, Yau P, Dong D, et al. Change of willingness to accept covid-19 vaccine and reasons of vaccine hesitancy of working people at different waves of local epidemic in Hong Kong, China: repeated cross-sectional surveys. Vaccines (Basel) 2021 Jan 18;9(1):1 [FREE Full text] [CrossRef] [Medline]
  33. Interim guidance notes on common medical diseases and COVID-19 vaccination in primary care settings. Center for Health Protection. 2022 Sep 2.   URL: https://www.covidvaccine.gov.hk/pdf/Guidance_Notes.pdf [accessed 2023-03-09]


AOR: adjusted odds ratio
DCE: discrete choice experiment
LCL: latent-class logit
NLM: nested logistic model


Edited by A Mavragani, T Sanchez; submitted 30.06.22; peer-reviewed by O Rahaman, DMA Awal, Z Han; comments to author 06.01.23; revised version received 27.01.23; accepted 23.02.23; published 27.03.23

Copyright

©Saba Asim, Kailu Wang, Elena Nichini, Faustina Fu Yip, Liling Zhu, Hin Chung Eddy Fung, Yan Zeng, Zhilan Fang, Annie Wai-Ling Cheung, Eliza Lai-yi Wong, Dong Dong, Eng-Kiong Yeoh. Originally published in JMIR Public Health and Surveillance (https://publichealth.jmir.org), 27.03.2023.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on https://publichealth.jmir.org, as well as this copyright and license information must be included.