Acceptance of and preference for COVID-19 vaccination in healthcare workers: a comparative analysis and discrete choice experiment

2.1 Participants and design

The study enrolled HCWs aged 20–59 years from hospitals, local Centers for Disease Control and Prevention (CDC), and health community centers. Based on the total number of confirmed COVID-19 cases reported until March 15, 2020 ¹⁸, we divided all provinces in mainland China into two categories: (i) provinces with high cumulative COVID-19 incidence (exceeding 8 cases per 100,000), including Hubei, Guangdong, Zhejiang, and Henan, and (ii) provinces with low cumulative incidence (at or below 8 cases per 100,000).

On March 17 and 18, 2020, following the cessation of large epidemics in Hubei province and other parts of China, we initiated an online survey via WeChat, China’s most popular messaging app that has a monthly user base of more than 1 billion people. Two rounds of Delphi and preliminary experiment were conducted before formal survey. We first enrolled 20 HCWs, who then invited their WeChat colleagues or contacts (30 to 40 subjects) to the survey from both high and low cumulative incidence provinces to participate in the survey by sending them a WeChat link. To compare how COVID-19 vaccination acceptance differed among HCWs compared to the general population, we also invited adult participants from non-health care backgrounds, recruiting a sample that was approximately half the size of the HCW sample. We powered the study to estimate a COVID-19 vaccination acceptance rate of 65% in both groups, with a standard error of 0.05 in the HCW group and 0.08 in the general population group, and with 5% significance. This is based on the acceptance level of H1N1 pandemic vaccine (64%) or seasonal influenza vaccine (64%) measured in August 2009¹⁹. Using this, we estimated that 348 HCW participants (σ=0.05, α=0.05) and 136 general population participants (σ=0.08, α=0.05) were required. The study approval was obtained from the Institutional Review Board of Zhejiang Chinese Medical University (ZCMU) and anonymity was guaranteed to participants.

2.2 Data collection

The survey was conducted through a self-administered, anonymous questionnaire, which comprised of five sections: (i) demography; (ii) knowledge of and attitudes about SARS-CoV-2 infection, including susceptibility, health outcomes, subgroups at high risk of mortality, effective treatment, virus mutation, future epidemic trends and infection risk; (iii) acceptability of COVID-19 vaccination, including their opinion about the necessity of getting vaccinated, time before the vaccine is likely to become available, population groups most in need of vaccination, desire to receive vaccination, minimum level of vaccine effectiveness, most serious adverse effects following vaccination, highest number of doses needed, highest acceptable cost, and confidence in domestically manufactured vaccines; (iv) post-epidemic behavior, including intention to receive seasonal influenza vaccination or pneumococcal vaccination, and maintenance of other (unspecified) protective measures; and (v) preference about receiving COVID-19 vaccination (eSupplement1). All questions in the survey were designed based on evidence in the literature about the determinants of people’s decisions to get vaccinated^{2, 3, 7, 10}. The SO JUMP software (Ranxing Info Technology Co., Ltd), an online survey platform, was used for the online survey.

2.3 Discrete choice experiment

2.4 Data statistics

We did not run through mixed logit model in that most attributes’ levels were qualitative variables. To estimate the influence of each attribute in the DCE on participants’ preferences about vaccination, we used a binary logistic regression model shown in the equation below. This assesses the preference weight (odds ratio [OR]) for each attribute level to the probability of a participant preferring to be vaccinated. Here, β_1–11 are coefficients corresponding to each of the attributes in the DCE; X_i is a vector of alternative levels of the attributes selected by each individual; β₀ is a constant term, and Error indicates a random term following a type I extreme value distribution (Table 1). A descriptive analysis of respondents’ knowledge of, attitude about, and acceptance of COVID-19 vaccination was conducted using SPSS Version 25.0 (IBM Corporation, New York, NY, United States). The statistical analysis of the DCE results was carried out in STATA 16 (StataCorp LLC, Texas, United States).

3.1 Demographic characteristics

We approached 583 individuals to complete the survey. Of these, 22 (3.85%) did not complete the survey, and 20 (3.56%) were excluded based on our selection criteria (13 were living outside China, and 7 did not fulfil the age criteria of being 20-59 years old). Of the 541 participants selected for the analysis, 352 (65%) were HCWs and 189 (35%) were general population respondents, from 26 Chinese provinces.

Approximately 303 (56%) of the respondents were from high cumulative incidence provinces, and 70 (23%) were from Wuhan. Nearly 60% (207/352) of the HCWs were female, and 69% (242/352) were younger than 40 years. There were no significant differences between HCW and general population respondents in terms of sex (χ²=0.719, p=0.396) or age (χ²=4.834, p=0.184). Approximately 90% (313/352) of HCWs had a bachelor’s degree or higher qualification and were better educated than the general population sample (59.3%, 112/189; χ²=65.057, p<0.001; Table 2).

3.2 Knowledge of and attitude toward COVID-19

The majority of HCWs recognized that all age groups were susceptible to SARS-CoV-2 (n=305, 86.6%), and there was an increased risk of death in infected individuals who were older or had chronic diseases (n=349, 99.1%). Nearly 90% (314/352) of the HCWs believed that there would be slight mutations in the virus from the time it was first identified, and 7% (24/352) believed there were dramatic mutations. Of the 141 HCWs who reported their views about future COVID-19 trends, 47.5% (n=67) believed that incidence would diminish or disappear by summer (like SARS), 44.7% (n=63) believed that it would show annual cycles in incidence (like seasonal influenza), and only 8% (n=11) believed that it would become a endemic disease with year-round circulation (like tuberculosis). With regard to the risk of COVID-19, 66%(233/352) of the HCWs thought that they were at risk of becoming infected in the future, which was much higher than the equivalent proportion in the general population (52%, 98/189) (χ²=10.65, p<0.001).

3.3 Expectation about and acceptance of COVID-19 vaccination

With regard to the COVID-19 vaccines that were under development at the time of the survey, 95% of the HCWs thought that it was essential to be vaccinated. HCWs showed a higher acceptance to being vaccinated by a future vaccine. Compared to the general population, the HCWs believed that more time would be needed before vaccines were ready to be introduced for population-wide use (χ²=8.612, p=0.035), would accept a lower minimum threshold for vaccine effectiveness (χ²=12.178, p=0.007), more severe adverse effects (χ²=7.212, p=0.007), and more minor lesions (in the hypothetical situation that vaccination caused lesions) (χ²=7.515, p=0.006) than the general population (Table 2).

Approximately half the HCWs surveyed (47.7%, 168/352) believed that the COVID-19 vaccines produced by domestic manufacturers would be better than those produced abroad, which was lower than in the general population (55.6%, 105/189) (χ²=13.016, p=0.005). When asked about reasons for the potential inferiority of domestic vaccines among those who believed that imported vaccines would be better, approximately 60% (13/22) of HCWs suggested that this could be due to lower quality due to less stringently controlled production processes, and 60% (3/5) of the general population respondents believed that vaccine safety could not be fully guaranteed. In the entire study population, 78% (276/352) of HCWs agreed that the vaccine should be available free of charge although nearly half could afford to pay a price of 100–299 RMB (US$14–42) for a full course of vaccination. Three quarters (76.4%, 269/352) of the HCWs would opt to receive vaccination against COVID-19; however, nearly one fifth (19.3%, 68/352) were hesitant and needed more information before they could make a decision.

3.4 Protection adopted post epidemic

Approximately 70% (252/352) of the HCWs intended to be vaccinated against influenza and/or Streptococcus pneumoniae. The HCWs preferred to receive influenza vaccination (65.3%, 230/352), whereas the general population preferred to receive pneumococcal vaccination (58.7%, 111/189) (p=0.001). Both HCWs and the general population indicated that they would decrease the frequency of venturing into crowed areas (74%, 402/541), wash their hands more frequently (96%, 518/541), participate in physical exercise (93%, 501/541), and wear a face covering (67%, 364/541) in the future.

3.5 Vaccine preferences

As shown in Table 3, the DCE respondents indicated that the largest preference for a decision to receive COVID-19 vaccination (all p<0.001) are a seasonal (OR 4.37, 95% confidence interval [CI] 3.72–5.13) or persistent (OR 3.07, 95% CI 2.61–3.61) COVID-19 epidemic; a high vaccine refusal rate among social contacts of the respondents (OR 0.40, 95% CI 0.34– 0.47), neutrality (OR 0.41, 95% CI 0.35–0.49), high risk of infection (>30%; OR: 2.08, 95% CI 1.78–2.43), fee of 300 RMB (OR 0.48, 95% CI 0.40–0.56) and 100 RMB (OR 0.58, 95% CI 0.50–0.67); vaccine efficacy >80% (OR 1.77, 95% CI 1.54–2.04) and mild side effects only following immunisation (OR 1.54, 95% CI 1.35–1.75).

The decision to get vaccinated were reported in the general population, although the odds ratios were generally smaller. However, the general population respondents focus more on vaccine safety (OR 2.29, p<0.001), decision to vaccinate among social contacts (OR 0.21 and 0.48, p<0.001), and the case-fatality ratio (OR 0.78, p<0.05) than the HCWs. A sensitivity analysis in which all participants were included regardless of whether they answered the consistency question correctly, gave effects for both HCWs and the general population (see Table 3).

4.1 HCWs’ attitude to and willingness for COVID-19 vaccination

A similar percentage of HCW and the general population respondents (76.4% and 72.5%) were willing to get vaccinated. It is believed that the Chinese have much higher confidence in vaccination, especially for the Expanded Immunization (EPI) vaccines, than many other parts of the world. Although the COVID-19 vaccine was not included in the EPI list till now, the public in China are prone to have less hesitancy and more uptake. For the general population, China had much higher COVID-19 vaccine acceptance (91%) compared to most countries (range 38-98%)²⁰. According to a survey conducted in May 2020, 49% of US respondents reported willingness to get vaccinated against COVID-19, which is much lower than the proportion in this study. Furthermore, 31% of US respondents were vaccine hesitant, a higher proportion than in this study (19.3% of HCWs and 24.9% of the general population) ²¹. The proportion of respondents in our study willing the get vaccinated was also higher than reported in some other countries during the 2009 H1N1 pandemic (56.1% in the UK, 64% in the US, and 54.7% in Australia)^{19, 22, 23}.

COVID-19 vaccines were approved in late 2020 and early 2021 for population use in many countries across the world. During that time, HCWs’ attitudes towards COVID-19 vaccination has varied across countries. HCWs’ hesitancy toward COVID-19 may also vary over time. Overall Chinese HCWs seems to have more confidence and willingness to receive the vaccine than many parts of the world. A survey conducted in France, French-speaking Belgium and Quebec, Canada reported 71.6% acceptance of COVID-19 vaccination among HCWs²⁴. A study in October and November 2020 showed that 36% of US HCWs were willing to receive COVID-19 vaccine as soon as it became available, while 56% were hesitant²⁵. In Germany, 57% of front-line HCWs in emergency medical services (EMS) were willing to be vaccinated in a survey from December 2020 to January 2021²⁶. A systematic review indicated that 22.5% out of 76 471 HCWs worldwide reported COVID-19 vaccine hesitancy, with the proportion ranging between countries from 4.3% to 72%. Concerns about vaccine safety, efficacy and potential adverse effects were the leading reasons for hesitancy in the survey²⁷. Hence, continued surveillance of the attitudes of HCWs from diverse socioeconomic and geographical backgrounds is essential.

Most HCWs expressed willingness to receive COVID-19 vaccination after vaccine introduction, which, to a large extent, is attributable to their risk perception of infection and confidence in the efficacy and safety of the vaccine. HCWs have a much higher risk of getting infected with SARS-CoV-2 compared to the general population. For example, by Feb 11, 2020, 6 deaths and 1716 infections in HCWs occurred in mainland China (0.4% and 3.8% of the total deaths and infection respectively) ²⁸. In this study, 66% of the HCWs believed that they were at risk of being infected by SARS-CoV-2 from close contact with COVID-19 patients. Furthermore, uncertainty around how the epidemic may develop contributed to the risk evaluation: 60% of HCWs in the study were uncertain about the future trend in COVID-19 epidemics.

The HCWs showed a positive attitude toward the vaccine and were willing to accept a vaccine with effectiveness as low as 60–70% (similar to seasonal influenza vaccines), more severe adverse effects, and a higher number of doses than respondents from the general population. HCWs are vital to the public’s vaccination decisions, so their attitudes to vaccination can influence overall vaccine coverage. Adequate knowledge about a vaccine has been found in previous studies to increase HCW’s willingness to recommend that vaccine. A study in the UK reported that nurses with high levels of knowledge about seasonal influenza vaccination were more likely to recommend both influenza vaccination and vaccination in general to their patients ²⁹. In a study on HPV vaccination in Cameroon, one of the most important determinants of whether nurses would recommend the vaccine to the patients was their understanding of the efficacy and safety of the vaccine ³⁰.

In this study, nearly 20% and 25% of HCWs and the general population surveyed, respectively, were hesitant about COVID-19 vaccination, which may hamper establishment of widespread vaccine-induced protection in the population. Lack of preparedness for advising patients about vaccination and lack of training can inhibit HCWs in making effective vaccine recommendations. Since HCWs play a key role in reducing the disease burden of COVID-19 and helping to increase vaccine coverage, efforts need to be invested in increasing their knowledge and acceptance of vaccination.

4.2 Preference and relevant factors for COVID-19 vaccination

We found that respondents’ expectation about the future trend in COVID-19 was far more influential on their decision to want to be vaccinated than their infection risk or the case-fatality ratio of COVID-19. Their preferences may be motivated by their experience of seasonal influenza vaccination, since the seasonal influenza virus displays annual seasonality in temperate regions. Participants in the study were also influenced to the safety of future COVID-19 vaccines.

The behavior and attitudes of social contacts such as relatives, friends, and neighbors, play an important role in the decision for both HCWs and the general population. When vaccine safety or efficacy is uncertain, external cues such as vaccination uptake by others may be highly influential on vaccination intent ¹⁷. Vaccine hesitancy has been one of the most important public health issues in many countries, including some of those in Europe and North America. In contrast, vaccine hesitancy has remained limited in China, and vaccine confidence has recovered rapidly after vaccine safety or quality-related incidents in the past^{31, 32}.The reason for the resilience of vaccine confidence has been attributed by other authors to the collective culture in China where nonconformity is more strongly associated with deviance ³³. This suggests that appealing to a communal responsibility to protect others via indirect protection may be a strategy to maintain COVID-19 vaccine uptake.

Surprisingly, disease severity was not found to have a strong relationship with vaccination decisions among HCWs, and actually showed a negative (although non-significant, with CIs for the OR overlapping 1.0) association in the general population sample. However, similar relationships have been observed with other vaccine-preventable diseases such as seasonal influenza ¹⁷.

Efficacy was found to be the factor that most influenced vaccine selection in a DCE survey in UK, especially for those aged ≥ 55 years³⁴. An online DCE survey administered in China found that public prefer high effectiveness of the vaccine, followed by long protective duration, very few adverse events and being manufactured overseas and price was the least important attribute³⁵. The studies support our findings that future COVID-19 disease incidence and infection risk could significantly increase the probability of vaccination acceptance in HCWs.

4.3 Attitude toward domestically produced COVID-19 vaccine

Due to concerns about the quality of domestically produced vaccines, some HCWs may be hesitant to be vaccinated, which could have knock on effects on the vaccine recommendations they make to their patients. In this study, the proportion of HCWs who considered domestic vaccines to be inferior was twice that in the general population (7% vs 3%). Sixty percent of Chinese HCWs were concerned about less stringent quality control of vaccine production rather than the vaccine itself. Vaccine quality incidents reported in China over the past decade may have contributed to a decline in confidence in domestically-produced vaccines among the Chinese population, including HCWs ^{31, 36}. For example, on July 15, 2018, after an unannounced inspection of the Changchun Changsheng Biotech drug producer, the China Food and Drug Administration (CFDA) confirmed that fraudulent records were made during rabies vaccine production. Six days later, an article entitled “King of Vaccine” by “Beast Father” was posted in the WeChat media social account that received more than 2 million views within an hour of its release and triggered a public outcry. Rumors of ‘poisonous’ vaccines were quickly spread throughout online media ³⁶. Like human immune memory which is reactivated following another challenge with a similar antigen, HCWs and the general public may react negatively toward a future COVID-19 vaccine if serious adverse events (e.g. allergic reactions and blood clots following vaccination seen recently) are reported after vaccine introduction.

4.4 Potential positive influence including seasonal influenza and pneumonia vaccination following the pandemic

The COVID-19 epidemic could be used as an opportunity to improve public awareness of and knowledge about other infectious disease threats. Study respondents indicated increased use of personal protective measures that could reduce the incidence of other infections, such as face covering use and uptake of non-COVID-19 vaccines. Public awareness of seasonal influenza burden has increased in China in recent years, as seen in the shortages of vaccine supplies experienced in recent years ^{37, 38}. This awareness may have been heightened by the perception that SARS-CoV-2 was an “enhanced influenza” virus, and may have contributed to the increased willingness to receive seasonal influenza vaccination among HCWs. In the past, seasonal influenza vaccine coverage among Chinese HCWs has been low (17.7%, 95%CI 12.8–22.6) ³⁹, so one positive development from the COVID-19 pandemic may be increased vaccine uptake for seasonal influenza. Among general population respondents, we saw higher uptake of pneumococcal vaccination, possibly because of fear that COVID-19 may seriously damage the lungs ⁴⁰. In China, knowledge about the pneumococcal vaccine is poor and the coverage is very low. For example, in a survey conducted in the city of Hangzhou in 2013, only 21.8% of older adults were willing to receive the 23-valent pneumococcal polysaccharide vaccine (PPV23) vaccination, and actual coverage among older adults was only 1.23%⁴¹.

4.5 Limitations

Several limitations exist in the present study. First, the small sample was not randomly recruited and may not be representative of HCWs in the country; thus, any generalization of these results should be undertaken with caution. In particular, online recruitment via WeChat may have biased the study towards a better-educated and/or higher income sample. For example, only 36.2% of the HCWs in China have a bachelor’s degree or higher qualification, which is much lower than that in the present study (88.9%) ⁴². However, the HCW and general population samples are at least likely to suffer from the same kind of biases, since they were recruited simultaneously using the same methods, and have similar demographics in terms of gender and age distribution. The sample size is not large enough for us to report detailed interaction between attribute (levels) and individual characteristics. Second, participants were recruited and surveyed online instead of in a face-to-face interview, which may have led to potential biases. Third, we did not distinguish between different types of HCWs, such as doctors and nurses in hospitals, health care providers in the community, and staff in local Centers for Disease Control and Preventions, all of whom may have had different levels of knowledge about vaccination and decision-making influences. Fourth, the data could not be well run through the mixed logit model which we had designed for DCE. One possible reason is in that we lack relative data to determine quantitative variables in March 2020, which may hinder the inference and implication. Fifth, as a tool to measure the preferences, the DCE cannot include too many attributes and levels which makes it different from real situation. In addition, the consistency between the predicted results of the model and the actual selection results of individuals has not been confirmed.