Effectiveness of heterologous mRNA vaccine boosters during an Omicron wave of COVID-19: a cross-sectional study in Macao (China)

Abstract

Background

Inactivated vaccines have been widely used in China and many low- and middle-income countries, but real-world evidence on the protective effectiveness of heterologous mRNA boosters after a SinoPharm primary regimen during Omicron waves in Chinese populations remains limited. This study aimed to evaluate the real-world infection protection effectiveness of a SinoPharm primary regimen with mRNA boosters during the 2022 Omicron (BF.7 and BA5.2 variants) wave in Macao, China.

Methods

An online survey conducted for two days among Macao residents towards the end of the 2022 Omicron wave gathered 4,879 responses (61.7% female, n=3,010). Among the participants, 83.2% (n=4,052) reported receiving the SinoPharm primary vaccination regimen, while 20% (n=827) reported receiving mRNA boosters; 71.4% of respondents reported confirmed infections, with an average duration of 5.4 days testing positive. Logistic and Ordinary Least Squared regressions were utilized to analyze the infection risk and the number of days testing positive.

Results

When compared to those who received a three-dose inactivated vaccine, those with one-dose mRNA booster shows a significantly lower likelihood of confirmed infection [odds ratio (OR) =0.421, P<0.01]. In the event of an infection, they also experienced a decreased likelihood of developing fever (OR =0.290, P<0.01), and fewer sick days (−0.931, P<0.05).

Conclusions

Heterologous vaccination is recommendable for both the Chinese population and other low-to-middle-income countries that have predominantly adopted inactivated vaccines as their primary regimen. Future studies are needed to examine the long-term effects, which could not be assessed in this survey.

Highlight box.

Key findings

• The study’s key findings reveal that heterologous vaccination, combining SinoPharm’s inactivated vaccine with mRNA boosters, significantly lowers coronavirus disease 2019 (COVID-19) infection risk and reduces symptom severity during the 2022 Omicron wave in Macao, China. This suggests a promising approach, particularly for populations in low-to-middle-income countries that primarily rely on inactivated vaccines.

What is known and what is new?

• Inactivated vaccines are widely used globally, especially in low-to-middle-income countries. Heterologous boosters (combining different vaccine types) have shown improved immune responses compared to homologous boosters.

• The study demonstrates that heterologous vaccination, particularly combining SinoPharm’s inactivated vaccine with an mRNA booster, significantly reduces COVID-19 infection risks and symptom severity during the 2022 Omicron wave in Macao, China.

What is the implication, and what should change now?

• This study suggests that heterologous vaccination strategies should be more widely adopted, especially in regions heavily reliant on inactivated vaccines. Governments and health authorities might consider revising vaccination guidelines to include mRNA boosters after primary inactivated vaccine regimens for enhanced protection against COVID-19, particularly during prevalent variant outbreaks.

Introduction

Inactivated vaccines account for approximately half of the coronavirus disease 2019 (COVID-19) vaccine doses distributed worldwide (1,2), and these vaccines have primarily been administered in low- and middle-income countries that have limited vaccine infrastructure (3). Systematic reviews and meta-analyses of clinical and real-world observational studies (4,5) indicate that heterologous vaccination booster doses, following the initial vaccination with inactivated vaccines, can elicit an improved immune response (4-6) and display an overall higher level of protection effectiveness when compared to homologous vaccinations (7-9). Strategic implementation of heterologous boosters hence offers a promising avenue for maximizing protection and mitigating the impact of the virus in these populations (3,10).

Several studies have investigated the efficacy of heterologous vaccination boosters using real-world observational data from different countries, including Malaysia (11), Chile (12), China (13), and global cohorts (14). Despite the fact that a large portion of the population in China has primarily received inactivated vaccines (Sinopharm COVID-19 vaccines and CoronaVac), the heterologous vaccination strategy was not actively promoted. Qin and colleagues (15) offer a real-world observational study at this stage, researching the heterologous vaccine regimens in mainland China and examining combinations with booster doses of protein-subunit vaccine or adenovirus-vectored vaccine. Studies on heterologous vaccination using mRNA vaccines among the Chinese population after the zero-COVID-19 strategy are still lacking.

The Macao Special Administrative Region (SAR) of China during the COVID-19 pandemic followed mainland China to adopt the “Zero-COVID” public health controls. In order to facilitate effective monitoring of vaccine effects, the Macao Health Authority adopted only two types of COVID-19 vaccines: inactivated vaccines manufactured by SinoPharm and mRNA vaccines developed by BioNTech. The Macao Health Authority has strongly promoted “mixed vaccination” since November 2021, when vaccine boosters became available (16). By the end of 2022, approximately 94.3% of the Macao population had completed the COVID-19 primary vaccine series, and 57.2% had at least one booster dose (17,18). In December 2022, Macao experienced a wave of COVID-19 cases caused by the Omicron variants, namely the BF.7 variant from Beijing and the BA5.2 variant from Guangdong, with approximately equal proportion (19). Therefore, a unique real-world observational study opportunity arose in Macao to assess the effectiveness of the mRNA heterologous vaccination within the context of a Chinese population.

Utilizing self-reported infection and vaccination information obtained through an online survey, this study aims to evaluate the protection effectiveness of heterologous vaccine boosters (SinoPharm primary regime with mRNA boosters) as measured by the infection status and days testing positive among Macao’s general population during the 2022 Omicron wave period. We present this article in accordance with the SURGE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1979/rc).

Methods

This study is a cross-sectional study based on population-level observational data collected using an online survey. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline was followed in this study.

Study location

Located in the western Pearl River Delta by the South China Sea, with a population of approximately 680,000 in 2022 and a land area of 32.9 square kilometers (12.7 square miles), Macao is the most densely populated region in the world (20).

The Macao Health Authority launched the COVID-19 vaccination program on February 9, 2021, encouraging residents with vaccine hesitancy to start with the inactivated vaccine, which was evaluated as safe with sufficient protection effects (19,20). SinoPharm was the only type of inactivated vaccine provided. The majority of Macao residents had inactivated vaccines for their primary series (21). Since March 3rd 2021, mRNA vaccines have become available for those aged above 16 years, which was adjusted to those aged above 12 years later. The bivalent mRNA booster vaccines have been available in Macao since December 1st 2022 for those aged 12 years and older. All vaccinations in Macao were on a voluntary basis, with sufficient information disclosed to the general public (21).

The wave of Omicron infection cases lasted from December 13th 2022 to January 4th 2023, with a peak around December 22nd of 2022 (22,23). It is estimated that approximately 60–70% of the Macao population were infected (22,23), and a total of 57 death cases (a population mortality of approximately 0.012%) were officially reported (23). From December 13th 2022 to January 4th 2023, a total of 57 people died due to COVID virus infection. About 90% of them are over 70 years old, and about 60% are over 80 years old. Most of them have underlying diseases, and more than half of the cases have not received any COVID-19 vaccines (24).

Ethical considerations

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the research ethics committee of Macao University of Science and Technology (No. MUST-FDCT-20230116001, dated 27 December 2022). At the beginning of the survey, respondents were informed that participating in this questionnaire was anonymous and voluntary, and their privacy would be strictly protected. Their digital personal information would not be tracked or stored in any form. Informed consent was obtained from all respondents by checking a consent statement at the beginning of the survey questionnaire.

Questionnaire and sampling

A questionnaire was developed to collect the demographic information of respondents, current COVID-19 infection status and previous history, symptoms if currently infected, and vaccination history. A COVID-19 infection was defined as having a positive Rapid Antigen Test (RAT) or Nucleic Acid Test (NAT) result. Seven experts in public health were asked to evaluate the validity of the questionnaire content. Before formally launching the survey, three rounds of pre-tests and improvements were conducted, involving approximately 100 volunteers from the local community with diverse social backgrounds.

The survey was conducted over a two-day period, from December 27th to December 28th 2022, during the ending stage of the Omicron wave. The wave began to rise significantly on December 14th, reached its peak around December 22nd, and the number of infection cases steadily declined (23). The Macao Health Bureau explained that, as a small city within an extremely densely populated urban environment, Macao experienced a relatively short peak period of infections compared to large-scale outbreaks in other countries or regions (23).

A convenient sampling method was utilized for this study. The e-questionnaire was initially posted on the web link (https://f.wps.cn/w/02djQDa9/) and then widely distributed through “WeChat” platforms and local community WeChat groups. Since the onset of the COVID-19 pandemic, WeChat groups for resident communities have been established to facilitate the rapid dissemination of pandemic updates and emergency notifications. During the Omicron wave, as the majority of residents stayed at home and remained vigilant, they promptly responded to incoming information and quickly addressed urgent calls shared within their WeChat groups. Consequently, sponsored by a research institute in a prominent local university, the survey link for this study was widely shared across various community WeChat groups and received active responses. The research team also sent out follow-up notice during the survey period to enhance the response rate (24).

To minimize responses with incomplete answers, all fields were set as mandatory. To minimize the possibility of duplicate or fraudulent responses from the same source, IP address restrictions were implemented. Each IP address was allowed to submit only one response to the survey.

A total of 5,803 original responses were completed and collected in this survey, accounting for about 0.85% of Macao population.

Data selection and management

The raw survey data was further confirmed and cleaned based on the following exclusion criteria:

1. The age groups of “10 years old and younger” (n=251) and “61 years and older” (n=178). These two age groups were excluded due to their limited online activity and potential self-selection bias among those respondents.

2. Respondents who reported having received five vaccine doses (n=11). These responses were incorrect because Macao had only administered four doses in total.

3. Respondents who reported receiving any vaccines other than SinoPharm, mRNA, and bivalent mRNA (n=170), which were not officially offered to the general public by Macao heath administration.

4. Respondents who stated that they do not clearly remember the time of having the latest vaccines (n=261) so that the report errors may be reduced.

5. Respondents in vaccine-combination groups with fewer than 20 observations (n=53) so that the small sample error may be reduced for statistical estimation.

The final sample totaled 4,879 observations, accounting for 84% of the original sample size. This is equivalent to approximately 0.72% of the population in Macao. Figure 1 presents a flowchart illustrating the data processing procedure.

Figure 1.

Flowchart of data processing.

Outcomes and covariates

This study includes three health outcome variables: COVID-19 infection status, fever symptoms, and number of days tested positive. The infection status was measured as “confirmed infection” or “being uninfected” according to RAT or NAT results. The symptom of a fever (≥37.5 ℃) was used as a proxy indicator of the severity of the infection. The number of days testing positive if infected was calculated as the date of first testing negative minus the date of first testing positive.

The key explanatory variables are the vaccination types. We used “G” to represent the inactivated vaccine (SinoPharm), “M” for mRNA (original virus strain), and “B” for bivalent (bivalent original and Omicron BA.4/BA.5) vaccine. Accordingly, “GGGM” indicates a 3-dose regimen of inactivated vaccine and a 1-dose regimen of mRNA booster. See Table 1. Vaccination types of GGB, GGM, GGGB, GGGM, GGMB, and GGMM are identified as heterologous vaccinations.

Table 1. Descriptive characteristics of respondents (full sample).

Variable	All respondents (n=4,879)	Respondents with a dose within 3 months (n=1,298)†	P value (χ2)
Gender			0.097*
Male	1,869 (38.3)	530 (40.8)
Female	3,010 (61.7)	768 (59.2)
Age, years			<0.001***
12–17	1,162 (23.8)	228 (17.6)
18–30	822 (16.8)	234 (18.0)
31–40	1,520 (31.2)	401 (30.9)
41–50	947 (19.4)	290 (22.3)
51–60	428 (8.8)	145 (11.2)
COVID-19 infection status			<0.001***
Confirmed infection	3,485 (71.4)	750 (57.8)
Unconfirmed infection	331 (6.8)	95 (7.3)
Uninfected	1,063 (21.8)	453 (34.9)
Fever			<0.001***
No fever	2,285 (46.8)	807 (62.2)
With fever	2,594 (53.2)	491 (37.8)
Last vaccination time			<0.001***
Recent 1 month	963 (19.7)	963 (74.2)
Past 1–3 months	335 (6.9)	335 (25.8)
Past 3–6 months	801 (16.4)	–
Past 6–12 months	1,803 (37.0)	–
More than 1 year	977 (20.0)	–
Residence/work situation
Meet many people at work (yes)	2,958 (60.6)	821 (63.3)	0.09*
Live with more than 5 people (yes)	1,110 (22.8)	259 (20.0)	0.03**
Heterologous vaccination type‡
G000	115 (2.4)	42 (3.2)	0.07*
GG00	1,465 (30.0)	72 (5.5)	<0.001***
GGG0	1,391 (28.5)	194 (14.9)	<0.001***
GGGG	92 (1.9)	71 (5.5)	<0.001***
GGB0	62 (1.3)	60 (4.6)	<0.001***
GGGB	198 (4.1)	197 (15.2)	<0.001***
GGGM	118 (2.4)	100 (7.7)	<0.001***
GGM0	473 (9.7)	199 (15.3)	<0.001***
GGMB	91 (1.9)	91 (7.0)	<0.001***
GGMM	47 (1.0)	30 (2.3)	<0.001***
M000	34 (0.7)	19 (1.5)	0.008***
MM00	395 (8.1)	17 (1.3)	<0.001***
MMB0	31 (0.6)	30 (2.3)	<0.001***
MMM0	295 (6.1)	109 (8.4)	0.002***
MMMB	43 (0.9)	43 (3.3)	<0.001***
MMMM	29 (0.6)	24 (1.8)	<0.001***
Days of testing positive	5.40±2.88	4.81±2.90	–

Data are presented as n (%) or mean ± SD. *, P<0.1; **, P<0.05; ***, P<0.01. ^†, including those that had latest vaccine within 1 month (n=963), and those in past 1–3 months (n=335); ^‡, “G” represents the inactivated vaccine, “M” for mRNA (original virus strain), “B” for bivalent (bivalent original and Omicron BA.4/BA.5) vaccine, “0” for taking no vaccine. Accordingly, “GGGM” indicates a 3-dose regimen of inactivated vaccine and a 1-dose mRNA booster. COVID-19, coronavirus disease 2019; SD, standard deviation.

Control variables include some risk factors associated with demographic characteristics, such as gender, age, work environment, and living conditions. The timing of the most recent vaccination was captured using a set of dummy variables, which included categories such as “within 1 month”, “past 1–3 months”, “past 3–6 months”, “past 6–12 months”, and “more than 1 year ago”.

Statistical analysis

Multivariate logistic regression was applied to analyze the health outcome variables of infection status and fever, which were binary dependent variables. The odds ratio (OR) was calculated to assess the relationship between the independent variables and the health outcomes. A value greater than 1 indicates a positive association, meaning that an increase in the independent variable is associated with an increased odds of experiencing the health outcome of infection or having a fever. Conversely, a value less than 1 suggests a negative association, indicating that an increase in the independent variable is associated with a decreased odds of the health outcome. On the other hand, a multiple linear regression model utilizing Ordinary Least Square (OLS) regression was chosen to analyze the outcome variable “days tested positive”. This decision was based on the nature of the variable, which represents continuous numerical values and exhibits a normal distribution. The use of OLS regression is appropriate in this context as it accommodates the variance in individual health responses to the virus, allowing for a comprehensive examination of the factors influencing the duration of positive test results.

In the regressions, potential risk factors such as gender, age (25), work, and living conditions (26), as well as the timing of the latest vaccination (27), were identified and controlled for.

In order to account for the potential decline in vaccine effectiveness over time (28), this study conducted a robustness check by focusing specifically on respondents who clearly reported receiving their most recent vaccine shot within the past three months.

The statistical software Stata 14 (Stata Corp LP, College Station, TX, USA) was used to conduct all statistical analyses.

Results

Descriptive characteristics

As numerically reported in Table 1 and illustrated in Figure 2, female respondents account for 61.7% of the total sample, which is much higher than 53.1%, the female ratio of the Macao population. Due to the online survey method, the age group in their 50s accounts for a relatively small share; 71.4% of respondents in the full sample reported having confirmed infection, which is highly consistent with the official estimation made by the Macao Health Bureau (23). Approximately half of the respondents experienced a fever.

Figure 2.

Descriptive characteristics of respondents (full sample). COVID-19, coronavirus disease 2019.

As reported in the lower section of Table 1, this sample contains 16 heterologous vaccination types. According to the self-reported vaccination records, all respondents had at least one dose of the vaccine, with the groups of GG00 and GGG0 being the largest groups, accounting for approximately 30% and 28% of the sample size, respectively. Approximately 83% of respondents took the inactivated vaccine as the primary series, and 20% of the sample (n=827) reported heterologous vaccination with a variety of combinations; 17% of the sample took only mRNA and its booster.

The second column of Table 1 reports the descriptive statistics of those who received a dose of the vaccine within three months (n=1,298, 26.6% of the respondents).

Table 2 presents the descriptive characteristics of the heterologous vaccination groups within the full sample. The groups with heterologous vaccination (GGB, GGGB, GGM, GGMB, and GGMM) reported a much lower confirmed infection ratio than the groups with 2–4 doses of inactivated vaccines or with 2–3 doses of mRNA vaccines. When compared across the groups, data from days tested positive also demonstrated a consistent pattern. See Table 2.

Table 2. Descriptive characteristics of heterologous vaccination groups in the full sample^†.

Variable	GG00‡	GGG0	GGGG	GGB0	GGGB	GGGM	GGM0	GGMB	GGMM	MM00	MMM0
Gender
Male	596 (40.7)	442 (31.8)	38 (41.3)	23 (37.1)	81 (40.9)	49 (41.5)	195 (41.2)	48 (52.7)	24 (51.1)	155 (39.2)	121 (41.0)
Female	869 (59.3)	949 (68.2)	54 (58.7)	39 (62.9)	117 (59.1)	69 (58.5)	278 (58.8)	43 (47.3)	23 (48.9)	240 (60.8)	174 (59.0)
Age, years
12–17	680 (46.4)	15 (1.1)	1 (1.1)	15 (24.2)	0 (0.0)	0 (0.0)	89 (18.8)	0 (0.0)	0 (0.0)	216 (54.7)	79 (26.8)
18–30	245 (16.7)	228 (16.4)	3 (3.3)	15 (24.2)	43 (21.7)	30 (25.4)	80 (16.9)	7 (7.7)	3 (6.4)	71 (18.0)	63 (21.4)
31–40	371 (25.3)	584 (42.0)	28 (30.4)	18 (29.0)	64 (32.3)	34 (28.8)	160 (33.8)	28 (30.8)	10 (21.3)	72 (18.2)	77 (26.1)
41–50	130 (8.9)	408 (29.3)	32 (34.8)	13 (21.0)	60 (30.3)	35 (29.7)	93 (19.7)	38 (41.8)	18 (38.3)	28 (7.1)	48 (16.3)
51–60	39 (2.7)	156 (11.2)	28 (30.4)	1 (1.6)	31 (15.7)	19 (16.1)	51 (10.8)	18 (19.8)	16 (34.0)	8 (2.0)	28 (9.5)
COVID-19 infection status
Confirmed infection	1,099 (75)	1,080 (77.6)	70 (76.1)	29 (46.8)	96 (48.5)	58 (49.2)	315 (66.6)	51 (56.1)	29 (61.7)	291 (73.7)	211 (71.5)
Unconfirmed infection	110 (7.5)	81 (5.8)	5 (5.4)	7 (11.3)	15 (7.6)	7 (5.9)	31 (6.6)	7 (7.7)	2 (4.3)	28 (7.1)	21 (7.1)
Uninfected	256 (17.5)	230 (16.5)	17 (18.5)	26 (41.9)	87 (43.9)	53 (44.9)	127 (26.8)	33 (36.3)	16 (34.0)	76 (19.2)	63 (21.4)
Fever
Yes	895 (61.1)	819 (58.9)	42 (45.7)	16 (25.8)	66 (33.3)	27 (22.9)	218 (46.1)	25 (27.5)	18 (38.3)	222 (56.2)	133 (45.1)
Last vaccination time
Recent 1 month	24 (1.6)	108 (7.8)	45 (48.9)	58 (93.5)	190 (96.0)	73 (61.9)	149 (31.5)	90 (98.9)	13 (27.7)	5 (1.3)	72 (24.4)
Past 1–3 months	48 (3.3)	86 (6.2)	26 (28.3)	2 (3.2)	7 (3.5)	27 (22.9)	50 (10.6)	1 (1.1)	17 (36.2)	12 (3.0)	37 (12.5)
Past 3–6 months	218 (14.9)	285 (20.5)	16 (17.4)	1 (1.6)	1 (0.5)	18 (15.3)	84 (17.8)	0 (0.0)	15 (31.9)	47 (11.9)	90 (30.5)
Past 6–12 months	610 (41.6)	765 (55.0)	4 (4.3)	1 (1.6)	0 (0.0)	0 (0.0)	167 (35.3)	0 (0.0)	1 (2.1)	137 (34.7)	83 (28.1)
More than 1 year	565 (38.6)	147 (10.6)	1 (1.1)	0 (0.0)	0 (0.0)	0 (0.0)	23 (4.9)	0 (0.0)	1 (2.1)	194 (49.1)	13 (4.4)
Residence/work situation
Meet many people at work (yes)	708 (48.3)	992 (71.3)	71 (77.2)	38 (61.3)	139 (70.2)	88 (74.6)	301 (63.6)	73 (80.2)	39 (83.0)	176 (44.6)	182 (61.7)
Live with more than 5 people (yes)	1,150 (78.5)	1,001 (72.0)	70 (76.1)	49 (79.0)	159 (80.3)	96 (81.4)	371 (78.4)	78 (85.7)	40 (85.1)	322 (81.5)	245 (83.1)
Days tested positive	5.63±2.78	5.60±2.84	5.40±2.76	3.66±2.19	4.27±3.03	4.50±2.76	5.17±3.03	4.77±3.04	4.59±2.44	5.43±2.90	5.04±2.97
Observations	1,465	1,391	92	62	198	118	473	91	47	395	295

Data are presented as n (%), mean ± SD, or number. ^†, heterologous vaccination group G000, M000, MMMB and MMMM are not reported here due to small observation numbers; ^‡, “G” represents the inactivated vaccine, “M” for mRNA (original virus strain), “B” for bivalent (bivalent original and Omicron BA.4/BA.5) vaccine, “0” for taking no vaccine. Accordingly, “GGGM” indicates 3-dose of inactivated vaccine and 1-dose mRNA booster, and so on. COVID-19, coronavirus disease 2019; SD, standard deviation.

Analysis of protection effects

As shown in Table 3 and illustrated in Figures 3,4, when compared with those respondents with the three-dose inactivated vaccines (GGG0), those with the two-dose inactivated vaccines (GG00) were more likely to report a confirmed infection (OR =1.219, P<0.1), less likely to report being uninfected (OR =0.767, P<0.05), and more likely to report having a fever (OR =1.305, P<0.01), though there was no significant difference in the days of testing positive. Meanwhile, the four-dose inactivated vaccines (GGGG)’s health outcome indicators were, overall, not significantly different from those of GGG0.

Table 3. Regression analysis of protection effectiveness of heterologous vaccination (logistic regression and OLS) (full sample).

Variables	Confirmed infection (odds ratios)	Uninfected (odds ratios)	With fever (odds ratios)	Days tested positive (OLS)
Male	1.04 (0.072)	0.966 (0.073)	0.982 (0.061)	−0.210** (0.103)
Age, years
12–17	(Ref. group)
18–30	1.224* (0.136)	0.824 (0.100)	1.314*** (0.136)	0.337** (0.169)
31–40	1.714*** (0.186)	0.608*** (0.073)	1.536*** (0.151)	0.314** (0.154)
41–50	1.729*** (0.212)	0.579*** (0.078)	1.434*** (0.158)	0.262 (0.175)
51–60	1.566*** (0.234)	0.783 (0.126)	0.845 (0.115)	0.369 (0.230)
Heterologous vaccination type‡
GGG0	(Ref. group)
G000	1.158 (0.264)	0.83 (0.210)	1.455* (0.298)	0.423 (0.321)
GG00	1.219* (0.132)	0.767** (0.094)	1.305*** (0.121)	0.188 (0.145)
GGGG	1.236 (0.328)	0.803 (0.235)	0.828 (0.189)	−0.042 (0.368)
GGB0	0.495** (0.140)	1.715* (0.499)	0.385*** (0.120)	−1.722*** (0.426)
GGGB	0.461*** (0.086)	2.115*** (0.409)	0.539*** (0.101)	−1.150*** (0.366)
GGMB	0.591** (0.145)	1.616* (0.416)	0.407*** (0.107)	−0.630 (0.471)
GGM0	0.790* (0.099)	1.288* (0.177)	0.741*** (0.084)	−0.340* (0.195)
GGGM	0.421*** (0.089)	2.618*** (0.569)	0.290*** (0.069)	−0.931** (0.386)
GGMM	0.579* (0.184)	2.061** (0.680)	0.588* (0.185)	−0.775 (0.475)
M000	1.11 (0.439)	1.112 (0.457)	0.946 (0.339)	−0.884 (0.588)
MM00	1.23 (0.187)	0.8 (0.136)	1.108 (0.146)	0.003 (0.216)
MMB0	0.417** (0.162)	1.681 (0.658)	0.276*** (0.130)	−1.480* (0.800)
MMM0	1.079 (0.166)	0.892 (0.153)	0.740** (0.100)	−0.446** (0.227)
MMMB	0.368*** (0.123)	3.220*** (1.083)	0.496** (0.175)	−0.340 (0.723)
MMMM	0.602 (0.237)	1.682 (0.694)	0.329** (0.147)	−3.090*** (0.629)
Last vaccination time
Recent 1 month	(Ref. group)
Past 1–3 months	1.300* (0.193)	0.637*** (0.102)	1.272* (0.183)	−0.227 (0.247)
Past 3–6 months	1.559*** (0.200)	0.565*** (0.078)	1.357** (0.165)	0.236 (0.207)
Past 6–12 months	1.988*** (0.243)	0.458*** (0.060)	1.684*** (0.193)	0.201 (0.194)
More than 1 year ago	1.604*** (0.226)	0.547*** (0.084)	1.628*** (0.213)	0.133 (0.218)
Meeting many people at work	1.456*** (0.107)	0.624*** (0.050)	1.354*** (0.091)	0.026 (0.112)
Living with more than 5 persons	1.416*** (0.120)	0.673*** (0.064)	1.172** (0.085)	0.02 (0.113)
Constant	0.904 (0.133)	0.859 (0.136)	0.540*** (0.075)	5.176*** (0.233)
Observations	4,879	4,879	4,879	3,485†
R-squared	–	–	–	0.029
Pseudo R-squared	0.0578	0.069	0.0533	–

*, P<0.1; **, P<0.05; ***, P<0.01. ^†, only 3,485 observations available, because other respondents still tested positive at survey time. ^‡, “G” represents the inactivated vaccine, “M” for mRNA (original virus strain), “B” for bivalent (bivalent original and Omicron BA.4/BA.5) vaccine, “0” for taking no vaccine. OLS, Ordinary Least Squares.

Figure 3.

Forest plot of odds ratios.

Figure 4.

Forest plot of OLS coefficients for days tested positive. OLS, Ordinary Least Squares.

The protection effects of an inactivated vaccine primary series followed by mRNA booster(s) resulted in GGM0 (OR =0.790, P<0.1), GGGM (OR =0.421, P<0.01), and GGMM (OR =0.579, P<0.1), showing a significantly lower likelihood of infection. GGGM (OR =0.421, P<0.01) demonstrates strong infection protection effects comparable to GGB (OR =0.461, P<0.01) and even slightly better than GGGB (OR =0.591, P<0.05), which had bivalent mRNA boosters instead. The protection effects of these heterologous vaccination types were strong and consistent when the health outcomes of a fever and days of testing positive were examined. GGGM recipients were least likely to report a fever (OR =0.290, P<0.01), and the days of testing positive are less than GGG (the reference group) by approximately 0.931 days. GGMM recipients were also less likely to report a fever (OR =0.588, P<0.1).

Meanwhile, when compared with GGG0, those with heterologous vaccination of a bivalent mRNA booster (GGGB had much less likelihood of infection, are reflected by GGB0 (OR =0.495, P<0.05), GGGB (OR =0.461, P<0.01), and GGMB (OR =0.591, P<0.05). Their protection effects are strong and consistent when the health outcomes of a fever and days of testing positive are examined.

When compared to GGG0, MMB0 (OR =0.417, P<0.05), and MMMB (OR =0.368, P<0.01) are significantly associated with a much lower likelihood of reporting being infected. If infected, the likelihood of reporting a fever is significantly lower among recipients of MMB (OR =0.276, P<0.001) and MMMB (OR =0.496, P<0.05). MMB has shorter days of testing positive by approximately 1–48 days, whereas MMMB is insignificant.

MMM and MMMM are not associated with a lower possibility of reporting being infected, though the likelihood of reporting a fever is lower among recipients of MMM (OR =0.740, P<0.05) and MMMM (OR =0.329, P<0.05). The days of testing positive are significantly shorter by approximately 3 days among the recipients of MMMM.

We performed a robustness check using 1,298 observations from respondents who reported having received their latest vaccine dose within the past 3 months. In this way, the time effects of vaccine effectiveness were better controlled. The overall pattern of results was broadly consistent with the main findings, as shown in Table S1.

Discussion

Based on respondents’ self-reported infection status of COVID-19 and their vaccination history, this study evaluated the effectiveness of a heterologous vaccination regime (SinoPharm primary series followed by a dosage of mRNA or bivalent mRNA booster) among Macao residents during a huge Omicron wave of 2022. Prior to this, there were no population level observatory studies on heterologous vaccinations using mRNA vaccines among the Chinese population in the real world, though recent Phase 3 trials offer supporting evidence (29).

The findings of this study indicate that, when compared to three-dose homologous vaccines of inactivated virus (GGG), the heterologous vaccination provided in 1–3 months significant and strong protection from COVID-19 infection (Omicron BA.5), lowering the infection risks by approximately 30% (GGM0), 40% (GGMM), and 58% (GGGM). The findings highlight the remarkable efficacy of heterologous vaccination using the Omicron BA.4–BA.5 bivalent booster, which exhibited a substantial reduction in infection risks ranging from approximately 40% (GGMB) to 50% (GGB0 and GGGB). In addition, heterologous vaccination using the mRNA booster as the fourth dose (GGGM) demonstrated strong infection protection effects (OR =0.421, P<0.01), comparable to MMB0 (OR =0.417, P<0.05), though slightly weaker than MMMB (OR =0.368, P<0.01).

The findings in this study are consistent with existing studies regarding heterologous booster doses for inactivated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccines (8,9,13,29,30). This real-world observatory study provides additional empirical support that heterologous vaccination regimes exhibit superior effectiveness of protection when compared to homologous ones. The extra protection effects of heterologous vaccination regimes could be attributed to higher titers of S-specific neutralizing antibodies (4,31) and higher levels T-cell responses (7,11,32), providing positive impacts on the serological immune response (5,6).

Given that studies have shown that the preference of Macao residents for inactivated vaccines over mRNA vaccines is mainly influenced by concerns about side effects and vaccine safety (17,18), there are no significant differences in underlying health conditions or individual preventive strategies. Therefore, self-selection bias in choosing vaccination and booster types was not a significant concern in this study.

This study remains two limitations. Firstly, as an online survey on social media, this study is subject to response bias (who chooses to participate in the survey), particularly with regards to potential disparities in internet access and online social media usage across different demographic groups, for example, senior individuals, children and non-Chinese nationalities in Macao. Further studies specifically designed to examine the effectiveness of heterologous vaccination among special groups such as senior individuals and children are needed. Secondly, a mild survivor bias exists within the respondent pool because patients of severe cases of COVID-19 infection might be too ill to participate in the online survey. However, there is no major concern regarding this bias because free primary care and hospitalization services were available in Macao, and only a total of 144 severe COVID-19 cases (approximately 0.02% of Macao’s total population) (23) were officially recorded and reported during the Omicron wave. This low number of severe cases indicates that the study population was not significantly impacted by severe illness or hospitalization, reducing the potential influence of selection bias on the study results. Since this study assesses the short-term protection of vaccine boosters administered in the past three months, future research is needed to evaluate the durability and long-term sustainability of these effects (32).

Conclusions

Based on real-world data collected from 4,879 respondents in a representative survey conducted during the conclusion of the 2022 Omicron wave in Macao (China), this study provides evidence that among the young and middle-aged groups, heterologous booster doses of the mRNA (or bivalent mRNA) vaccine may offer a higher level of effective protection against Omicron variant infection within a three-month period, compared to homologous vaccination using inactivated vaccines. Additionally, this vaccination strategy was associated with a shorter recovery period.

This study, conducted as an online survey on social media, has several limitations. The respondents who received heterologous vaccine boosters were not randomly assigned, introducing a potential self-selection bias. The study excluded senior citizens and children due to their underlying health conditions, limited internet access, and online social media usage. No severe cases were reported by the survey respondents, partly due to the low number of severe COVID-19 cases during the Omicron wave (only 144 cases officially reported, accounting for 0.02% of Macao’s total population), as well as survivor bias. Future studies are needed to examine the long-term effects and safety of heterologous vaccinations, which could not be assessed in this online survey.

During widespread outbreaks of SARS-CoV-2 Omicron variants and potential future mutations, vaccination and boosters continue to be the most effective methods for safeguarding residents’ health. The findings of this study provide valuable reference information for government agencies to make effective evidence-based vaccination strategy recommendations. The implementation of heterologous vaccination is particularly recommended for both the Chinese population and other low- to middle-income countries that have predominantly adopted inactivated vaccines as their primary regimen.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the research ethics committee of Macao University of Science and Technology (No. MUST-FDCT-20230116001, dated 27 December 2022). Informed consent was obtained from all respondents by checking a consent statement at the beginning of the survey questionnaire.

Data Sharing Statement

Available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1979/dss

jtd-18-04-296-dss.pdf ^{(72.7KB, pdf)}

DOI: 10.21037/jtd-23-1979