Effectiveness of Enterovirus 71 inactivated vaccines against hand, foot, and mouth disease: A test-negative case-control study

ABSTRACT

The Enterovirus A71 (EV-A71) vaccine was introduced in China in December 2015 as a preventive measure against hand, foot, and mouth disease (HFMD) caused by EV-A71. However, the effectiveness of the vaccine (VE) in real-world settings needs to be evaluated. We conducted a test-negative case-control study to assess the effectiveness of EV-A71 vaccines in preventing EV-A71-associated HFMD. Children aged 6–71 months with HFMD were enrolled as participants. The case group comprised those who tested positive for EV-A71, while the control group comprised those who tested negative for EV-A71. To estimate VE, a logistic regression model was employed, adjusting for potential confounders including age, gender, and clinical severity. In total, 3223 children aged 6 to 71 months were included in the study, with 162 in the case group and 3061 in the control group. The proportion of children who received EV-A71 vaccination was significantly lower in the case group compared to the control group (p < .001). The overall VE_adj was estimated to be 90.8%. The VE_adj estimates for partially and fully vaccinated children were 90.1% and 90.9%, respectively. Stratified by age group, the VE_adj estimates were 88.7% for 6 to 35-month-olds and 95.5% for 36 to 71-month-olds. Regarding disease severity, the VE_adj estimates were 86.3% for mild cases and 100% for severe cases. Sensitivity analysis showed minimal changes in the VE point estimates, with most changing by no more than 1% point. Our study demonstrates a high level of vaccine effectiveness against EV-A71-HFMD, especially in severe cases. Active promotion of EV-A71 vaccination is an effective strategy in preventing EV-A71 infections.

KEYWORDS:

• Enterovirus A71
• Enterovirus A71 inactivated vaccine
• vaccine effectiveness
• test negative design
• hand, foot, and mouth disease

Introduction

Hand, foot, and mouth disease (HFMD) is a prevalent infectious disease predominantly affecting children under 5 y old, caused by enterovirus. While most cases of HFMD are self-limiting, a small proportion of patients may experience severe complications, including death.¹ The main viruses implicated in HFMD are Enterovirus A71 (EV-A71) and Coxsackievirus A16 (CV-A16),² with EV-A71 being the predominant strain associated with severe cases and mortality.^3,4

EV-A71 infections pose a significant threat to the health of infants and young children, leading to a substantial economic burden on society, and are a critical public health concern. A study conducted in Asian countries, where there is a high burden of HFMD, found that HFMD accounted for 96,900 disability-adjusted life years (DALYs) annually. The majority of these DALYs were due to years of life lost.⁵ In a prospective study conducted between 2016 and 2017 in three hospitals in Ho Chi Minh City, Vietnam, the average economic burden per hospitalized case of EV-A71 infection was estimated to be $400.80.⁶ It has been observed that EV-A71 infection with a longer duration of illness is associated with higher total costs and a greater loss of quality-adjusted life years.⁷

The first inactivated EV-A71 vaccine was licensed in China on December 3, 2015. Since then, three additional inactivated vaccines have been approved. Phase III clinical trials have demonstrated that these EV-A71 vaccines are highly effective against HFMD caused by EV-A71 (EV-A71-HFMD), with efficacy rates ranging from 94.8% to 97.4%^8,9 and 100% effectiveness in preventing severe cases of EV-A71-HFMD.¹⁰ It is important to note, however, that the robustness of the vaccine’s effectiveness against severe EV-A71-HFMD cases is limited due to the relatively small number of severe cases observed during Phase III clinical trials.

Test-negative case-control studies are commonly employed to assess the efficacy of influenza and rotavirus vaccines in real-world scenarios. Zhang et al. conducted a reanalysis of data from a randomized controlled trial (RCT) on EV-A71 vaccine efficacy using the test-negative case-control study design. The findings demonstrated that the vaccine effectiveness (VE) determined by both methods was essentially identical, implying that the test-negative case-control study can be utilized to evaluate the effectiveness of the EV-A71 vaccine in real-world settings.¹¹ To date, only three studies conducted in Beijing,¹² Henan,¹³ and Guangxi¹⁴ have employed the test-negative case-control study to evaluate the effectiveness of the EV-A71 vaccine against EV-A71-HFMD. However, these retrospective observational studies were constrained by limited statistical power due to recall bias or small sample sizes.

In order to address these concerns, a one-year prospective study was carried out using the test-negative case-control method in sentinel hospitals across three provinces. The objective of the study was to evaluate the effectiveness of EV-A71 vaccines in preventing both mild and severe cases of HFMD caused by EV-A71.

Materials and methods

Study design and setting

A prospective study was conducted to assess vaccine effectiveness using the test-negative case-control method. The study required sentinel hospitals to monitor over 60% of all cases in the respective provinces. Therefore, Henan, Hunan, and Yunnan provinces were chosen as the study sites. This study included patients with HFMD who visited sentinel hospitals in these provinces from January 1 to December 31, 2019. The EV-A71 vaccine coverage rate in these sentinel provinces was found to be 20% during the study period.¹⁵

Participants

Children between the ages of 6 and 71 months who were diagnosed with HFMD at select hospitals in Henan, Hunan, and Yunnan provinces were invited to participate in this test-negative case-control study. HFMD cases were diagnosed by clinicians using the Diagnostic Criteria for Hand, Foot, and Mouth Disease (2018 version)¹⁶ (Supplementary Material 1). Trained research personnel administered a standardized questionnaire (Supplementary Material 2) to collect demographic information, medical history, and EV-A71 vaccination history. Throat swabs and stool or anal swabs were collected within 24 hours of admission and stored at −80°C for testing. Any cases with unknown EV-A71 vaccination history or lack of specimen detection information were excluded from this study.

Prior to enrollment, written informed consent was obtained from the parents or legal guardians of all participants. The participants’ parents or guardians were provided with the informed consent form, which can be found in Supplementary Material 3.

Variable definitions

EV-A71 infections were confirmed using reverse transcription-polymerase chain reaction (RT-PCR) testing of throat swabs and stool or anal samples. HFMD cases that tested positive for EV-A71 from any single specimen by RT-PCR were classified as EV-A71-HFMD cases and assigned to the case group. HFMD cases that tested negative for EV-A71 from both specimens by RT-PCR were assigned to the control group.

For children who developed symptoms of illness 28 d after receiving the vaccination, we categorized those who received one dose as partially vaccinated and those who received two doses as fully vaccinated. The term “non-vaccinated” is used to describe individuals who have not received any dose of the EV-A71 vaccine or who had already contracted HFMD prior to receiving the first dose of the vaccine.

Data sources/measurement

Pathogen test results were collectively obtained by the Centers for Disease Control and Prevention (CDC) in three provinces. Prior to the study commencement, all researchers underwent uniform training and successfully passed a blinded sample assessment. The CDCs conducted RT-PCR testing for EV-A71 using standardized testing methods and uniform kits provided by Jiangsu Bioperfectus Technologies Co., Ltd. Regular reviews of the testing were conducted by provincial CDCs to mitigate potential biases (Supplementary Material 4).

The vaccination status for EV-A71 was obtained from multiple sources including the vaccination information registration system, vaccination record book, and parental self-report. In order to evaluate the reliability of the study, sensitivity analyses were conducted to exclude samples where vaccination status was obtained through parental self-report.

Study size

Using a significance level (α) of 0.05 (two-tailed) and a power (β) of 0.10, we calculated the sample size for an unmatched design. Assuming a vaccine effectiveness of 80% and a vaccine coverage rate of 20%, the optimal sample size was determined using the PASS 2021 software. Both the case group and control group should have a sample size of N1 = N2 = 93 cases.

According to a previous study, EV-A71 infection accounted for 40.0% of mild HFMD cases and 73.8% of severe HFMD cases.¹⁷ The detection rate of EV-A71 in HFMD patients was found to be 70.0%.¹⁸ Therefore, it is estimated that 332 mild HFMD cases and 180 severe HFMD cases should be investigated. To account for potential data loss during analysis, the estimated sample size will be multiplied by 1.2. Therefore, a total of 614 cases will be investigated, including 398 mild HFMD cases and 216 severe HFMD cases.

Statistical methods

Quantitative data, including age, were assessed for normality using the Shapiro-Wilk test. Non-normally distributed variables were presented using the median and interquartile range (IQR). Categorical variables, such as gender, underlying disease, disease severity, and vaccination status, were described using frequencies and percentages. Differences between the case and control groups were analyzed using the chi-square test or Fisher’s exact test. The relationship between EV-A71 infections and vaccination status was examined using univariate logistic regression to estimate the crude odds ratio (OR). To account for potential confounders such as age, gender, and disease severity, an unconditioned multivariate logistic regression model was used to calculate the adjusted odds ratio (OR_adj) and explore the association between EV-A71 infections and vaccination status.

VE was determined by comparing the odds of vaccination between cases and controls. VE was calculated as $\mathit{VE}=\left(1-\mathit{O}{R}_{\mathit{adj}}\right)\times 100\mathrm{\%}$. The 95% confidence interval (CI) for VE was calculated as $95\%\mathrm{CI}=\mathrm{VE}\pm 1.96\times \mathit{se}$, where se is the standard error of OR_adj in the multivariate logistic regression. Stratified analysis of VE estimates was performed based on age groups (6–35 months, 36–71 months), clinical severity (mild and severe), and dosage (partially and fully vaccinated).

Statistical analysis was performed using R software version 4.3.0 (R Foundation for Statistical Computing, Vienna, Austria, https://cran.r-project.org/). Statistical significance was defined as p < .05.

Results

Basic characteristics of HFMD cases

A total of 3,306 cases of HFMD were reported between January 1 and December 31, 2019. Out of these, 83 cases were excluded from the analysis. The reasons for exclusion were as follows: 23 cases were less than 6 months old at the time of HFMD diagnosis, 11 cases were more than 71 months old at the time of HFMD diagnosis, and 46 cases had received the EV-A71 vaccine within 28 d before the illness onset. After excluding these cases, a total of 3,223 cases were included in the analysis. Among these, there were 162 cases in the case group and 3,061 cases in the control group, as shown in Figure 1.

Figure 1. Flow chart of subject enrollment in the test-negative design case-control study.

The median age of patients with HFMD was 24 months (IQR: 16, 38), with 70.1% of cases occurring in children under 36 months of age. The male-to-female ratio was 1.5. Of the total cases, 2649 (82.2%) were classified as mild and 574 (17.8%) as severe. Among the severe cases, 49 (30.2%) tested positive for EV-A71, while among the mild cases, 113 (69.8%) were positive for EV-A71.

Table 1 presents a comparison of EV-A71-positive cases and EV-A71-negative controls. The proportion of severe cases in the case group was significantly higher than the control group (30.2% vs 17.2%, p < .001). Additionally, the proportion of vaccination in the case group was significantly lower than the control group (3.7% vs 29.4%, p < .001). No significant differences were observed in age group, gender, underlying diseases, and hospitalization between the two groups (p > .05).

Table 1. Characteristics and comparison of HFMD cases between EV-A71-positive cases and EV-A71-negative cases.

Characteristic	Total (N = 3223) No. (%)	EV-A71-positive cases (N = 162) No. (%)	EV-A71-negative cases (N = 3061) No. (%)	P
Age group (months)
6–35	2262(70.2)	115(71.0)	2147(70.1)	.818
36–71	961(29.8)	47(29.0)	914(29.9)
Gender
Male	1926(59.8)	98(60.5)	1828(59.7)	.845
Female	1297(40.2)	64(39.5)	1233(40.3)
Underlying diseases
Yes	17(0.5)	1(0.6)	16(0.5)	.585
No	3206(99.5)	161(99.4)	3045(99.5)
Severity of HFMD
Mild	2649(82.2)	113(69.8)	2536(82.8)	＜.001
Severe	574(17.8)	49(30.2)	525(17.2)
Hospitalization
Yes	1768(54.9)	92(56.8)	1676(54.8)	.669
No	1455(45.1)	70(43.2)	1385(45.2)
Vaccination
Yes	905(28.1)	6(3.7)	899(29.4)	＜.001
No	2318(71.9)	156(96.3)	2162(70.6)

Effectiveness of the inactivated EV-A71 vaccine

The adjusted vaccine effectiveness (VE_adj) against the target disease was estimated to be 90.8% (95% CI: 80.9%, 96.4%) overall. For fully vaccinated individuals, the VE_adj was estimated to be 90.1% (95% CI: 55.4%, 99.4%), and for partially vaccinated individuals, the VE_adj was estimated to be 90.9% (95% CI: 80.1%, 96.8%). Stratified analyses by age group and clinical severity yielded similar results, with higher VE estimates observed for fully vaccinated individuals compared to partially vaccinated individuals.

The overall VE_adj was 88.7% (95% CI: 74.9%, 96.0%) for children aged 6 to 35 months, with partially vaccinated children in this age group showing an VE_adj of 86.5% (95% CI: 38.3%, 99.2%), and fully vaccinated children in this age group showing an VE_adj of 89.1% (95% CI: 73.9%, 96.7%). For children aged 36 to 71 months, the VE_adj was 95.5% (95% CI: 75.7%, 99.7%), with partially vaccinated children in this age group showing a VE_adj of 100%, and fully vaccinated children in this age group showing a VE_adj of 94.8% (95% CI: 75.7%, 99.7%).

The overall VE_adj for mild cases was 86.3% (95% CI: 71.4%, 94.7%). For partially vaccinated mild cases, the VE_adj was 86.2% (95% CI: 37.3%, 99.2%), and for fully vaccinated mild cases, the VE_adj was 86.4% (95% CI: 69.8%, 95.2%). In severe cases, none of the EV-A71-positive cases had received the EV-A71 vaccine, resulting in a VE_adj of 100% for both partially and fully vaccinated individuals, as presented in Table 2.

Table 2. Crude and adjusted estimates of EV-A71 vaccine effectiveness against EV-A71-HFMD for different doses, ages, and clinical severity.

		Vaccinated No./total (%)		Vaccine effectiveness
Groups		Cases	Controls	Crude (95%CI)	Adjusted (95% CI)
Total		6/162 (3.7)	899/3061 (29.4)	90.8 (96.4,80.8)	90.8 (80.9,96.4)^‡
Partially vaccinated		1/157 (0.6)	143/2305 (6.2)	90.3 (56.3,99.5)	90.1 (55.4,99.4)^‡
Fully vaccinated		5/161 (3.1)	756/2918 (25.9)	90.8 (79.8,96.8)	90.9 (80.1,96.8)^‡
Age groups (months)
6–35		5/115 (4.3)	619/2147 (28.8)	88.8 (75.1,96.0)	88.7 (74.9,96.0)*
	Partially vaccinated	1/111 (0.9)	108/1636 (6.6)	87.1 (41.6,99.3)	86.5 (38.3,99.2)*
	Fully vaccinated	4/114 (3.5)	511/2039 (25.1)	89.1 (74.0,96.7)	89.1 (73.9,96.7)*
36–71		1/47 (2.1)	280/914 (30.6)	95.1 (77.3,99.7)	95.5 (79.0,99.7)*
	Partially vaccinated	0/46 (0)	35/669 (5.2)	100	100*
	Fully vaccinated	1/47 (2.1)	245/879 (27.9)	94.4 (74.1,99.7)	94.8 (75.7,99.7)*
Clinical severity
Mild		6/113 (5.3)	741/2536 (29.2)	86.4 (71.5,94.7)	86.3 (71.4,94.7)^⁑
	Partially vaccinated	1/108 (0.9)	121/1916 (6.3)	86.1 (37.1,99.2)	86.2 (37.3,99.2)^⁑
	Fully vaccinated	5/112 (4.5)	620/2415 (25.7)	86.5 (70.0,95.2)	86.4 (69.8,95.2)^⁑
Severe		0/49 (0)	158/525 (30.1)	100	100^⁑
	Partially vaccinated	0/49 (0)	22/389 (5.7)	100	100^⁑
	Fully vaccinated	0/49 (0)	136/503 (27.0)	100	100^⁑

^‡Adjusted for gender, age and clinical severity. Adjusted for gender and clinical severity. ^⁑Adjusted for age and gender.

The sensitivity analysis demonstrated minimal changes in the VE estimates, except for a slight variation of 1% point in the VE_adj for the mild-case group, when compared to the partially vaccinated group (86.2% vs 77.8%). Supplementary Material 5 provides further detailed results.

Discussion

The results of our study, which utilized a test-negative case-control design to assess the real-world effectiveness of the EV-A71 vaccine, indicate a VE_adj of 86.3% for mild cases and 100% for severe cases. Interestingly, the VE_adj for severe cases was higher compared to mild cases regardless of full vaccination status. This demonstrates that the EV-A71 vaccine is highly effective in providing protection against EV-A71-HFMD. In a Phase III clinical trial, the EV-A71 vaccine demonstrated 90% and 100% efficacy against mild and severe cases, respectively.^10,19 However, our VE estimates were slightly lower than those observed in the Phase III trial, due to various uncontrollable factors that could influence vaccine effectiveness in real-world settings.²⁰ Clinical trials, on the other hand, have more rigorous control over subjects, groups, types of vaccines administered, and vaccination timing. These factors may impact VE calculations to differing extents.

In our study, we observed that the VE for both partially and fully vaccinated individuals in mild cases was 86.2% and 86.4%, respectively. These values were slightly lower than the VE reported by Li (91.1% for fully vaccinated),¹³ but similar to the findings of Wang (83.7% for fully vaccinated).¹² Different sample inclusion criteria could explain the variation in results. Li’s study focused on hospitalized cases, while our study included outpatient cases, resulting in differences in clinical severity among the samples. Nonetheless, the definitions of severe and mild cases in Wang’s study were similar to ours, and the study populations had similar characteristics. Remarkably, the VE for severe cases was 100%, consistent with Wang’s results.¹² However, a study by Jiang et al. in Guangxi, evaluating VE of the EV-A71 vaccine for severe HFMD, reported a VE of 86.6%,¹⁴ The dissimilarities in population characteristics and sample size could account for these disparate observations. Jiang et al. included 2,779 subjects in their study, with six severe cases of EV-A71 infection occurring after vaccination. In contrast, we did not observe any severe cases of infection after vaccination in our study.

The prevalence of VE was found to be higher among children aged 36–71 months compared to those aged 6–35 months, regardless of their vaccination status (partially vaccinated: 100% vs. 86.5%; fully vaccinated: 94.8% vs. 89.1%). Similar trends have been observed in studies conducted in other regions, indicating that the EV-A71 vaccine offers greater protection in older age groups.^12–14 Furthermore, a comparative analysis of EV-A71 vaccine immunogenicity in children of different ages showed that the geometric mean titer (GMT) of EV-A71 neutralizing antibodies was higher in children aged 36–59 months compared to those aged 3–35 months.²¹

In order to assess the reliability of VE estimates, we conducted a sensitivity analysis for the EV-A71 vaccine by altering the source of vaccination records. Specifically, we excluded 18 cases in which vaccination records were obtained through parental recall. Our analysis revealed that the main findings remained unchanged, with minimal changes observed in VE point estimates (within 1% point). These results indicate that the inclusion of parent recalls in our study did not significantly impact the findings.

According to Cowling B. J., the most significant source of measurement error in estimating influenza vaccine effectiveness using the test-negative design is the determination of vaccination status.²² In our prospective study, we implemented a rigorous experimental design to address this issue. We obtained vaccination records from the Vaccination Information Registry System or the vaccination record book, ensuring highly accurate information and minimizing recall bias. To reduce classification errors from false-negative test results, we performed two tests on patient specimens and combined the results. Additionally, our study included a large number of severe cases. We analyzed vaccine effectiveness in severe cases and adjusted for clinical severity to minimize bias in the analysis of other subgroups.

A limitation of this study is its geographic scope, as it focused solely on three Chinese provinces with high incidence rates of HFMD: Henan, Hunan, and Yunnan. As a result, generalizing the findings to other regions or countries should be done cautiously.

In conclusion, our study demonstrated high vaccine effectiveness against EV-A71-HFMD in children, especially in severe cases. Active promotion of EV-A71 vaccination is an effective strategy for preventing EV-A71. Nonetheless, the current vaccination rate among children is low, and local governments should actively encourage age-appropriate EV-A71 vaccine uptake.

Author contributions

Zhaorui Chang designed the study and conducted the survey. Yutong Zhang analyzed the data and drafted the paper. Jinzhao Cui, Fengfeng Liu, Yang Song, and Yanzhe Liu revised the paper critically for intellectual content. Quanyi Wang, Yanping Zhang and Zhongjie Li completed final approval of the version to be published. All authors contributed to the discussions and the final draft.

Ethical approval

The study was approved by the Ethics Committee of the Chinese Center for Disease Control and Prevention [No.201904] (Supplementary Material 6).

Acknowledgments

We would like to express our gratitude to the participants for their participation in the study. We would like to express our gratitude to the Hunan Province Center for Disease Control and Prevention, Yunnan Province Center for Disease Control and Prevention, and Henan Province Center for Disease Control and Prevention for their valuable assistance in carrying out this project. Authors acknowledge Zhaorui Chang for her valuable ideas.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available on request from the corresponding author, CHANG, upon reasonable request.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2024.2330163.

Additional information

Funding

This study was supported by the National Science and Technology Major Project of China [grant number 2016ZX09101120-004], Beijing Municipal Natural Science Foundation [grant number L192014], and Key research projects of Beijing Natural Science Foundation-Haidian District Joint Fund [grant number L192012].