5C psychological antecedents of vaccine hesitancy toward maternal RSV vaccines and infant monoclonal antibodies among pregnant and postpartum women and their partners in China: A multicenter, dyadic survey design, cross-sectional study

ABSTRACT

We assessed vaccine hesitancy toward two novel respiratory syncytial virus (RSV) immunization strategies, maternal vaccines and infant monoclonal antibodies (mAbs), among 4422 couples across eight Chinese provinces, via a multicenter cross-sectional survey with convenience sampling from medical institutions. Vaccine hesitancy was measured using the 5C scale (confidence, complacency, constraints, calculation, collective responsibility), with total scores (higher indicating lower hesitancy) as the primary outcome. Linear mixed-effects models analyzed sociodemographic and pregnancy-related factors. Women had marginally lower total 5C scores than partners for maternal vaccines (45.88 [SD 5.90] vs. 46.25 [5.74]; Cohen’s d = –0.06, 95% CI: −0.09 to −0.03; p < .001), with no significant difference for mAbs. For both strategies, calculation scored highest in both sexes. Within-couple correlations across domains for the same strategy were low (r ranged from 0.31 to 0.48). In multivariable models, older age, rural residence and lower income were associated with greater hesitancy for both strategies (e.g. age: maternal vaccines, β = −0.09, 95% CI: −0.12 to −0.05; mAbs, β = −0.09, 95% CI: −0.12 to −0.06). Among women, third-trimester status and prior children were each independently associated with higher hesitancy. Prior influenza vaccination during pregnancy was associated with lower hesitancy toward maternal vaccines (β = 0.73, 95% CI: 0.32 to 1.13) but not mAbs. Ultimately, Chinese expectant and new parents show a hesitancy profile toward RSV immunization strategies dominated by active information-seeking. Prenatal counseling and broader communication efforts should include both parents. Efforts should prioritize older, rural, and lower-income couples.

Background

Respiratory syncytial virus (RSV) is a leading cause of acute lower respiratory tract infections in young children and poses its greatest threat during the first 6 months of life, before infants mount a robust adaptive immune response.^1,2 In China, national surveillance from 2009 to 2020 identified RSV as the predominant pathogen among children hospitalized with severe community-acquired pneumonia.³

Two complementary immunization strategies have been developed to address this vulnerability: maternal immunization during pregnancy and infant immunization with long-acting monoclonal antibodies. Maternal RSV vaccines (e.g., Abrysvo) are administered during pregnancy to induce transplacental antibody transfer, providing passive protection from birth through the first 6 months of life.^4,5 Alternatively, long-acting monoclonal antibodies (mAbs, e.g., Nirsevimab/Beyfortus™) are administered directly to newborns, conferring immediate humoral immunity through the RSV season, reducing severe RSV-related outcomes through viral neutralization.^6,7 The implementation status of RSV preventive products currently differs in mainland China. Maternal RSV vaccines had not yet been licensed, whereas the infant mAbs (Nirsevimab and Clesrovimab) had received regulatory approval but remained accessible mainly through market-based, voluntary self-paid pathways rather than a universally implemented national immunization program.^8–10 In this context, understanding likely demand for and hesitancy toward these products is important for future implementation planning.

Vaccine hesitancy, defined as the delay in acceptance or refusal of vaccination despite the availability of services, has emerged as a significant barrier to achieving high immunization coverage. In 2012, the Strategic Advisory Group of Experts on Immunization (SAGE) of the World Health Organization (WHO) introduced the “3C model,” identifying confidence (trust in vaccine safety and systems), complacency (lack of perceived disease risk), and constraints (accessibility and affordability of vaccination) as primary determinants of hesitancy.¹¹ This model was later expanded to the 5C framework, incorporating calculation (engagement in extensive decision-making) and collective responsibility (willingness to protect others through herd immunity), providing a more comprehensive psychological lens to assess vaccine hesitancy.¹² The 5C model has been widely validated and is increasingly applied in public health research and intervention design.

Among pregnant and postpartum individuals, decisions regarding immunization are influenced by additional risk-benefit deliberations distinct from the general population, including heightened concerns about fetal safety, medical interventions during pregnancy, and trust in the health system.^13–15 In addition, women’s vaccination decisions may be influenced by their partners, making it important to consider the broader household decision-making context rather than focusing on mothers alone.^16–18 Moreover, previous studies have predominantly measured overall acceptance, which alone cannot explain the reasons for hesitancy, further examination of psychological antecedents is warranted.^19,20 We therefore conducted a large multicentre dyadic survey across eight Chinese provinces to: (1) characterize the psychological profile of vaccine hesitancy toward maternal RSV vaccines and infant mAbs among pregnant and postpartum women and their partners; (2) evaluate within-couple concordance; (3) identify sociodemographic and perinatal factors associated with hesitancy toward these RSV immunization strategies.

Methods

Study design and setting

This was a dyadic, cross-sectional study conducted between December 2024 and August 2025 across eight provinces in five regions of China: North China (Beijing, Tianjin), Northwest China (Gansu), East China (Zhejiang, Fujian), Central-South China (Guangdong, Henan), and Southwest China (Sichuan). These provinces were purposefully selected to reflect geographic diversity and variation in socioeconomic development levels, supported by regional Gross Domestic Product (GDP) data published by the National Bureau of Statistics of China.²¹ The study employed a dyadic survey design, collecting paired data from pregnant or postpartum women and their partners. This study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.²²

Participants

Eligible participants were women who were currently pregnant (in any trimester) or postpartum within 6 weeks, along with their cohabiting spouses or long-term male partners. The inclusion criteria for women and their partners were: (1) age 18 or above; (2) ability to communicate and comprehend normally; and (3) provision of written informed consent. Participants were excluded if either partner had cognitive or psychiatric impairments that interfered with questionnaire completion, or if only one member of the couple agreed to participate despite eligibility. Recruitment was conducted using convenience sampling across 14 medical institutions, including provincial/municipal maternal and child health hospitals, tertiary general hospitals, and community health service centers. Clinical staff informed eligible women during routine prenatal/postnatal visits or hospitalization, and interested couples were referred to trained investigators for consent, enrollment, and data collection. Each member of the couple completed the survey independently, with trained investigators available to clarify procedural questions. Participants’ questionnaire data were collected using an online questionnaire platform (https://www.wjx.cn). Researchers performed daily checks on completeness and logical validity to ensure data quality.

Ethical approval

The study protocol was approved by the Ethics Committee of Zhejiang University School of Public Health (ZGL202410–1). All participants provided written informed consent, and data confidentiality was maintained throughout the study.

Survey questionnaire

A self-administered questionnaire was developed, structured in three main sections. The first section collected sociodemographic characteristics, including age, urban or rural residence, education level, occupation, whether any family member was a healthcare professional, annual household income per capita and self-rated health. For pregnant/postpartum women, perinatal variables were additionally collected, including current perinatal stage, routine prenatal care completion, number of children (except for this pregnancy or postpartum period), pregnancy-related complications (which included gestational hypertension, gestational diabetes mellitus, thyroid disorders, and other complications such as anemia and thrombophilia), and history of influenza and COVID-19 vaccination during any pregnancy. The second section measured vaccine hesitancy using the validated 5C psychological antecedents of vaccination scale (Table S1). This model captures five domains: confidence (trust in vaccine safety and systems), complacency (lack of perceived disease risk), constraints (accessibility and affordability of vaccination), calculation (engagement in extensive decision-making), and collective responsibility (willingness to protect others through herd immunity). Each domain was measured using three items on a 5-point Likert scale ranging from “strongly disagree” to “strongly agree.” The 5C model was utilized among both pregnant/postpartum women and their partners regarding two immunization strategies. Internal consistency reliability of the model was assessed using McDonald’s ω coefficient (Table S2). Values for the main domains showed good reliability (McDonald’s ω coefficient > 0.7).²³

We calculated the mean score of all items within each domain. The higher the domain mean score, the stronger the confidence in RSV immunization, the greater the complacency about the risk of RSV infection, the more constraints in accessing RSV immunization services, the higher the degree of calculation when searching for RSV immunization information, and the stronger the collective responsibility to protect others. The total score for 5C model was also calculated (with a full score of 75). Previous studies showed that high confidence in vaccine safety and efficacy, or a strong sense of collective responsibility, typically leads to greater recognition of the benefits of vaccination, indicating a low level of vaccine hesitancy.¹¹ On the other hand, high complacency, various constraints, and excessive calculation often signify a high level of vaccine hesitancy.¹² Therefore, when calculating the total score, items in the complacency, constraints, and calculation domains were reverse-coded, such that a higher total score reflects a lower level of vaccine hesitancy. This reverse coding approach for the 5C scale has been adopted in previous relevant research.²⁴

Sample size calculation

Based on the vaccine hesitancy rate for influenza vaccination among pregnant women in previous studies,²⁵ the rate of willingness to receive maternal RSV vaccines/mAbs (p) was assumed to be 24.7%. Type I error (α) was set at 0.05, corresponding to a Z-score of 1.96 for the 95% confidence interval (CI). The permissible error (δ) was set at 3%. Accounting for potential non-response rates, the calculated sample size was increased by 20%. Consequently, a minimum of 954 women per stage (i.e., first trimester, second trimester, third trimester, and postpartum) were required, resulting in a total of 3816 women. Simultaneously, 3816 expectant/new fathers were to be recruited in pairs with the enrolled women. After excluding dyads with unmatchable data, the final analytical sample included 4422 couples (8844 individuals).

$$\mathit{n}=\frac{Z_{\alpha }^2\ast \mathit{p}\ast \left(1-\mathit{p}\right)}{{\mathit{\delta }}^2}$$

Statistical analysis

Descriptive statistics were used to summarize the characteristics of participants. For continuous variables, means and standard deviations (SD) were used to describe normally distributed data, while medians and interquartile ranges (IQR) were applied to characterize non-normally distributed data. Categorical variables were described using frequencies and percentages. All 5C domain scores were treated as continuous variables. Both maternal RSV vaccines and infant mAbs were assessed separately.

The percentage of missing values for variables ranged from 0.02% to 5.56% (Table S3). Age below 18 y was treated as an outlier. The participant-level missingness rate (defined as the proportion of participants with at least one missing or abnormal value) was 4.2% for women and 6.5% for partners. Missing or abnormal values were treated as missing observations and handled using Multiple Imputation by Chained Equations (MICE). The MICE algorithm was run under the Missing at Random (MAR) assumption, generating five imputed datasets with ten iterations per dataset. Convergence was confirmed by visual inspection of trace plots (Figure S1), ensuring stable mean and variance trajectories. All statistical analyses were performed separately on each imputed dataset, and the final pooled parameter estimates were calculated using Rubin’s rules.

To compare the differences in the 5C domains and total scores of immunization strategies within couples, we conducted a paired t-test. After visual inspection of histograms and Q-Q plots (Figures S2 and S3), the data were approximately consistent with normality. Pearson correlation coefficients were calculated to assess associations between 5C scores within individuals (i.e., comparing maternal vaccines and infant mAbs scores for the same individual) and within couples (i.e., comparing the same strategy across each couple). Correlation strength was interpreted using conventional thresholds: negligible (0.00–0.30), low (0.30–0.49), moderate (0.50–0.69), and strong (≥0.70).²⁶

Given the dyadic nature of the survey, data from women and their partners were clustered within households and violate the assumption of independent observations. To rigorously account for this non-independence and avoid underestimating standard errors, Linear mixed-effects models (LMMs) were employed. In these models, the dyad was specified as a random intercept, while socio-demographic and perinatal factors were entered as fixed effects. Prior to model fitting, Harman’s single-factor test was used to assess common method bias, which indicated no severe bias (the first factor explained 23.9% of the variance). To ensure the robustness of LMMs, residual normality and homoscedasticity were visually confirmed (Figure S4), and multicollinearity among fixed effects was evaluated using the variance inflation factor (VIF). Values >5 were considered to indicate potentially problematic multicollinearity.²⁷ Most VIF values were low (1.07–1.24), whereas the VIF for gender was 6.68, which was expected given its inclusion in interaction terms and likely reflects structural rather than problematic multicollinearity (Table S4).

LMMs were fitted separately for maternal RSV vaccines and infant mAbs strategies. In each model, the 5C total score was specified as the dependent variable. A random intercept was included at the couple level to account for within-couple clustering. Furthermore, the intraclass correlation coefficient (ICC) was calculated to quantify the proportion of total variance attributed to this dyadic interdependence. Fixed effects included sociodemographic covariates, including gender, age (continuous), residence (urban/rural), education level (ordinal), occupation (categorical), any family member as a healthcare professional (yes/no), annual household income per capita level (ordinal) and self-rated health (ordinal). Covariates reflecting perinatal status were included for women participants only and comprised perinatal stage (first trimester/second trimester/third trimester/postpartum), completion of routine prenatal care (completely finish/incompletely finish/no check), No. of children (0/1/≥2), pregnancy-related complications (yes/no), influenza vaccination history during pregnancy (yes/no), and COVID-19 vaccination history during pregnancy (yes/no). These variables were incorporated into the models via interaction terms with a binary indicator for female gender, thereby restricting their effects to women. Model estimates were expressed as unstandardized β coefficients with corresponding 95% CIs.

To validate the robustness of these results, a sensitivity analysis was conducted, which was restricted to participants without missing observations (Table S5). Furthermore, we applied the false discovery rate (FDR) method to correct for multiple testing, and adjusted p-values were reported.

All reported p-values were two-sided, with statistical significance defined as p < .05. Statistical analyses were conducted using R version 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria), and mixed-effects models were estimated using the lme4 package.

Results

Participants’ characteristics

A total of 4422 couples (8844 individuals) were included in the primary analysis (Figure 1). Among women participants, 77.7% were currently pregnant and 22.3% were postpartum within 6 weeks.

Figure 1.

Flowchart of participant recruitment and final sample inclusion.

Notes. MICE, Multiple Imputation by Chained Equations.

The mean age was 29.17 y (SD = 4.11) for women and 30.21 y (SD = 4.32) for partners. The majority of participants resided in urban areas (women: n = 3552, 80.3%; partners: n = 3532, 79.9%) and held a college degree or higher (women: n = 3187, 72.1%; partners: n = 2969, 67.1%). A minority of participants reported having a healthcare professional in the family (women: n = 1245, 28.2%; partners: n = 1347, 30.5%). Approximately 40% of participants (women: n = 1693, 38.3%; partners: n = 1763, 39.9%) reported an annual household income per capita above 100,000 CNY. Among women, approximately 70% (n = 3168, 71.6%) completed all routine prenatal care. In addition, more than half of women had no children (n = 2886, 65.3%). 68.2% of women (n = 3014) had any pregnancy-related complications. Only 19.9% of women (n = 881) reported receiving influenza vaccination during pregnancy, and 20.7% (n = 916) reported receiving COVID-19 vaccination. The full distribution of demographic and perinatal characteristics was presented in Table 1.

Table 1.

Characteristics of the included participants.

Characteristics		Women	Partners
Socio-demographic variables (n = 8844)
Age, mean (SD)		29.17 (4.11)	30.21 (4.32)
Residence, No. (%)
Urban		3552 (80.3)	3532 (79.9)
Rural		870 (19.7)	890 (20.1)
Education level, No. (%)
Below college degree		1235 (27.9)	1453 (32.9)
College degree or higher		3187 (72.1)	2969 (67.1)
Occupation, No. (%)
Medical worker		331 (7.5)	178 (4.0)
Civil servant/Enterprise employee		1781 (40.3)	1816 (41.1)
Self-employed/Freelancer		1469 (33.2)	1882 (42.6)
Others		291 (6.6)	301 (6.8)
Unemployed		550 (12.4)	245 (5.5)
Any healthcare professional in the family, No. (%)
Yes		1245 (28.2)	1347 (30.5)
No		3177 (71.8)	3075 (69.5)
Annual household income per capita (CNY), No. (%)
≤50,000		1775 (40.1)	1801 (40.7)
50,000–100,000a		954 (21.6)	858 (19.4)
>100,000		1693 (38.3)	1763 (39.9)
Self-rated health, No. (%)
Very good		773 (17.5)	1006 (22.8)
Good		1811 (40.9)	1848 (41.8)
Fair		1501 (33.9)	1295 (29.3)
Not good		337 (7.6)	273 (6.2)
Perinatal variables (n = 4422)
Perinatal stage, No. (%)
First trimester		539 (12.2)	–
Second trimester		1072 (24.2)	–
Third trimester		1824 (41.3)	–
Postpartum		987 (22.3)	–
Routine prenatal care completion, No. (%)
Completely finish		3168 (71.6)	–
Incompletely finish		1045 (23.6)	–
No check		209 (4.7)	–
No. of childrenb, No. (%)
0		2886 (65.3)	–
1		1068 (24.1)	–
≥2		468 (10.6)	–
Any pregnancy-related complicationsc, No. (%)
No		1408 (31.8)	–
Yes		3014 (68.2)	–
Influenza vaccination during pregnancyd, No. (%)
No		3541 (80.1)	–
Yes		881 (19.9)	–
COVID-19 vaccination during pregnancyd, No. (%)
No		3506 (79.3)	–
Yes		916 (20.7)	–

Notes: SD, standard deviation.

^aThe range of 50,000–100,000 does not include 50,000.

^bNo. of children refers to the number of children the participant had prior to the current pregnancy.

^cAny pregnancy-related complications include gestational hypertension, gestational diabetes mellitus, thyroid disorders, and other complications.

^dInfluenza and COVID‑19 vaccination during pregnancy refer to receipt of the respective vaccine during the index pregnancy or during any prior pregnancy.

Vaccine hesitancy profiles by 5C model

Among pregnant/postpartum women, the mean 5C total score for maternal RSV vaccines was 45.88 (SD = 5.90), and for infant mAbs it was 46.25 (SD = 5.47). Across both strategies, the highest domain score was found in calculation (maternal vaccines: 3.60 ± 0.91; mAbs: 3.62 ± 0.93), while the lowest scores were recorded for constraints (maternal vaccines: 2.75 ± 0.88; mAbs: 2.73 ± 0.90) and complacency (maternal vaccines: 2.89 ± 0.81; mAbs: 2.85 ± 0.81) (Table 2).

Table 2.

Within-couple comparisons of 5C scores for RSV immunization strategies.

		Domains (mean ± SD)		Paired-samples t-test
		Women	Partners	Adjusted p-values	Cohen’s d (95% CI)
5C model of maternal vaccines
Confidence		3.26 ± 0.86	3.27 ± 0.85	.662	−0.01 (−0.04 to 0.02)
Complacency		2.89 ± 0.81	2.85 ± 0.87	.007	0.05 (0.02 to 0.08)
Constraints		2.75 ± 0.88	2.71 ± 0.94	.027	0.04 (0.01 to 0.07)
Calculation		3.60 ± 0.91	3.57 ± 0.91	.103	0.03 (0.00 to 0.06)
Collective responsibility		3.28 ± 0.62	3.28 ± 0.60	.994	0.00 (−0.03 to 0.03)
Total score		45.88 ± 5.90	46.25 ± 5.74	<.001	−0.06 (−0.09 to −0.03)
5C model of mAbs
Confidence		3.35 ± 0.79	3.36 ± 0.80	.420	−0.02 (−0.04 to 0.01)
Complacency		2.85 ± 0.81	2.83 ± 0.88	.305	0.02 (−0.01 to 0.05)
Constraints		2.73 ± 0.90	2.72 ± 0.95	.437	0.01 (−0.02 to 0.04)
Calculation		3.62 ± 0.93	3.60 ± 0.95	.195	0.02 (−0.01 to 0.05)
Collective responsibility		3.27 ± 0.61	3.25 ± 0.60	.103	0.03 (0.00 to 0.06)
Total score		46.25 ± 5.47	46.39 ± 5.69	.195	−0.02 (−0.05 to 0.01)

Notes: SD, standard deviation; CI, confidence interval; maternal vaccines, maternal RSV vaccines; mAbs, infant monoclonal antibodies.

Partners exhibited similar patterns. For maternal vaccines, the mean 5C total score was 46.25 (SD = 5.74), and for mAbs it was 46.39 (SD = 5.69). The calculation domain remained highest (maternal vaccines: 3.57 ± 0.91; mAbs: 3.60 ± 0.95), the constraints (maternal vaccines: 2.71 ± 0.94; mAbs: 2.72 ± 0.95) and complacency (maternal vaccines: 2.85 ± 0.87; mAbs: 2.83 ± 0.88) domains remained lowest (Table 2).

When comparing individuals within dyads, women had marginally lower total 5C scores for maternal RSV vaccines than their partners (p < .001; Cohen’s d = −0.06, 95% CI: −0.09 to −0.03), indicating slightly higher overall hesitancy. Consistent with this, women also scored marginally higher on complacency (p = .007; Cohen’s d = 0.05, 95% CI: 0.02 to 0.08) and constraints (p = .027; Cohen’s d = 0.04, 95% CI: 0.01 to 0.07), while no significant gender difference was observed in other domains. In contrast, no gender difference was observed in total scores for the mAbs strategy (p = .195). No domain differences by gender were significant (Table 2).

Correlation of vaccine hesitancy within individuals and couples

Strong within-individual correlations were observed between hesitancy toward maternal RSV vaccines and infant mAbs across all 5C domains and total scores (Pearson’s r ≥ 0.70), except for the correlation (r) between partners’ collective responsibility scores (r = 0.69) (Figure 2). Within-couple correlations for the same strategy were generally low (r ranged from 0.31 to 0.48). The highest couple-level correlation was observed for the constraints domain (maternal vaccines: r = 0.45; mAbs: r = 0.48), and the lowest for confidence (maternal vaccines: r = 0.31; mAbs: r = 0.36) (Figures 2B, D). Notably, cross-strategy correlations within couples (e.g., one’s maternal RSV hesitancy vs. partner’s mAbs hesitancy) were also low (range: r = 0.29–0.47).

Figure 2.

Correlation heatmaps of 5C model (women vs. partners).

Notes: All these correlations were statistically significant (p < .01). Circle size indicated correlation strength. Women Maternal: pregnant/postpartum women’s 5C total and domain scores for maternal RSV vaccines; Partner Maternal: partners’ 5C total and domain scores for maternal RSV vaccines; Women mAbs: pregnant/postpartum women’s 5C total and domain scores for infant mAbs; Partner mAbs: partners’ 5C total and domain scores for infant mAbs.

Factors associated with vaccine hesitancy

Multivariable LMMs were employed to identify factors associated with hesitancy toward both maternal RSV vaccines and infant mAbs. These models incorporated a random intercept to account for clustering within couples and adjusted for all sociodemographic and perinatal variables. The calculated ICCs indicated substantial within-couple clustering, with values of 0.36 for the maternal vaccines model and 0.35 for the infant mAbs model. Full results were presented in Table 3.

Table 3.

Associated factors of vaccine hesitancy toward RSV immunization strategies.

Characteristics		Maternal vaccines		mAbs
		β (95% CI)	Adjusted p-values	β (95% CI)	Adjusted p-values
Sociodemographic variables (n = 8844)
Age, mean (SD)		−0.09 (−0.12 to −0.05)	<.001	−0.09 (−0.12 to −0.06)	<.001
Gender
Men		ref		ref
Women		0.09 (−0.40 to 0.57)	.781	0.53 (0.06 to 1.00)	.050
Residence
Urban		ref		ref
Rural		−0.44 (−0.75 to −0.13)	.011	−0.58 (−0.87 to −0.28)	<.001
Education level
Below college degree		ref		ref
College degree or higher		0.20 (−0.08 to 0.48)	.265	0.14 (−0.13 to 0.41)	.462
Occupation
Medical worker		ref		ref
Civil servant/Enterprise employee		−0.02 (−0.53 to 0.50)	.991	0.15 (−0.34 to 0.65)	.621
Self-employed/Freelancer		−0.35 (−0.87 to 0.17)	.276	−0.20 (−0.70 to 0.30)	.569
Others		−0.39 (−1.05 to 0.28)	.331	−0.25 (−0.88 to 0.39)	.569
Unemployed		−0.26 (−0.88 to 0.35)	.478	−0.40 (−0.99 to 0.20)	.299
Any healthcare professional in the family
Yes		ref		ref
No		1.12 (0.85 to 1.40)	<.001	1.07 (0.81 to 1.33)	<.001
Annual household income per capita (CNY)
≤50,000		ref		ref
50,000–100,000a		0.69 (0.37 to 1.00)	<.001	0.87 (0.56 to 1.18)	<.001
>100,000		0.22 (−0.07 to 0.51)	.225	0.63 (0.35 to 0.91)	<.001
Self-rated health
Very good		ref		ref
Good		0.55 (0.23 to 0.86)	.002	0.42 (0.12 to 0.72)	.015
Fair		−0.32 (−0.65 to 0.01)	.104	−0.33 (−0.65 to −0.02)	.067
Not good		−0.93 (−1.45 to −0.41)	.002	−0.87 (−1.37 to −0.38)	.002
Perinatal variables (n = 4422)
Perinatal stage
First trimester		ref		ref
Second trimester		−0.00 (−0.56 to 0.56)	.998	−0.19 (−0.73 to 0.35)	.589
Third trimester		−0.71 (−1.18 to −0.25)	.006	−0.68 (−1.13 to −0.24)	.006
Postpartum		0.23 (−0.18 to 0.65)	.338	−0.02 (−0.42 to 0.38)	.911
Routine prenatal care completion
Completely finish		ref		ref
Incompletely finish		−0.43 (−0.81 to −0.05)	.059	−0.43 (−0.80 to −0.07)	.043
No check		0.29 (−0.47 to 1.04)	.519	−0.13 (−0.85 to 0.59)	.757
No. of childrenb
0		ref		ref
1		−0.93 (−1.31 to −0.55)	<.001	−0.79 (−1.15 to −0.42)	<.001
≥2		−1.55 (−2.11 to −0.99)	<.001	−1.39 (−1.93 to −0.85)	<.001
Any pregnancy-related complicationsc
No		ref		ref
Yes		−0.34 (−0.71 to 0.02)	.114	−0.25 (−0.60 to 0.10)	.261
Influenza vaccination during pregnancyd
No		ref		ref
Yes		0.73 (0.32 to 1.13)	.002	0.07 (−0.32 to 0.46)	.757
COVID-19 vaccination during pregnancyd
No		ref		ref
Yes		0.27 (−0.14 to 0.69)	.276	0.15 (−0.25 to 0.55)	.569

Notes. Maternal vaccines, maternal RSV vaccines; mAbs, infant monoclonal antibodies; CI, confidence interval.

^aThe range of 50,000–100,000 does not include 50,000.

^bNo. of children refers to the number of children the participant had prior to the current pregnancy.

^cAny pregnancy-related complications include gestational hypertension, gestational diabetes mellitus, thyroid disorders, and other complications.

^dInfluenza and COVID‑19 vaccination during pregnancy refer to receipt of the respective vaccine during the index pregnancy or during any prior pregnancy.

For socio-demographic variables, older age was associated with significantly greater vaccine hesitancy for both immunization strategies. Specifically, each additional year of age was associated with a decrease in total 5C score (maternal vaccines: β = −0.09, 95% CI: −0.12 to −0.05; mAbs: β = −0.09, 95% CI: −0.12 to −0.06). Compared to those living in urban areas, participants living in rural areas reported significantly greater hesitancy for both maternal vaccines (β = −0.44, 95% CI: −0.75 to −0.13) and mAbs (β = −0.58, 95% CI: −0.87 to −0.28). Income was negatively associated with hesitancy. Compared to the lowest income group (≤50,000 CNY), those in the middle-income group (50,000–100,000 CNY) showed lower hesitancy (maternal vaccines: β = 0.69, 95% CI: 0.37 to 1.00; mAbs: β = 0.87, 95% CI: 0.56 to 1.18), and those in the highest income group (>100,000 CNY) had lower hesitancy for mAbs only (β = 0.63, 95% CI: 0.35 to 0.91).

Self-rated health status also influenced vaccine attitudes. Compared to those with “very good” self-rated health, participants who reported “good” health demonstrated lower hesitancy (maternal vaccines: β = 0.55, 95% CI: 0.23 to 0.86; mAbs: β = 0.42, 95% CI: 0.12 to 0.72), while those rating their health as “not good” reported higher hesitancy (maternal vaccines: β = −0.93, 95% CI: −1.45 to −0.41; mAbs: β = −0.87, 95% CI: −1.37 to −0.38). Gender difference, educational level and non-medical occupations were not significantly associated with hesitancy in either model. Participants reporting no healthcare professional in the family had significantly lower hesitancy than those with a healthcare professional in the family (maternal vaccines: β = 1.12, 95% CI: 0.85 to 1.40; mAbs: β = 1.07, 95% CI: 0.81 to 1.33).

Among women participants, perinatal variables also contributed to hesitancy. Being in the third trimester was associated with significantly greater hesitancy for both maternal vaccines (β = −0.71, 95% CI: −1.18 to −0.25) and mAbs (β = −0.68, 95% CI: −1.13 to −0.24). Incomplete prenatal care was associated with higher hesitancy for mAbs (β = −0.43, 95% CI: −0.80 to −0.07) and showed a marginal association for maternal vaccines (β = −0.43, 95% CI: −0.81 to −0.05, p = .059), in comparison to those who completed all routine prenatal care. The number of children was associated with increased hesitancy. Women with one prior child had higher hesitancy than primiparas (maternal vaccines: β = −0.93, 95% CI: −1.31 to −0.55; mAbs: β = −0.79, 95% CI: −1.15 to −0.42), and those with two or more prior children showed even stronger associations (maternal vaccines: β = −1.55, 95% CI: −2.11 to −0.99; mAbs: β = −1.39, 95% CI: −1.93 to −0.85). No significant associations were observed between pregnancy-related complications and vaccine hesitancy. Having a history of influenza vaccination during pregnancy was significantly associated with lower hesitancy toward maternal vaccines (β = 0.73, 95% CI: 0.32 to 1.13), but not toward mAbs. Meanwhile, no significant associations were observed between history of COVID-19 vaccination during pregnancy and vaccine hesitancy.

The results of the sensitivity analysis were presented in Table S5. After excluding samples with missing observations, the analysis results were basically consistent with the aforementioned analysis, except that incomplete prenatal care also became significantly associated with higher hesitancy toward maternal vaccines (β = −0.50, 95% CI: −0.90 to −0.10).

Discussion

The present study is, to our knowledge, the largest dyadic survey design investigation of hesitancy toward maternal RSV vaccines and infant mAbs conducted in any setting, and the first to apply the 5C psychological framework to both RSV immunization strategies simultaneously among pregnant and postpartum women and their cohabiting partners. We found that the dominant psychological profile of hesitancy across both strategies and both members of the couple was characterized by higher calculation scores rather than by low confidence, complacency, or perceived constraints. Within-couple concordance on hesitancy was uniformly low across all 5C domains. Older age, rural residence, lower household income, third-trimester status and prior children were each independently associated with greater hesitancy, while prior influenza vaccination during pregnancy was specifically associated with lower hesitancy toward maternal vaccines but not toward infant mAbs. Sensitivity analyses showed a broadly similar pattern, supporting the robustness of the main findings, except that women with incomplete prenatal care exhibited higher levels of vaccine hesitancy toward both strategies.

The predominance of calculation as the highest-scoring domain carries important implications for how public health authorities should frame RSV immunization strategies. In the 5C framework, calculation reflects active information seeking and deliberation about the benefits and risks of vaccination, and therefore does not necessarily indicate opposition. However, it may also signal indecision, uncertainty, information overload, or greater susceptibility to conflicting information.^12,28 Interpretation of this domain should therefore be contextual. In China, only the long-acting mAb (e.g., Nirsevimab) is currently available and maternal vaccines remain unlicensed.⁹ Public health communication should therefore prioritize authoritative information on safety, efficacy, and post-marketing performance.^8,29–31 Communication strategies that emphasize authoritative data, address specific safety concerns about immunization during pregnancy or in neonates, and acknowledge residual uncertainty are more likely to convert cautious deliberation into informed acceptance than campaigns that simply promote vaccination in general terms.^32,33 This approach may be more effective than generic promotional messaging, consistent with evidence from European studies of maternal pertussis and influenza vaccination.³⁴

Women reported statistically significantly lower total 5C scores than their partners for maternal vaccines, but this difference was trivially small (0.37 points on a 75-point scale; Cohen’s d = −0.06, 95% CI: −0.09 to −0.03) and likely reflected the inflated statistical power of a large sample rather than a meaningful attitudinal gap. Women also scored only slightly higher on complacency and constraints, with similarly trivial effect sizes. For infant mAbs, no significant difference between women and partners was detected. The absence of any meaningful gender gap for mAbs is consistent with the hypothesis that when the immunization target shifts from the pregnant woman herself to the newborn infant, both parents adopt comparable evaluative stances. Previous qualitative research has suggested that pregnant women experience heightened concern about fetal safety when asked to accept a vaccine during pregnancy,^35,36 and this concern may account for the marginally less favorable attitudes observed here, but the effect is too small to support confident causal inference. Taken together, these results suggest that RSV immunization communication strategies should engage both parents rather than assume substantial gendered differences in hesitancy, though qualitative research may still reveal gender-specific concerns that are masked by the composite 5C score.

Within-couple Pearson correlations for the same strategy ranged from 0.31 to 0.48 across the ten domain-strategy combinations, and the couple-level ICCs from the LMMs were consistent with this range. These values indicate that a substantial proportion of between-individual variation in hesitancy exists within rather than between couples. The constraints domain showed the strongest within-couple agreement, which is expected given that structural barriers such as geographical access to healthcare and financial resources are shared material circumstances.^37,38 Confidence, by contrast, showed the weakest concordance, suggesting that trust in vaccine safety, the healthcare system, and public health authorities is shaped more by individual experience than by shared household context.³⁰ It suggests that, although immunization decisions may be shaped within a shared spousal context, partners are not necessarily well aligned in the psychological antecedents captured by the 5C domains. This interpretation remains speculative, however, as we did not measure dyadic communication directly. Regardless of the mechanism, the low concordance indicates that prenatal immunization counseling that targets only the pregnant woman risks leaving a skeptical or uninformed partner whose opposition may subsequently impede uptake.^31,39 Both parents should be included in educational interventions.

Older age, rural residence, and lower income were each associated with higher hesitancy. Previous studies also noted a positive correlation between younger age and higher vaccination willingness, which may be primarily due to greater concern among older individuals about uncertainty or unknown future side effects.⁴⁰ Participants with lower income and rural residence, who generally represent groups with lower socioeconomic status, may have limited access to health information and services.³⁷ We found that having a healthcare professional in the household was associated with greater hesitancy toward both strategies, despite the woman’s or her partner’s own occupation as a healthcare professional being non-significant. This finding is inconsistent with much of the existing literature and should therefore be interpreted cautiously. One possible explanation is that, given that RSV preventive products are not yet routinely available in mainland China, healthcare professionals may adopt a more cautious stance toward these products, and their reserved attitude may shape household attitudes.^34,41,42 At the same time, we strongly acknowledge the possibility of residual confounding. Variables such as the family member’s specific medical specialty, infectious disease-related experience, or the household’s baseline health anxiety were not measured. We also cannot exclude the possibility that this was a spurious association. This specific finding therefore warrants further targeted investigation. Individuals reporting poor self-rated health also exhibited higher hesitancy, possibly reflecting heightened concern about immunization-related risks.⁴³

Among pregnancy and postpartum covariates, we found that women in the third trimester, those with prior children, and those with incomplete prenatal care exhibited higher levels of vaccine hesitancy. Third-trimester women may have lower risk tolerance regarding maternal vaccine safety.^44,45 Among women, incomplete prenatal care was associated particularly with higher hesitancy toward infant mAbs, while a similar but less certain pattern was observed for maternal RSV vaccines in the primary analysis. This may reflect lower perception of RSV-related risk among these women, as standardized prenatal care is a key channel through which the healthcare system conveys vaccine confidence and educates women about the health risks of RSV infection.⁴⁶ Multiparity-associated hesitancy may reflect reliance on prior immunization experience, differing perceptions of the health risks of infant RSV, or a more cautious attitude toward novel immunization products, similar findings have been reported in previous studies.⁴⁷ In our study, prior influenza vaccination during pregnancy was associated with lower maternal hesitancy toward RSV vaccines, but no such association was observed for COVID-19 vaccination. This divergence was consistent with findings from France showing that knowledge of vaccine-preventable diseases predicted willingness for both influenza and RSV vaccination but not for COVID-19 vaccination.³⁴ Both influenza and RSV maternal vaccines share the common protective goal of benefiting infants, and thus may elicit similar hesitancy profiles.⁴⁸ In contrast, COVID-19 vaccination decisions appear to be driven by distinct factors, including the influence of public discourse, misinformation, and pandemic-related anxiety.^49,50

This study possesses several strengths. First, this is the first study in China to investigate vaccine hesitancy toward RSV immunization strategies from a dyadic design perspective. Second, we employed validated and well-established scales to assess vaccine hesitancy. Third, our sample encompassed pregnant/postpartum women covering all trimesters of pregnancy, providing a panoramic view of how hesitancy varies across the perinatal journey. Fourth, our study covered regions across China with diverse geographic and socioeconomic profiles. Nevertheless, several limitations should be acknowledged. First, convenience sampling may have introduced selection bias, with highly educated and urban populations potentially overrepresented. This may limit the generalizability of the findings to rural residents, people with lower educational attainment, and other medically underserved populations, and may mean that the observed hesitancy profile underestimates barriers in these groups. Second, all data were self-reported and may be subject to social desirability bias. Additionally, the cross-sectional design captures vaccine hesitancy only at a single point in time. Future longitudinal studies are warranted to examine the dynamic evolution of vaccine hesitancy as parental risk perceptions may change over the course of pregnancy. Third, because recruitment was based on convenience sampling in busy clinical settings, we were unable to systematically record all individuals approached, refusals, or characteristics of non-participants. Therefore, response and refusal rates could not be calculated, and selection bias cannot be ruled out. Fourth, infant clinical characteristics were not included in the analyses, as many participants completed the survey during pregnancy before such information was available. Future postnatal studies should assess whether infant health status affects parental hesitancy toward RSV preventive products. And the communication ability and cognitive/psychiatric status were assessed by clinical observation and self-report rather than standardized tools, which may affect the consistency of eligibility assessment and the accuracy of self-reported data.

Conclusion

In conclusion, pregnant and postpartum women and their partners in the surveyed regions of China exhibit a hesitancy profile toward RSV immunization strategies dominated by active information-seeking rather than by distrust or complacency. The primary barrier to uptake therefore appears to be an unmet need for accessible, evidence-based information about the safety and efficacy of maternal RSV vaccines and infant mAbs, rather than attitudinal opposition to immunization itself. This implies that public health interventions should prioritize transparent communication and proactive information provision by healthcare providers. Low within-couple concordance in attitudes underscores the importance of engaging both parents in prenatal immunization counseling and broader communication. Public health strategies should prioritize populations at elevated risk of hesitancy, notably older parents, rural residents, lower-income households, and women in late pregnancy, or with prior children. Longitudinal research is needed to track how hesitancy evolves across pregnancy and the postpartum period and to evaluate whether targeted communication interventions can effectively convert rational caution into informed acceptance.

Biography

Peige Song is a Research Fellow of the School of Public Health, Zhejiang University School of Medicine, where she also serves as Deputy Director of the Global Health Center. Dr Song focuses on global health metrics and health resource allocation, with major contributions in disease burden estimation and optimizing health resources, which also extends to maternal and child health, specifically in the optimization of immunization strategies for vulnerable populations.

Funding Statement

The research was supported by a grant from Merck MISP [102659].

Disclosure statement

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

Data availability

The data that support the findings of this study are available from the corresponding author, Peige Song (peigesong@zju.edu.cn), upon reasonable request, but restrictions apply to the availability of these data due to the nature of the study.

Institutional review board statement

This study was conducted following the guidelines of the Declaration of Helsinki. Ethics approval was obtained from the Ethics Committee of Zhejiang University School of Public Health (reference number: ZGL202410–1).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2026.2716519