Preventive effectiveness of respiratory syncytial virus vaccination in older adults: a multi-institutional cohort study

Abstract

Background

This multi-institutional study evaluated the real-world effectiveness of respiratory syncytial virus (RSV) vaccination in preventing RSV infections among adults aged 50 and older.

Methods

Using data from the TriNetX Research Network, we included individuals aged ≥ 50 years between January 1, 2023, and May 31, 2026. Patients were categorized into vaccinated and unvaccinated groups, defined as those receiving any RSV vaccine and any non-RSV vaccine, respectively. After 1:1 propensity score matching, we estimated hazard ratios (HR) for the composite outcome of RSV infection using Cox proportional-hazards models.

Results

The study included 179,871 patients in each group after matching. The vaccinated group had a lower risk of RSV infection compared to the unvaccinated group (incidence rate: 0.17 vs. 0.27 per 100 person-years, HR, 0.65; 95% CI, 0.58–0.73). Additionally, the vaccinated group showed lower risks of RSV-associated lower respiratory tract disease (incidence rate: 0.12 vs. 0.19 per 100 person-years, HR, 0.69; 95% CI, 0.60–0.79). Risk reductions were consistent across subgroups defined by age, sex, and comorbidity.

Conclusions

This large real-world study demonstrates the substantial effectiveness of RSV vaccination in preventing infections among older adults, a high-risk population.

Clinical trial number

Not applicable.

Introduction

Respiratory syncytial virus (RSV) is increasingly recognized as a significant cause of lower respiratory tract disease in older adults, often resulting in severe illness and substantial healthcare costs [1, 2]. It is also associated with notable morbidity due to exacerbations of underlying cardiopulmonary conditions [3, 4]. Recent meta-analysis including 22 studies suggest that in 2019, RSV caused over 5 million cases of acute respiratory infections, 470,000 hospitalizations, and 33,000 in-hospital deaths among adults aged 60 years and older in high-income countries [5]. Within the United States specifically, the annual burden in adults aged 65 and above is estimated to range from 600,000 to 1 million medically-attended visits, 140,000 to 177,000 hospitalizations, and approximately 11,000 to 14,000 deaths [6–8]. Furthermore, older adults hospitalized with RSV-associated lower respiratory tract disease appear to fare worse than those with SARS-CoV-2 or influenza, being 2–3 times more likely to require supplemental oxygen, 1.5 times more likely to be admitted to the intensive care unit, and having a 2-fold higher odds of mechanical ventilation or death compared to those with influenza [9, 10]. However, there are few evidence-based RSV-specific treatments currently available, with supportive care being the main modality [2].

Given the substantial burden of RSV in older adults, effective prevention strategies are essential to reduce the impact of this disease in this vulnerable population. Developing and implementing RSV vaccines specifically for older adults could play a crucial role in mitigating the significant morbidity and mortality associated with this respiratory pathogen. Based on the findings of several randomized clinical trials,^11–13 demonstrating the efficacy and safety of RSV vaccine in older individuals, two protein subunit vaccines for preventing severe RSV in adults aged 60 and older were licensed in 2023 [14]. The first, a recombinant RSV prefusion F protein-based vaccine (RSVpreF, bivalent and adjuvanted with aluminum hydroxide), was approved in the United States in January 2023 after a phase 3 study showed it reduced the risk of severe RSV-associated lower respiratory tract disease by 82.6% compared to placebo [11]. The second, a purified, bivalent RSV F protein subunit vaccine formulated with the AS01E adjuvant, received regulatory approval by the FDA in May 2023 following positive results from a phase 3 trial that reported a significant reduction in RSV-associated lower respiratory tract illness in vaccinated older adults [12]. Most recently, in September 2023, the FDA granted approval for a novel mRNA-based RSV vaccine encoding the prefusion F protein for use in individuals aged 60 and above in the United States [13, 14]. While these clinical trial results [11–13] are promising, the real-world evidence is still needed to fully assess the effectiveness and uptake of these new RSV vaccines in the older adult population [15]. However, real-world studies assessing the effectiveness of RSV vaccines in older individuals are limited [16, 17]. Therefore, this multi-institutional study was conducted to investigate the effectiveness of RSV vaccines in the prevention of RSV infection in older adults.

Methods

Data source

This retrospective cohort study queried clinical information from the TriNetX Global Collaborative Network (TriNetX Inc, Cambridge, MA, USA), a federated research platform containing real-time electronic health records of more than 150 healthcare organization (HCOs), the vast majority of which are U.S.-based academic medical centers, community hospitals, and integrated health systems [18]. The network aggregates anonymized medical records, including demographics, diagnoses (ICD-10-CM), medications, laboratory results, vital signs, clinical encounters, and genomic information, from participating healthcare organizations (HCOs) [19]. These records are standardized and maintained in compliance with the Health Insurance Portability and Accountability Act requirements for limited datasets.

Patient selection

Patients aged ≥ 50 years who had at least one visit to a HCO and received a general medical examination between January 1, 2023, and May 31, 2026, were included in the study. Participants were categorized into vaccinated and unvaccinated groups based on their RSV vaccination status. The vaccinated group was identified by CPT codes for 90,679 (recombinant prefusion F protein vaccine with adjuvant), 90,678 (bivalent subunit prefusion F protein vaccine), and 90,683 (mRNA lipid nanoparticle vaccine encoding prefusion F protein) as well as the corresponding RxNorm code (2636589) for the prefusion F protein vaccine. In contrast, the unvaccinated group comprised patients who received any vaccine other than an RSV vaccine. The index date was defined as the date of RSV vaccination for the vaccinated group and the date of receipt of a non-RSV vaccine for the unvaccinated group. This active-comparator design was employed to reduce healthy vaccinee bias, as individuals who receive vaccinations are more likely to share similar healthcare utilization patterns and health-seeking behaviors.

Outcomes

The primary outcome was any RSV infection identified by ICD-10-CM codes for RSV as the cause of diseases classified elsewhere (B97.4), RSV pneumonia (J12.1), acute bronchitis due to RSV (J20.5) and acute bronchiolitis due to RSV (J21.0) over a 1-year follow-up. Secondary outcomes included RSV-associated lower respiratory tract disease (RSV-LRTD), defined as the composite of RSV pneumonia, acute bronchitis due to RSV, and acute bronchiolitis due to RSV, analyzed as a single bundled outcome.

Statistical analysis

Baseline characteristics were presented as numbers (proportions), except for age which was presented as means (standard deviations). Propensity score matching (PSM) was performed using the built-in 1:1 greedy nearest-neighbor algorithm of the TriNetX platform. Comorbidities were selected a priori as matching variables based on their established associations with both RSV vaccination likelihood and RSV infection risk, rather than on observed baseline imbalance alone, consistent with Advisory Committee on Immunization Practices guidance identifying older adults with chronic cardiopulmonary disease, diabetes mellitus, obesity, chronic kidney disease, and immunocompromising conditions as priority populations for vaccination and as recognized risk factors for severe RSV infection. Matching variables included age, sex, race, and comorbidities (hypertension (ICD-10-CM: I10-I1A), hyperlipidemia (E78.5), neoplasms (C00-D49), overweight and obesity (E66), diabetes mellitus (DM) (E08-E13), chronic ischemic heart disease (I25), obstructive sleep apnea (G47.33), chronic obstructive pulmonary disease (J44), heart failure (I50), and nicotine dependence (F17) [20, 21]. Covariate balance was regarded as adequate when standardized mean differences (SMDs) were less than 0.1.

Hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) were estimated using Cox proportional-hazards models. Statistical analyses were conducted with the built-in statistical module on the TriNetX platform. Kaplan-Meier curves and forest plots were generated with the R ggplot package. Cumulative probabilities were estimated with Kaplan-Meier analyses, and p-values less than 0.05 were considered statistically significant.

Prespecified subgroup analyses were performed by age, sex, and comorbidity status, including cardiovascular disease (ICD-10-CM: I00–I99), chronic lower respiratory disease (ICD-10-CM: J40–J47), chronic kidney disease (CKD) (ICD-10-CM: N18), chronic liver disease (ICD-10-CM: K70–K77), type 2 diabetes mellitus (ICD-10-CM: E11), and obesity (ICD-10-CM: E66). For each subgroup, a new propensity score-matched cohort was constructed within the respective stratum to ensure adequate covariate balance prior to outcome analysis.

Results

Patient selection

On June 16, 2026, we identified 2,941,567 individuals aged ≥ 50 years who received immunization across 177 HCOs. Among these individuals, 179,906 received an RSV vaccine, whereas 2,658,747 received a non-RSV vaccine. After 1:1 propensity score matching (PSM) based on age, sex, race, and comorbidities, 179,871 patients remained in each cohort, resulting in well-matched vaccinated and comparator groups (Fig. 1).

Fig. 1.

The algorithm of patient selection

Baseline characteristics of included patients

Before PSM, the vaccinated group was older than the unvaccinated group (mean age, 72.0 ± 7.6 years vs. 65.4 ± 9.7 years). Although the sex distribution was similar between the groups, White individuals were more prevalent in the vaccinated group. In addition, the vaccinated group had a higher prevalence of hypertension, hyperlipidemia, neoplasms, diabetes mellitus, chronic ischemic heart disease, obstructive sleep apnea, chronic kidney disease, chronic lower respiratory diseases, and heart failure than the comparator group (all SMD > 0.1) (Table 1). After PSM, the vaccinated and unvaccinated groups were well balanced with respect to age, sex, race, and comorbidities, with all SMDs < 0.1 (Table 1).

Table 1.

Demographic information of patients before and after propensity score matching

Overall patient, n (%)	Before matching					After matching
	Vaccinated		Unvaccinated		SMDs	Vaccinated		Unvaccinated		SMDs
	N = 179,871		N = 2,708,833			N = 179,871		N = 179,871
Age at index
Mean ± SD	72.0 ± 7.58		65.4 ± 9.7		0.83	72.0 ± 7.58		72.0 ± 7.81		0
Sex, n (%)
Male	76263	-42.4	1206426	-44.5	0.04	76263	-42.4	76371	-42.5		0
Female	103595	-57.6	1499372	-55.4	0.08	103595	-57.6	103426	-57.5		0
Race, n (%)
White	150481	-83.7	1937919	-71.5	0.29	151908	-42.4	76371	-42.5		0
Black or African American	17,788	-9.9	3,35,836	-12.4	0.08	17,788	-9.9	17,430	-9.70%		0
Asian	5032	-2.8	128124	-4.7	0.1	5032	-2.8	4909	-2.70%		0
Comorbidities, n (%)
Hypertensive diseases	134318	-74.7	1572909	-58.1	0.36	134318	-74.7	134985	-75		0.01
Hyperlipidemia	101605	-56.5	1063221	-39.3	0.35	101605	-56.5	102383	-56.9		0.01
Neoplasms	82296	-45.8	862806	-31.9	0.29	82296	-45.8	81854	-45.5		0
Diabetes mellitus	58899	-32.7	691678	-25.5	0.16	58899	-32.7	58527	-32.5		0
Overweight and obesity	61248	-34.1	705488	-26	0.18	61248	-34.1	60787	-33.8		0.01
Chronic ischemic heart disease	46427	-25.8	406813	-15	0.27	46427	-25.8	45896	-25.5		0.01
Obstructive sleep apnea	39547	-22	359077	-13.3	0.23	39547	-22	37513	-20.9		0.03
Chronic kidney disease	37550	-20.9	328820	-12.1	0.24	37550	-20.9	36209	-20.1		0.02
COPD	28643	-15.9	219420	-8.1	0.24	28643	-15.9	27247	-15.1		0.02
Heart failure	24869	-13.8	210448	-7.8	0.2	24869	-13.8	22890	-12.7		0.03
Nicotine dependence	22284	-12.4	298208	-11	0.04	22284	-12.4	21518	-12		0.01

COPD, chronic obstructive pulmonary disease; SMD, standardized mean difference

Primary outcomes

During the one-year follow-up period, 458 patients in the vaccinated group and 834 patients in the unvaccinated group developed RSV infections. Compared to the unvaccinated group, the vaccinated group had a lower risk of RSV infection (incidence rate: 0.17 vs. 0.27 per 100 person-years, HR, 0.65; 95% CI, 0.58–0.73; p < 0.0001) (Table 2). Kaplan-Meier survival analysis confirmed these findings, showing a lower accumulated incidence of RSV infections in the vaccinated group during the follow-up period (log-rank p < 0.0001, Fig. 2).

Table 2.

The hazard ratio for comparing vaccinated with unvaccinated cohort for the risk of RSV infection

Outcomes	Vaccinated (n = 179,871)		Unvaccinated (n = 179,871)		Hazard ratio (95% CI)	p-value
	Events (n)	Incidence rate Per 100 person-years	Events (n)	Incidence rate Per 100 person-years
Primary outcome
RSV infection	458	0.17	834	0.27	0.65 (0.58–0.73)	< 0.0001
Secondary outcome
RSV-LRTD	329	0.12	569	0.19	069 (0.60–0.79)	< 0.0001

RSV, respiratory syncytial virus; RSV-LRTD, RSV-associated lower respiratory tract disease

Fig. 2.

Kaplan-Meier time-to-event free curve for the vaccinated group and the unvaccinated group

Figure 3 displays subgroup analyses stratified by age, sex, and underlying conditions or comorbidities. The vaccinated group consistently showed lower RSV infection risks across all subgroups. Significant risk reductions were observed in males (HR, 0.71; 95% CI, 0.59–0.85), females (HR, 0.57; 95% CI, 0.49–0.67), age groups of 60–69 years (HR, 0.68; 95% CI, 0.54–0.85), 70–79 years (HR, 0.52; 95% CI, 0.44–0.63), and ≥ 80 years (HR, 0.58; 95% CI, 0.48–0.71). Significant reductions were also noted in individuals with CKD (HR, 0.64; 95% CI, 0.52–0.79), chronic lower respiratory diseases (HR, 0.61; 95% CI, 0.53–0.71), chronic liver disease (HR, 0.53; 95% CI, 0.42–0.66), cardiovascular diseases (HR, 0.61; 95% CI, 0.54–0.69), type 2 DM (HR, 0.53; 95% CI, 0.44–0.64) and obesity (HR, 0.62; 95% CI, 0.53–0.73).

Fig. 3.

Forest plots of respiratory syncytial virus (RSV) infection between the vaccinated and unvaccinated group according to the subgroup

Secondary outcomes

Compared with the unvaccinated group, the vaccinated group had a lower risk of RSV-related lower respiratory tract disease (RSV-LRTD), with incidence rates of 0.12 versus 0.19 per 100 person-years (hazard ratio [HR], 0.69; 95% CI, 0.60–0.79). Consistent findings (HR < 1) were observed across subgroups stratified by sex, age, and comorbidities (Fig. 4). Significant risk reductions were observed in both males (HR, 0.79; 95% CI, 0.64–0.99) and females (HR, 0.59; 95% CI, 0.51–0.82), as well as in age groups 60–69 years (HR, 0.68; 95% CI, 0.52–0.89), 70–79 years (HR, 0.52; 95% CI, 0.42–0.64), and ≥ 80 years (HR, 0.64; 95% CI, 0.51–0.82). Among individuals with comorbidities, RSV vaccination was associated with significantly lower risks in those with CKD (HR, 0.69; 95% CI, 0.53–0.88), chronic lower respiratory diseases (HR, 0.68; 95% CI, 0.57–0.81), chronic liver disease (HR, 0.62; 95% CI, 0.48–0.82), cardiovascular diseases (HR, 0.67; 95% CI, 0.58–0.78), type 2 DM (HR, 0.58; 95% CI, 0.47–0.73), and obesity (HR, 0.64; 95% CI, 0.53–0.78).

Fig. 4.

Forest plot of the risk of RSV associated lower respiratory tract disease (RSV-LRTD) between the vaccinated and unvaccinated group

Discussion

This real-world study assessed the effectiveness of RSV vaccination and demonstrated that RSV vaccination is associated with a significantly lower risk of RSV infections compared to the unvaccinated group. After PSM for age, sex, race, and comorbidities, the vaccinated group consistently showed a 35% reduction in the risk of RSV infection. This protective effect remained significant across various subgroups, including males, females, patients aged 60 years and older, and those with chronic conditions such as CKD, chronic liver disease, cardiovascular diseases, chronic lower respiratory diseases, DM and obesity. Furthermore, vaccination was associated with a 31% reduction in RSV-LRTD, with all subgroup analyses supporting these findings. Overall, this large real-world study highlights the substantial effectiveness of RSV vaccination in preventing infections among older adults, a high-risk population. The results support recent regulatory approvals and emphasize the importance of implementing vaccination strategies to reduce the significant RSV burden in this vulnerable group.

The finding of a significant reduction in RSV infection risk among vaccinated older adults is highly significant and aligns with the promising efficacy results seen in clinical trials [11–13]. The AReSVi-006 trial reported that the efficacy of RSVpreF3 against RSV-associated lower respiratory tract disease and RSV-associated acute respiratory disease was 82.6% and 71.7%, respectively, during a median follow-up of 6.7 months [11]. The RENOIR trials showed that the RSVpreF vaccine exhibited efficacy of 66.7%, 85.7%, and 62.1% against RSV-associated lower respiratory tract disease with at least two signs or symptoms, with at least three signs or symptoms, and RSV-associated acute respiratory disease, respectively [12]. The ConquerRSV trial demonstrated that the efficacy of mRNA-1345 was 83.7%, 82.4%, and 68.4% against RSV-associated lower respiratory tract disease with at least two signs or symptoms, with at least three signs or symptoms, and RSV-associated acute respiratory disease, respectivel [13]. However, in contrast to these promising RCT which involved highly selected participants, our study present real-world evidence provide insights into treatment effectiveness in broader, more diverse patient populations. While RCTs offer controlled, internal validity, real-world study enhances generalizability and external validity, making both types of studies complementary in evaluating healthcare.

The study’s key strengths include its large sample size, with 179,871 patients in each the vaccinated and unvaccinated groups after PSM. This helped balance baseline characteristics between the two groups, reducing potential confounding factors. The use of PSM to balance baseline characteristics between the vaccinated and unvaccinated groups is a methodological strength that enhances the validity of the study’s findings. Another important strength is the use of a comprehensive healthcare database, providing a real-world representation of the older adult population. This enhances the generalizability of the results compared to narrowly-defined clinical trial cohorts. The consistent risk reduction observed across subgroups, including those with chronic conditions, suggests the vaccine’s broad effectiveness.

This study had several limitations. First, despite rigorous matching, there is potential for residual confounding. Unmeasured variables such as frailty status, functional status, or socioeconomic factors could potentially influence both the likelihood of vaccination and the risk of RSV infection. Additionally, despite the large matched cohort, the absolute number of RSV outcome events was relatively small. This limits the precision of the hazard ratio estimates, increases sensitivity to outcome misclassification, and reduces statistical power for subgroup analyses. Second, although this study lacked data on clinical outcomes beyond the occurrence of RSV infections, a test-negative design analysis based on an electronic health records network reported that RSV vaccination was effective in preventing RSV-associated hospitalizations and emergency department encounters among adults aged 60 years and older in the USA during the 2023-24 RSV season [16]. Nevertheless, a more comprehensive understanding of the vaccine’s public health benefits is needed, particularly regarding its impact on healthcare utilization, hospitalization rates, illness severity, and mortality. Third, because RSV vaccines were not approved until mid-2023, vaccinated individuals’ index dates were necessarily concentrated after this period, whereas unvaccinated index dates (defined by a non-RSV vaccination encounter) spanned the full study window. The TriNetX platform does not support direct matching on calendar index date between cohorts, so residual calendar-time confounding, such as secular trends in RSV circulation or testing practices, cannot be fully excluded. Fourth, the TriNetX platform captures vaccinations administered within participating HCOs but does not systematically incorporate vaccination records from external sources such as community pharmacies, retail chains, or non-participating providers. A substantial proportion of adult RSV vaccinations are administered in pharmacy settings in the United States. As a result, some individuals in the unvaccinated comparator group may in fact be vaccinated, representing non-differential exposure misclassification that would bias our vaccine effectiveness estimates toward the null. Our reported hazard ratios may therefore underestimate true vaccine effectiveness. Fifth, ICD-10-CM coding for RSV is known to underestimate true incidence, particularly in adults, because RSV testing is not routinely performed outside of severe presentations. This ascertainment bias may also be differential by vaccination status: vaccinated individuals who experience a breakthrough infection may undergo more extensive diagnostic evaluation, increasing the likelihood of an RSV-specific code being recorded, whereas similar mild infections among unvaccinated individuals without a healthcare encounter would go uncaptured. If such differential detection occurred, it would bias the observed event rate upward in the vaccinated group relative to the unvaccinated group, suggesting that the vaccine effectiveness estimated in this study may be conservative rather than overstated. Lastly, this study could not compare effectiveness across the three available RSV vaccine products. Although vaccination was identified using product-specific CPT codes (90679, 90678, 90683) as well as a non-specific RxNorm code (2636589), the majority of vaccinated individuals were captured only through the RxNorm code, which does not distinguish between products. This precluded a sufficiently powered stratified analysis by vaccine type. Future studies with more granular product-level data are needed to determine whether the observed associations differ across available RSV vaccine formulations.

In conclusion, this large-scale real-world study provides supportive evidence that RSV vaccination is associated with a lower risk of RSV infection among older adults, a population at high risk for severe disease. These findings should be interpreted with consideration of the relatively small number of absolute outcome events, potential outcome misclassification from ICD-10-based ascertainment, incomplete capture of vaccinations administered outside participating HCOs, and residual calendar-time confounding. As such, our results should be regarded as supportive rather than definitive evidence of vaccine effectiveness. Confirmation in studies specifically designed to address these methodological challenges, alongside ongoing real-world monitoring as the RSV vaccine landscape evolves, will be important for clarifying the magnitude of benefit and guiding implementation in this vulnerable population.

Author contributions

C.M.C. and C.C.L. conceptualized and designed the study. C.M.C., Y.W.T., and B.Z. contributed to data acquisition, statistical analysis, and interpretation of the data. C.M.C. drafted the initial manuscript. C.C.L. and C.K.T. provided critical revision of the manuscript for important intellectual content and supervised the study. All authors reviewed, edited, and approved the final manuscript.

Funding

This study received no specific funding from public, commercial, or not-for-profit funding agencies.

Data availability

The data underlying this study were accessed through the TriNetX Global Collaborative Network (https://trinetx.com). The federated architecture of the TriNetX platform does not permit direct patient-level data transfer or public release in accordance with participating institutions’ data governance agreements. Aggregate summary statistics and propensity score–matched cohort characteristics are presented in full in the manuscript and supplementary materials. Researchers interested in replicating this analysis may apply for independent access to the TriNetX network through their institution.

Declarations

Ethics approval and consent to participate

The Chi Mei Hospital institutional review board approved this TriNetX database study (approval number: 11402-E02) and waived the requirement for informed consent since the research used only aggregated statistical data from de-identified sources. The study was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.