Estimated Vaccine Effectiveness for Respiratory Syncytial Virus–Related Acute Respiratory Illness in Older Adults: Findings From the First Postlicensure Season

Abstract

Background

Respiratory syncytial virus (RSV) is an important cause of severe respiratory illness among older adults. Previously, we showed that the RSVpreF vaccine (Abrysvo; Pfizer) prevents RSV-related lower respiratory tract disease–related emergency department (ED) visits/hospitalizations in older adults. Here, we evaluate its effectiveness against additional acute respiratory illness (ARI) end points, including severe disease, among high-risk persons.

Methods

This was a retrospective test-negative case-control study of adults aged ≥60 years at Kaiser Permanente Southern California with ARI ED visits/hospitalizations, defined by International Classification of Diseases, Tenth Revision discharge code, from 24 November 2023 to 9 April 2024. Case patients tested positive for RSV. Controls in the primary analysis tested negative for RSV, human metapneumovirus, influenza, and severe acute respiratory syndrome coronavirus 2 and positive for a non–vaccine-preventable pathogen. The exposure was RSVpreF receipt ≥21 days before ARI diagnosis. Vaccine effectiveness (VE) was calculated from adjusted odds ratios via multivariable logistic regression.

Results

Overall, 8965 ARI ED visits/hospitalizations with RSV testing were included; 7.8% of patients were RSV positive, among whom 0.3% had received RSVpreF, compared with 3.6% of controls. The adjusted VE was 92% (95% confidence interval, 64%–98%). We estimated similar VE among patients with risk conditions (92% [95% confidence interval: 65%–98%]), the oldest subgroup (age ≥75 years; 95% [60%–99%]), those with critical outcomes (intensive care unit admission, mechanical ventilation, respiratory failure, vasopressor use, or death; 90% [16%–99%]), and those with severe disease (defined as ED visit or hospitalization requiring oxygen; 92% [35%–99%]).

Conclusions

Among older adults, RSVpreF demonstrated high VE against RSV-related ARI hospitalization or ED visits, including among high-risk subgroups, and against severe outcomes. RSV vaccination programs can protect groups at the highest risk of severe disease.

The RSVpreF vaccine was 92% effective against respiratory syncytial virus–related acute respiratory illness emergency department visits/hospitalizations among adults ≥60 years. Vaccine effectiveness was 90%–95% among those with risk conditions, age ≥75, and severe disease, and against critical outcomes.

Respiratory syncytial virus (RSV) is an important cause of severe acute respiratory illness (ARI) among older adults in the United States and globally. The estimated annual incidence rates of RSV-related hospitalizations and emergency department (ED) visits among adults ≥65 years are 282/100 000 and 200/100 000 [1] in the United States and 347/100 000 [2] and 729/100 000 in high-income countries [3], respectively. RSV manifestations can be severe [4, 5], requiring hospitalization or respiratory support. A 2023 study of prospectively enrolled hospitalized older adults reported that 80% of those testing RSV positive required standard-flow oxygen supplementation, and 23% required high-flow oxygen. Patients hospitalized with RSV had approximately 2- or 3-fold higher odds of requiring oxygen compared with patients with influenza or coronavirus disease 2019 (COVID-19), respectively [6]. The RENOIR prelicensure pivotal trial demonstrated the efficacy of the RSVpreF vaccine (Abrysvo; Pfizer) against symptomatic RSV-related lower respiratory tract infections (LRTIs) (89% efficacy against RSV-related LRTI with ≥3 symptoms at the end of season 1 and 77.8% at the end of season 2), but the trial was not powered to provide estimates for more severe outcomes (eg, hospitalization or oxygen use) [7].

RSV vaccines were first licensed for adults aged ≥60 years in 2023; the Advisory Committee on Immunization Practices (ACIP) initially recommended a single dose for this age group, using shared clinical decision making [8]. Studies from the first season after these recommendations demonstrated RSV vaccine effectiveness (VE) against RSV-related respiratory illnesses in real-world settings, including our prior analysis, which found a VE of 89% against ED visits/hospitalizations in older adults with lower respiratory tract disease (LRTD) [9]. LRTD is a subset of ARI, with ARIs including upper respiratory illnesses. The IVY study (RSVpreF/RSVpreF3 combined) reported VE of 75% (95% confidence interval [CI]: 50%–87%) against RSV-related ARI hospitalizations [10], and the VISION study (RSVpreF/RSVpreF3 combined) reported VE of 80% (71%–85%) against RSV-like hospitalizations (including upper respiratory illnesses/ARI) and 77% (70%–83%) against RSV-like ED visits [11].

In June 2024, ACIP recommendations were revised to a single dose for all adults aged ≥75 years and those aged 60–74 years at increased risk for severe RSV [10]. By December 2024, estimated US uptake of RSV vaccines was 34% among adults aged ≥75 years and 31% among those aged 60–74 years at higher risk for severe RSV [12]. In the present analysis, we evaluate the VE of RSVpreF against ARI end points, including severe disease requiring oxygen, by treatment setting, among those with high-risk chronic medical conditions, and among those aged ≥75 or ≥80 years.

METHODS

This was a retrospective test-negative case-control study of adults aged ≥60 years at Kaiser Permanente Southern California (KPSC) with an ARI ED visit or hospitalization from 24 November 2023 to 9 April 2024. KPSC is an integrated healthcare system whose patient population reflects the racial/ethnic and sociodemographic diversity of the underlying population [13, 14]. ARI was defined by International Classification of Diseases, Tenth Revision discharge code (Supplementary Table 1) in any position. We included ARI ED visits/hospitalizations with RSV-tested nasal/nasopharyngeal swab samples collected from 14 days before to 3 days after admission. ED encounters were defined as admission and discharge from the ED within <2 days. ED encounters of ≥2 days or with direct hospital transfer were defined as hospitalizations. Hospitalizations starting in inpatient settings required duration of ≥24 hours. Swab samples were tested using the Roche Diagnostics Cobas eplex RP2 multiplex respiratory pathogen panel, covering 16 viral and 2 bacterial targets. More than a single ARI encounter was included for a patient if the subsequent encounter occurred >30 days later (ie, >30 days from discharge from the first encounter to admission for the subsequent encounter). Multiple ARI ED visits/hospitalizations comprised a single encounter if occurring within ≤30 days of each other. Participants were required to have ≥1 year of health plan membership before the ARI (allowing for a 45-day membership gap to account for delays in membership renewal) to determine medical history. Patients receiving an RSV vaccine other than RSVpreF (eg, RSVpreF3) and those receiving RSVpreF <21 days before the ARI were excluded.

To reduce bias associated with infrequent standard-of-care (SOC) RSV testing, this study included RSV results obtained from both SOC testing and from specimens salvaged from ARIs for which testing was ordered for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) or influenza only. Salvage and SOC specimens were tested on the Cobas panel. Case patients were RSV positive. Separate analyses were conducted with 2 sets of prespecified controls. The primary analysis used “strict” controls, who were negative for RSV, human metapneumovirus (hMPV), and other vaccine-preventable diseases (VPDs) (influenza and SARS-CoV-2), and positive for ≥1 non-VPD pathogen (eg, adenovirus, coronavirus (229E, HKU1, NL63, OC43), human rhinovirus/enterovirus, parainfluenza 1–4, Chlamydia pneumoniae, and Mycoplasma pneumoniae). “Broad” controls (sensitivity analysis) were RSV negative. The strict approach accounted for potential bias associated with VPD controls [15], reduced polymerase chain reaction sensitivity among adults with single specimen testing [16], and potential RSVpreF effectiveness against hMPV [17].

Exposure was RSVpreF receipt ≥21 days before the ARI. Eligible KPSC members (per ACIP guidelines) received RSV vaccines at no cost based on FDA indications. KPSC electronic health records captured all vaccines administered within the health system and were supplemented with data from California's Immunization Registry, to which healthcare providers are legally required to report RSV vaccinations [18]. RSVpreF was administered at 7 study hospitals and their associated medical offices. The analysis also included RSV-unvaccinated ED/hospitalization events among an eligible population in which RSV testing was conducted and RSVpreF was not available (Figure 1).

Figure 1.

Study population flowchart, Kaiser Permanente Southern California (KPSC)—November 2023 to April 2024. Samples that were not tested for respiratory syncytial virus (RSV) include nonsalvageable point-of-care swab samples and samples positive for influenza or severe acute respiratory syndrome coronavirus 2 that were not salvaged for RSV testing. The RSV-unvaccinated patients from nonstudy sites include those with emergency department (ED)/hospitalization events at KPSC in which RSV testing was conducted and the RSVpreF vaccine was not available for a population meeting eligibility criteria. Abbreviations: ARI, acute respiratory illness; LRTD, lower respiratory tract disease; SOC, standard-of-care.

Adjusted odds ratios (ORs) and 95% CIs were estimated from multivariable logistic regression and used to calculate VE (VE = [1 − OR] × 100%). Prespecified and model-specified covariates included encounter month, age, sex, self-reported race/ethnicity, modified Charlson Comorbidity Index, and healthcare use in the prior year. The only variable with missing values was race/ethnicity, which was combined as a non-Hispanic multiple/other/unknown category. VE was estimated against ED visits/hospitalizations; by age; for those with severe ARI (defined as ED/hospitalization events requiring oxygen supplementation); for those at high risk of severe disease based on underlying comorbid conditions (asthma, chronic obstructive pulmonary disease [COPD], congestive heart failure [CHF], coronary artery disease, other chronic lung, cardiac, renal, and liver diseases, diabetes mellitus, neurologic conditions, stroke, autoimmune disorders, immunocompromising conditions and medications, human immunodeficiency virus, AIDS, cancers, organ transplant, and blood disorders), as well as among those with CHF or COPD; and for those with critical outcomes (defined as intensive care unit [ICU] admission, invasive/noninvasive mechanical ventilation, new diagnosis of respiratory failure or vasopressor use, or death). Sensitivity analyses were conducted to include only the first ARI episode per patient, and to exclude RSV case patients coinfected with another respiratory pathogen. Programming and analyses were conducted using SAS statistical software, version 9.4. KPSC's institutional review board approved this study and waived informed consent requirements.

RESULTS

Of 15 452 patients aged ≥60 years with ARI ED visits/hospitalizations at study sites, 11 330 (73.3%) had a nasal/nasopharyngeal swab sample collected; of these, 6963 (61.5%) were RSV tested (5820 salvage specimens and 1143 SOC specimens). After the application of eligibility criteria and inclusion of unvaccinated RSV-tested patients from nonstudy sites, 8965 ARI events with tested specimens were included in the study population (Figure 1), 3680 specimens (41.0%) were RSV tested as part of the SOC, and 5285 (59.0%) were salvaged. The majority had comorbid conditions (92.7% had a Charlson Comorbidity Index of ≥1), 45.7% had COPD, 14.6% were immunocompromised, and 56.4% were aged ≥75 years (Table 1). Overall, 0.3% of case patients (2 of 695), 3.6% of strict controls (37 of 1034), and 3.4% of broad controls (282 of 8270) had received the RSVpreF vaccine. Compared with those who remained unvaccinated, patients receiving RSVpreF were older, more often non-Hispanic Asian/Pacific Islander or white, had more outpatient encounters in the prior year, and had more select comorbid conditions. Thirty-two RSV case patients were coinfected.

Table 1.

Characteristics of Patients ≥60 Years of Age With Acute Respiratory Illness Hospital or Emergency Department Encounters, Kaiser Permanente Southern California—November 2023 to April 2024

Characteristic	Patients, No (%)			Strict Analysis P Value	Broad Analysis Controlsb, No. (%) (n = 8270)	Broad Analysis P Value
	Total Study Population (n = 8965)	Case Patients (n = 695	Strict Analysis Controlsa (n = 1034)
Age at index date, y
60–74	3910 (43.6)	283 (40.7)	501 (48.5)	.002	3627 (43.9)	.11
≥75	5055 (56.4)	412 (59.3)	533 (51.5)		4643 (56.1)
Sex
Female	4900 (54.7)	440 (63.3)	549 (53.1)	<.001	4460 (53.9)	<.001
Male	4065 (45.3)	255 (36.7)	485 (46.9)		3810 (46.1)
Months of encounter
Nov–Dec	2566 (28.6)	288 (41.4)	282 (27.3)	<.001	2278 (27.5)	<.001
Jan–Feb	4660 (52)	365 (52.5)	489 (47.3)		4295 (51.9)
Mar–Apr	1739 (19.4)	42 (6)	263 (25.4)		1697 (20.5)
Race/ethnicity
Non-Hispanic Asian/Pacific Islander	1055 (11.8)	98 (14.1)	131 (12.7)	.52	957 (11.6)	.002
Non-Hispanic black	1499 (16.7)	84 (12.1)	148 (14.3)		1415 (17.1)
Hispanic	3023 (33.7)	260 (37.4)	402 (38.9)		2763 (33.4)
Non-Hispanic multiple/other/unknown	95 (1.1)	9 (1.3)	15 (1.5)		86 (1)
Non-Hispanic white	3293 (36.7)	244 (35.1)	338 (32.7)		3049 (36.9)
Modified CCI
0	654 (7.3)	46 (6.6)	102 (9.9)	.13	608 (7.4)	.43
1	854 (9.5)	77 (11.1)	114 (11)		777 (9.4)
2	1103 (12.3)	89 (12.8)	132 (12.8)		1014 (12.3)
≥3	6354 (70.9)	483 (69.5)	686 (66.3)		5871 (71)
No. of hospital encounters in the year before index date
0	5390 (60.1)	501 (72.1)	698 (67.5)	.054	4889 (59.1)	<.001
1	1910 (21.3)	101 (14.5)	195 (18.9)		1809 (21.9)
≥2	1665 (18.6)	93 (13.4)	141 (13.6)		1572 (19)
No. of ED encounters in the year before index date
0	3364 (37.5)	310 (44.6)	405 (39.2)	.03	3054 (36.9)	<.001
1	2173 (24.2)	170 (24.5)	247 (23.9)		2003 (24.2)
≥2	3428 (38.2)	215 (30.9)	382 (36.9)		3213 (38.9)
No. of outpatient encounters in the year before index date
<10	2535 (28.3)	225 (32.4)	310 (30)	.29	2310 (27.9)	.01
≥10	6430 (71.7)	470 (67.6)	724 (70)		5960 (72.1)
Comorbid conditions
Myocardial infarction	1444 (16.1)	89 (12.8)	150 (14.5)	.32	1355 (16.4)	.01
CHF	3277 (36.6)	242 (34.8)	322 (31.1)	.11	3035 (36.7)	.32
Cerebrovascular disease	1339 (14.9)	76 (10.9)	130 (12.6)	.30	1263 (15.3)	.002
Peripheral vascular disease	6056 (67.6)	457 (65.8)	682 (66)	.93	5599 (67.7)	.29
Moderate or severe liver disease	162 (1.8)	10 (1.4)	15 (1.5)	.98	152 (1.8)	.45
Cancer	1555 (17.3)	109 (15.7)	163 (15.8)	.96	1446 (17.5)	.23
Renal disease	3768 (42)	304 (43.7)	382 (36.9)	.005	3464 (41.9)	.34
COPD	4098 (45.7)	314 (45.2)	477 (46.1)	.70	3784 (45.8)	.77
Dementia	1079 (12)	97 (14)	114 (11)	.07	982 (11.9)	.11
AIDS/HIV	27 (0.3)	0 (0)	2 (0.2)	.25	27 (0.3)	.13
Diabetes mellitus
HbA1c ≥7.5	1385 (15.4)	98 (14.1)	178 (17.2)	.33	1287 (15.6)	.50
HbA1c <7.5	3045 (34)	231 (33.2)	323 (31.2)		2814 (34)
Unknown HbA1c	186 (2.1)	18 (2.6)	22 (2.1)		168 (2)
No diabetes diagnosis	4349 (48.5)	348 (50.1)	511 (49.4)		4001 (48.4)
Receipt of RSVpreF vaccine
No	8681 (96.8)	693 (99.7)	997 (96.4)	<.001	7988 (96.6)	<.001
Yes	284 (3.2)	2 (0.3)	37 (3.6)		282 (3.4)
Encounter setting
Hospitalization	4559 (50.9)	317 (45.6)	438 (42.4)	.18	4242 (51.3)	.004
ED	4406 (49.1)	378 (54.4)	596 (57.6)		4028 (48.7)
Immunocompromised
No	7657 (85.4)	606 (87.2)	880 (85.1)	.22	7051 (85.3)	.17
Yes	1308 (14.6)	89 (12.8)	154 (14.9)		1219 (14.7)
Severe ARI
No	4005 (44.7)	327 (47.1)	525 (50.8)	.13	3678 (44.5)	.19
Yes	4960 (55.3)	368 (52.9)	509 (49.2)		4592 (55.5)
High-risk chronic medical conditionsc
No	704 (7.9)	53 (7.6)	104 (10.1)	.08	651 (7.9)	.82
Yes	8261 (92.1)	642 (92.4)	930 (89.9)		7619 (92.1)

Abbreviations: ARI, acute respiratory illness; CCI, Charlson Comorbidity Index; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; ED, emergency department; HbA1c, hemoglobin A1c; HIV, human immunodeficiency virus.

^aIn the strict analysis, case patients were positive for RSV. Controls were positive for a non-vaccine-preventable disease (VPD) and negative for RSV, human metapneumovirus, and VPDs. Identified pathogens for non-VPDs included adenovirus, Chlamydia pneumoniae, coronavirus (229E, HKU1, NL63, OC43), human rhinovirus/enterovirus, parainfluenza (serotypes 1–4), and Mycoplasma pneumoniae.

^bIn the broad analysis, case patients were positive for RSV, and controls included all who were negative for RSV.

^cConditions included asthma, COPD, CHF, coronary artery disease, other chronic lung, cardiac, renal, and liver diseases, diabetes mellitus, neurologic conditions, stroke, autoimmune disorders, immunocompromising conditions and medications, HIV, AIDS, cancers, organ transplant, and blood disorders.

Overall, patient characteristics were similar between patients with salvaged versus SOC specimens. However, compared with SOC specimens, salvaged specimens were more likely to be from younger, nonwhite patients, those with ED presentations, those with a lower Charlson Comorbidity Index, and those receiving RSVpreF (Supplementary Table 2). Compared with those tested for RSV, patients with ARI ED visits/hospitalizations who had no swab sample collected or were not RSV tested were more likely to have had ED visits, earlier in the RSV season, and were more likely to have a lower Charlson Comorbidity Index and lower healthcare use (Supplementary Table 2). Compared with unvaccinated persons, vaccinees were more likely to be older, have an encounter later in the RSV season, and have more prior outpatient encounters and were less likely to be Hispanic (Supplementary Table 3).

To construct the strict control group, the following were removed from the broad controls: 696 testing VPD positive (246 with hMPV, 290 with SARS-CoV-2, and 172 with influenza; not mutually exclusive) and 6540 testing negative for all pathogens, leaving 1034 testing positive for non-VPD pathogens. Comparing 695 case patients with 1034 strict controls, the adjusted VE was 92% (95% CI: 64%–98%), with vaccinees receiving RSVpreF a median (interquartile range [IQR]) of 56 (39–75) days before ARI (Figure 2 and Supplementary Table 4). Comparing 695 case patients with 8270 broad controls (ie, all RSV-negative controls), the adjusted VE was 90% (95% CI: 62%–98%), with vaccinees receiving RSVpreF a median (IQR) of 62 (41–82.5) days before ARI.

Figure 2.

Vaccine effectiveness (VE; with 95% confidence interval [CI]) of the RSVpreF vaccine against respiratory syncytial virus (RSV)–related acute respiratory illness (ARI) hospital and/or emergency department (ED) encounters, by select outcome—Kaiser Permanente Southern California, November 2023 to April 2024. ^aVE was adjusted for age, sex, encounter months, race/ethnicity, Charlson Comorbidity Index, and previous outpatient, inpatient encounters, and ED encounters. ^bIn the strict analysis, case patients were positive for RSV. Controls were positive for a non–vaccine-preventable disease (VPD) and negative for RSV, human metapneumovirus, and VPDs. ^cIn the broad analysis, case patients were positive for RSV, and controls included all who were negative for RSV. ^dSevere ARI was defined as that requiring supplemental oxygen use. ^eHigh-risk ARI was ARI among those with high-risk chronic medical conditions, including asthma, chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), coronary artery disease, other chronic lung, cardiac, renal, and liver diseases, diabetes mellitus, neurologic conditions, stroke, autoimmune disorders, immunocompromising conditions and medications, human immunodeficiency virus, AIDS, cancers, organ transplant and blood disorders. ^fCritical outcomes include intensive care unit admission, mechanical ventilation, respiratory failure, vasopressor use, and death.

VE estimates against additional ARI end points demonstrated a similar degree of protection, ranging from 84% to 95%, although age-stratified estimates had wide CIs for the youngest patients (Figure 2 and Supplementary Table 4). VE against ED visits/hospitalizations was high among vulnerable populations, such as those aged ≥75 years (VE, 95% [95% CI: 60%–99%]) or ≥80 years (95% [62%–99%]) and those with high-risk chronic medical conditions (92% [65%–98%]). VE against hospitalization was 91% [95% CI: 30%–99%]. Estimates for VE against severe ARI were similar to those for VE against overall ARI, using strict and broad controls (strict controls 92% [95% CI: 35%–99%]; broad controls, 90% [26%–99%]).

Overall, 329 patients contributed >1 ARI >30 days apart. Sensitivity analyses including only the first ARI for each patient demonstrated results similar to those of the main analyses (VE for strict analysis, 92% [95% CI: 64%–98%]; VE for broad analysis, 90% [61%–98%]). Removal of 32 RSV case patients with coinfections resulted in similar VE similar to the main ED/hospitalization findings (VE for strict controls, 91% [95% CI: 63%–98%]; VE for broad controls, 90% [60%–98%]; Table 2).

Table 2.

Sensitivity Analyses of RSVpreF Vaccine Effectiveness Against Respiratory Syncytial Virus–Related Acute Respiratory Illness Hospitalizations or Emergency Department Visits—Kaiser Permanente Southern California, November 2023 to April 2024

Analysis	Total Population, No.	Test-Negative Controls, No. (%)		Test-Positive Case Patients, No. (%)		VE (95% CI)
		Unvaccinated	Vaccinated	Unvaccinated	Vaccinated	Crude	Adjusteda
RSV-related ARI ED visits or hospitalizations—1st event only
Strict definition/primary analysisb	1720	991 (96.4)	37 (3.6)	690 (99.7)	2 (0.3)	92 (68–98)	92 (64–98)
Broad definition/sensitivity analysisc	8631	7680 (96.6)	267 (3.4)	682 (99.7)	2 (0.3)	92 (66–98)	90 (61–98)
RSV-related ARI ED visits or hospitalizations—coinfected patients removed
Strict definition/primary analysisb	1697	997 (96.4)	37 (3.6)	661 (99.7)	2 (0.3)	92 (66–98)	91 (63–98)
Broad definition/sensitivity analysisc	8933	7988 (96.6)	282 (3.4)	661 (99.7)	2 (0.3)	91 (65–98)	90 (60–98)

Abbreviations: ARI, acute respiratory illness; CI, confidence interval; ED, emergency department; RSV, respiratory syncytial virus; VE, vaccine effectiveness.

^aVE was adjusted for age, sex, encounter months, race/ethnicity, Charlson Comorbidity Index, previous outpatient encounters, previous inpatient encounters, and previous ED encounters.

^bIn the strict analysis, case patients were positive for RSV. Controls were positive for a non­–vaccine-preventable disease (VPD) and negative for RSV, human metapneumovirus, and VPDs. Identified pathogens for non-VPDs included adenovirus, Chlamydia pneumoniae, coronavirus (229E, HKU1, NL63, OC43), human rhinovirus/enterovirus, parainfluenza (serotypes 1–4), and Mycoplasma pneumoniae.

^cIn the broad analysis, case patients were positive for RSV, and controls included all who were negative for RSV.

DISCUSSION

The RSVpreF vaccine was 92% effective against both RSV-related ARI ED visits/hospitalizations and severe ARI in this older adult real-world population, indicating that it has higher VE against severe outcomes than other current respiratory vaccines, including SARS-CoV-2, influenza, and pneumococcal vaccines [19–21]. To our knowledge, this is the first real-world estimate of protection against severe RSV disease among adults that is not based exclusively on treatment setting but rather uses an objective severity definition (ie, oxygen supplementation). Importantly, the effectiveness of RSVpreF remained high when the population was limited to those at the highest risk for severe RSV disease, that is, persons with underlying medical conditions (91%) and those aged ≥75 years (95%), in contrast to the VE reduction seen in the oldest adults for influenza vaccine [22]. These results from first season of RSVpreF use indicate potential for a substantial population-level impact on severe RSV-related illness among older adults with sustained vaccine uptake.

The pivotal prelicensure trial for the RSVpreF vaccine included an objective estimate of VE against severe RSV-associated LRTD, defined as any medically attended visit requiring hospitalization or new/increased respiratory support [23]. However, the study accrued too few of these events to generate a stable VE estimate [8]. The phase 3 trial for RSVpreF3 defined severe disease by clinical signs or assessment by the site investigator, reporting vaccine efficacy of 94.1% (95% CI: 62.4%–99.9%) against severe RSV-associated LRTD [24]. Only 2 of these events included hospitalization. Our VE results are consistent with these estimates but assess VE in a real-world population that includes more people at high risk for severe outcomes (eg, immunocompromised persons and >90% with comorbid conditions).

Other postlicensure studies have provided VE estimates against RSV-related medically attended events. For RSVpreF and RSVpreF3 vaccines combined, VE estimates were 76%–80% for RSV-related hospitalizations [11, 25–27] and 77%–79% [11, 26] for RSV-related ED/urgent care visits (Figure 3). Furthermore, a Medicare analysis among patients with end-stage-renal disease found VE of 72% against RSV hospitalizations but relied on International Classification of Diseases codes rather than laboratory results to determine RSV status [28]. To our knowledge, however, no studies have evaluated a severity end point outside of that defined by care setting. Clinician judgment regarding patient hospital admission may yield a clinically heterogeneous patient population with respect to disease severity [29]. Thus, relying on setting alone may not uniformly capture severe disease, and adding the oxygen supplementation requirement, as we have done, may further standardize results compared with treatment setting alone. Given that about 80% of hospitalized older adults with RSV infection require standard-flow oxygen [6], capturing vaccine performance in this population is clinically meaningful. KPSC uses regionally standardized protocols that limit oxygen therapy to members presenting to the ED or hospitalized ARI patients with evidence of substantial respiratory compromise, and, on this basis, only 55% of the study's ARI ED/hospitalization events required oxygen supplementation. Taken together, our findings and others indicate that the VE of RSVpreF for more severe disease is higher than for events managed exclusively in the outpatient setting.

Figure 3.

Respiratory syncytial virus (RSV) vaccine effectiveness estimates in the first season after vaccine introduction, by select outcomes, from additional US real-world studies [11, 25–27]. The Veterans Healthcare Administration (VHA) study used target trial emulation methods. Population sizes represent the number in each analysis age stratum followed for RSV outcome; all other estimates from test-negative design studies and population size represent the number of outcome events included. Abbreviations: ARI, acute respiratory illness; CI, confidence interval; ED, emergency department; RLI, RSV-like illness, defined by clinical discharge diagnosis code of ARI, respiratory symptoms consistent with ARI, or nonrespiratory symptoms that could be consistent with ARI (eg, sepsis).

Reassuringly, our study similarly found a VE of 90% against critical outcomes (ie, ICU admission or mechanical ventilation), similar to the VISION study’s results: 81% VE against RSV-related critical illness (ie, ICU admission or in-hospital death). Although similar, VE in the VISION study may have been slightly lower due to reliance on SOC testing alone, which is often performed more frequently in patients with more severe disease. The VISION study reported that 20% of their hospitalized patients met these criteria, similar to other studies reporting that approximately 16%–20% of older adults with RSV hospitalization will have an ICU admission [30, 31].

The RSVpreF vaccine also conferred protection against RSV-related ARI ED visits/hospitalizations among the oldest age groups. The VE was 95% (95% CI: 60%–99%) for those aged ≥75 and 95% (62%–99%) for those aged ≥80 years. The RENOIR study also reported high VE in the oldest age groups (100% for those aged 70–79 years [95% CI: −51% to 100%]), though CIs were wide due to the limited number of events. Real-world studies have provided estimates in high-risk groups for RSVpreF and RSVpreF3 combined (Figure 3). Among those aged ≥75 years, the VISION and IVY studies reported 76%–79% VE against RSV-related hospitalizations or ED visits [11, 25], and Fry et al [27] reported 76% VE against hospitalization. The Veterans Health Administration (VHA) study found VE against any RSV infection of 72%–79% by age group, with slightly lower VE in the oldest category [26]. This latter finding could be due to the end point (any RSV infection), regardless of diagnosis or acuity, and more lower-acuity (eg, outpatient) events may have been included if testing was more frequent in older adults. Notably, our study demonstrates that the unadjuvanted RSVpreF vaccine has high VE in the oldest age groups, unlike what has been observed with influenza vaccines, where adjuvanted vaccines were required to maintain high VE in this population [32–34] and were preferentially recommended for older adults by ACIP in 2023 [35]. The potential public health impact of RSVpreF among persons ≥75 years has emerged globally, with substantial population-level reduction of RSV-related hospitalizations in this age group documented after the introduction of RSVpreF among 75–79 year-olds in the United Kingdom (eg, a 62% decrease with 69% vaccine uptake in Scotland) [36, 37].

We also demonstrated protective effectiveness of RSVpreF in patients with high-risk conditions. VE was 92% (95% CI: 65%–98%) among those with any high-risk medical condition and 93% (49%–99%) among those with CHF/COPD specifically. The pivotal phase 3 study for RSVpreF found high VE at the end of season 1 among those with ≥1 high-risk condition (82% [95% CI: 17%–98%]). However, CIs were wide, and LRTI events not requiring hospitalization or ED visits were included. With respect to high-risk conditions, both the VISION and VHA studies report VE among the immunocompromised, which is informative, but our study also provides a VE estimate for RSV-related severe disease among all CDC-defined high-risk conditions and among those with specific high-risk diagnoses, such as CHF/COPD [38].

Salvaging respiratory specimens that were otherwise not RSV tested was a notable strength of our study. Complicating the evaluation of RSV vaccine VE in observational postlicensure studies is the reliance on SOC testing. Undertesting of older adults during routine clinical care contributes to underrecognition of RSV [39]. One US study found that, across healthcare institutions, the median percentage of LRTI hospitalizations with RSV testing by the hospital was only 4.3% [40]. While the use of multiplex assay platforms in the post–COVID-19 pandemic era may have increased testing, the proportion of hospitalizations for respiratory tract illness with RSV testing remains relatively low, even among high-risk populations [30, 41]. Therefore, we salvaged specimens collected for influenza/SARS-CoV-2 testing, thereby boosting RSV detection. Indeed, data from KPSC have found that the addition of salvaged swab samples to those collected in routine care increases RSV detection by nearly 4-fold, compared with estimates obtained using only SOC swab samples [42]. Given the introduction of new preventive measures to protect against disease and death due to RSV, expansion of RSV testing is warranted, particularly in high-acuity settings—to better characterize populations at highest risk for RSV infection and potentially to improve treatment algorithms.

Another strength of our analysis was the completeness of vaccine exposure data, obtained through KPSC electronic health records, which are linked with California's Immunization Registry. Given clinicians’ legal mandate to report RSV vaccinations to the registry, misclassification of vaccination status is unlikely. This is in contrast with the heterogeneous sources used in the VISION study, the self-report of vaccination status accepted in the IVY study, or the reduced ability to capture vaccines administered external to VHA hospitals in the VHA study. These potential sources of vaccination status misclassification may have resulted in the lower VE estimates seen in the other studies.

The current study also has limitations. The analysis included shorter follow-up time after vaccination than in other real-world studies, likely contributing to our higher VE estimates. Indeed, our results, which included events a median (IQR) of 62 (41–82.5) days after vaccination, were comparable to those of the VISION study’s sensitivity analysis, which found a VE of 90% against RSV-associated hospitalization with vaccine receipt 14–59 days earlier, compared with ≥60 days earlier [11]. At the time of our study, 2 RSV vaccines were licensed for use in older adults (RSVpreF and RSVpreF3), but our study is limited to evaluating RSVpreF. Additional limitations include low RSVpreF uptake in the first season after licensure but before routine recommendation for adults aged ≥75 years and those aged 60–74 years and at increased risk of severe RSV disease, which limited the ability to conduct additional subgroup analyses (eg, among immunocompromised and frail populations). Additional analyses will be conducted to address this data gap as ARI ED visits/hospitalizations accrue in future RSV seasons.

In conclusion, the RSVpreF vaccine provides protection to older adults against RSV-related hospitalization, ED visits, and severe RSV disease requiring oxygen, among high-risk subgroups and among those aged ≥75 or ≥80 years, indicating that RSV vaccination programs can protect against serious outcomes due to RSV in groups at the highest risk of severe disease due to age and underlying conditions. With pending expansion of recommendations to adults aged 50–59 years [42], continued evaluation of the potential protective benefits of this vaccine is warranted.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.