Household Transmission and Genomic Diversity of Respiratory Syncytial Virus in the United States, 2022–2023

Abstract

Background

Household transmission of respiratory viruses may drive community spread. Few recent studies have examined household respiratory syncytial virus (RSV) transmission in the United States.

Methods

We conducted a prospective community-based cohort study from 1 June 2022 to 31 May 2023. Participants had blood samples collected and completed nasal swabs and surveys at least weekly, irrespective of symptoms. We tested serum for RSV antibody, nasal swabs by quantitative reverse transcription polymerase chain reaction (RT-qPCR), and performed whole genome sequencing. We evaluated secondary RSV transmission and associated risk factors based on a log-linear Poisson regression model.

Results

RSV was detected among 310 (10%) participants within 200 (20%) households. Most (94%) index cases were symptomatic. We identified 37 cases of potential secondary transmission within 14 days of a distinct index case (10%, 95% confidence interval [CI]: 7%, 14%); median age of index and secondary cases were 6 (interquartile range [IQR]: 3–10) and 35 (7–41) years, respectively, with 89% (24/27) of index cases aged 6 months to 12 years. Factors associated with increased risk of RSV transmission included index case viral detection ≥1 week and contact age ≤12 years. Of 120 sequenced specimens, the main lineages represented were A.d.5.2 (n = 37) and A.d.1 (n = 30). Sequenced viruses from households with ≥2 RSV infections were similar when occurring within ≤14 days (mean pairwise difference 4 [range 0–13], n = 17 households), compared to those >14 days (137 [37–236], n = 2).

Conclusions

Most RSV household transmission occurs from infants and young children to adults. Viral genome sequencing demonstrated that multiple household infections within a 14-day period are likely due to within-household transmission.

In a US-based cohort study, respiratory syncytial virus (RSV) transmission occurred in 10% of household contacts. Index and secondary cases were frequently young children and adults, respectively. Genomic sequencing demonstrated infections ≤2 weeks apart were likely due to within-household transmission.

Households are an important setting for the spread of respiratory viruses and provide an important framework to understand risk factors for viral transmission. Prior to the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, household-based studies in South Africa and Kenya indicated that school-aged children frequently introduce respiratory syncytial virus (RSV) into households, but few recent US-based studies using molecular testing methods have been conducted [1] and viral genomics have rarely been used to confirm epidemiologic inferences of household transmission [1–4].

Two years into the coronavirus disease 2019 (COVID-19) pandemic, RSV experienced a global resurgence with increased hospitalization rates among children [5]. Updated estimates of RSV household transmission are needed to better characterize transmission patterns after pandemic mitigation measures were lifted. Furthermore, with the recent introduction of new vaccines and monoclonal antibodies to prevent RSV-associated lower respriatory tract disease [6–8], updated demographic, clinical, and household characteristics that impact household transmission dynamics are critical for identifying specific populations for preventative measures.

As part of a community-based cohort study, we estimated the risk of secondary RSV infection (ie, proportion of household contacts infected after exposure to an index case, sometimes referred to as the “secondary attack rate”) and identified clinical, virologic, and immunologic factors associated with household transmission in the pre-RSV immunization era during 2022–2023. We additionally used viral genome sequencing to corroborate epidemiologic inferences of household transmission.

METHODS

Study Design and Population

We analyzed data collected from 1 June 2022, to 31 May 2023, from the CASCADIA study, a community-based prospective cohort study among households of infants, children, and adults in the metropolitan regions of Seattle, Washington, and Portland, Oregon, in the United States [9, 10]. The study used remote active surveillance of respiratory viruses among households, with at least weekly at-home collection of nasal swab specimens from participants with and without symptoms. Study enrollment was open to those aged ≥6 months to 50 years living within the University of Washington (UW) catchment area (ie, King, Snohomish, and Pierce Counties in Washington State) or the Kaiser Permanente Northwest catchment area (ie, Oregon and southern Washington). Not all household members were required to enroll in the study.

This study was reviewed and approved by the Kaiser Permanente Interregional Institutional Review Board. This activity was conducted consistent with applicable federal law and CDC policy (See 45 C.F.R. part 46.114; 21 C.F.R. part 56.114).

Data Collection

All eligible and consenting participants or their parent/guardian completed an enrollment survey and attended an in-person baseline blood draw appointment. Additionally, each household received nasal swab kits delivered to their home, which were replenished regularly and included visual guides for self-swabbing [10]. Participants ≥13 years were instructed to self-collect nasal swabs, while parents/guardians collected nasal swabs for children ≥6 months to 12 years. All swabs were tested for multiple respiratory viruses including RSV, allowing capture of both asymptomatic and symptomatic RSV infection.

At least weekly, participants (or their parent/guardian) completed online symptom surveys [11, 12] and collected nasal swabs, irrespective of symptoms to allow for capture of both asymptomatic and symptomatic RSV infection. Participants could also report new or worsening symptoms at any time. If a participant reported ≥1 new acute respiratory illness (ARI) symptom >72 hours after their last swab, they were prompted to collect an additional swab. ARI symptoms included fever, cough, sore throat, shortness of breath, myalgia, and rhinorrhea. Participants completed illness surveys with additional questions about symptoms, care-seeking, and absenteeism up to 30 days after symptom(s) or positive test result. Individuals underwent semi-annual and annual blood draws in addition to draws 30 days after infection or vaccination for influenza or SARS-CoV-2.

Laboratory Testing and Genomic Sequencing

All nasal swab specimens were tested for RSV, influenza, and SARS-CoV-2 using a quantitative reverse transcription polymerase chain reaction (RT-qPCR) platform developed by UW Brotman Baty Laboratory as previously described [13–15]. Specimens from participants who were symptomatic within 72 hours from their swab collection, and those positive/inconclusive on RT-qPCR testing, underwent secondary multiplex RT-qPCR testing using a custom, TaqMan-based Open Array panel (Thermo Fisher), including RSV subtype (A&B) and additional respiratory pathogen targets.

We defined all specimens with a cycle threshold (C_t) ≤37 or relative cycle threshold (C_rt) <28 as RSV-positive, including those initially classified as “inconclusive” (C_t ≤37 or C_rt <28 in 1 of 2 replicates, rather than 2 of 2 replicates). Specimens with RSV C_t ≤28.5 or C_rt <18.5 were selected for sequencing to ensure high-quality results and enable genome assembly. We used C_rt values as a proxy for semiquantitative viral load.

All serum samples were tested for RSV prefusion IgG binding antibody using a commercial electrochemiluminescence immunoassay (ECLIA) (Mesoscale Diagnostics), with antibody titers presented in log-transformed arbitrary units per milliliter (log-AU/mL). Further laboratory and sequencing methods are previously described [13–17] and detailed in the supplemental appendix.

Statistical Analysis

We analyzed demographic, clinical, and behavioral characteristics among households using R version 4.3.2 [18]. We restricted our analysis to multi-person households (ie, ≥2 household members enrolled). An index case was defined as the first household member with RSV detected. Co-primary index cases were defined as 2 or more household members with the same date of RSV detection. Potential secondary household transmission was defined as additional household member(s) with RSV detected 1–14 days after the identified index case(s) to align with the weekly testing protocol and prior RSV transmission studies [1, 3, 19]. We defined a possible repeat RSV infection within the same individual as two RSV detections separated by >14 days with no positive results in the intervening period. Generation interval was defined as the number of days between the index case's first RSV-positive specimen collection and the secondary case's first RSV-positive specimen collection. See Supplementary Table 1 for terms and definitions.

In our primary model, the analytical cohort consisted of household contacts of a distinct index case (each household contact provides one observation in the data set); the outcome of interest was secondary RSV infection among contacts within ≤14 days of the index case. We evaluated potential secondary RSV transmission and risk factors during the incident household illness episode based on a log-linear Poisson regression model fitted using generalized estimating equations (GEE). Variables included in the model are listed in the Supplemental appendix. We constructed 95% confidence intervals (CIs) and obtained P values using robust standard error estimates to account for clustering by household and violations of the Poisson distributional assumption.

We conducted sensitivity analyses examining the risk of household RSV transmission and associated risk factors including all households with co-primary index cases, as well as testing more conservative generation interval timeframes to define potential secondary transmission within 3–14 days or 1–7 days (compared to 1–14 days). Additionally, we examined the risk of RSV transmission restricting to households with all members enrolled.

RESULTS

Of 1188 total eligible households, we included 1009 multi-person households with 3058 participants. Overall, we had 97% retention of enrolled households over the study period (Figures 1, Supplementary 1; Tables 1, Supplementary 2). Households consisted of 2–8 participants (median: 3) and 52% had complete ascertainment (ie, all members enrolled). Most (67%) reported an annual household income of >$100 000. Approximately half of the participants were infants and children <18 years: 3% 6 months–1 year, 8% 2–4 years, 30% 5–12 years, and 10% 13–17 years.

Figure 1.

Study flow diagram, June 2022 to May 2023. *Fully enrolled = individual who completed consent, baseline enrollment survey, and ≥1 nasal swab during the study period. ^‡Index case = first household member with RSV detected; Co-primary index cases = 2 + household members with same date of RSV detection. ^† Potential secondary transmission defined as occurring when a household member with RSV is detected 1–14 d after index case. ^§Unlikely secondary transmission defined as occurring when there is only 1 RSV case detected in the household OR when there are ≥2 RSV cases in a household but secondary case detected >14 d after index case. +Individuals not enrolled at the time of the incident household illness episode (but enrolled later during the study period); This does not include household members never enrolled. Abbreviation: RSV, respiratory syncytial virus.

Table 1.

Demographic characteristics of households with and without a secondary RSV infection within 14 d of a distinct index case during the incident household illness episode, Washington and Oregon, United States, 2022-2023 (n = 170 households)

Characteristic		Potential secondary transmissiona	Unlikely secondary transmissionb	Total
Characteristic of household		(n = 27)	(n = 143)	(n = 170)
Household densityc	2–4 individuals	19 (70.4%)	116 (81.1%)	135 (79.4%)
	5+ individuals	8 (29.6%)	27 (18.9%)	35 (20.6%)
Housing type	House, condo, or townhouse	26 (96.3%)	138 (96.5%)	164 (96.5%)
	Apartment	1 (3.7%)	3 (2.1%)	4 (2.4%)
	Other	0 (0.0%)	2 (1.4%)	2 (1.2%)
Household enrollment	Number of participants, median [min, max]	4 [2, 6]	3 [2, 5]	3 [2, 6]
	No child <5	14 (51.9%)	95 (66.4%)	109 (64.1%)
	Child <5, but no childcare	3 (11.1%)	8 (5.6%)	11 (6.5%)
	Child <5 in childcared	10 (37.0%)	40 (28.0%)	50 (29.4%)
Income $100,000+		21 (77.8%)	108 (75.5%)	129 (75.9%)
Smoker in household		3 (11.1%)	11 (7.7%)	14 (8.2%)
Study site	Kaiser Permanente Northwest (KPNW)	6 (22.2%)	73 (51.0%)	79 (46.5%)
	University of Washington (UW)	21 (77.8%)	70 (49.0%)	91 (53.5%)
Characteristic of index case		(n = 27)	(n = 143)	(n = 170)
Age	Median [min, max] years	6 [1, 48]	9 [0, 50]	9 [0, 50]
	6 m to 1 y	4 (14.8%)	9 (6.3%)	13 (7.6%)
	2–4 y	7 (25.9%)	22 (15.4%)	29 (17.1%)
	5–12 y	13 (48.1%)	64 (44.8%)	77 (45.3%)
	13–50 y	3 (11.1%)	48 (33.6%)	51 (30.0%)
Female sex at birth		13 (48.1%)	84 (58.7%)	97 (57.1%)
Gender	Female	14 (51.9%)	83 (58.0%)	97 (57.1%)
	Male	13 (48.1%)	58 (40.6%)	71 (41.8%)
	Other	0 (0.0%)	2 (1.4%)	2 (1.2%)
Race	American Indian/Alaska Native	0 (0.0%)	1 (0.7%)	1 (0.6%)
	Asian	1 (3.7%)	6 (4.2%)	7 (4.1%)
	Black/African American	0 (0.0%)	1 (0.7%)	1 (0.6%)
	White	19 (70.4%)	113 (79.0%)	132 (77.6%)
	Multiracial	7 (25.9%)	20 (14.0%)	27 (15.9%)
	Other	0 (0.0%)	2 (1.4%)	2 (1.2%)
Hispanic		2 (7.4%)	12 (8.4%)	14 (8.2%)
Any smoking		1 (3.7%)	1 (0.7%)	2 (1.2%)
Any comorbiditiese		3 (11.1%)	32 (22.4%)	35 (20.6%)
Vaccinated for influenza in 2021–2022 seasonf		22 (81.5%)	116 (81.1%)	138 (81.2%)
Masking in public	Any	16 (59.3%)	100 (69.9%)	116 (68.2%)
	Never	9 (36.0%)	40 (28.6%)	49 (29.7%)
	Missing	2 (7.4%)	3 (2.1%)	5 (2.9%)
Baseline binding antibody titers	Median [min, max]	4.6 [1.5, 5.7]	5.1 [1.5, 5.9]	5.1 [1.5, 5.9]
	median (5.1 log-AU/mL)	17 (63.0%)	72 (50.3%)	89 (52.4%)
RSV subtype detected 2022–2023	RSV A	24 (88.9%)	115 (80.4 %)	139 (81.8%)
	RSV B	4 (14.8%)	32 (22.4%)	36 (21.2%)
Relative cycle threshold (Crt)	Median [min, max]	16.6 [8.5, 27.7]	18.9 [4.6, 27.4]	18.9 [4.6, 27.7]
	≤ median (18.9 Crt)	17 (63.0%)	60 (42.0%)	77 (45.3%)
Duration of viral detection ≥ 1 wk	12 (44.4%)	10 (7.0%)	22 (12.9%)
Presence of viral codetectiong	Any	6 (22.2%)	52 (36.4%)	58 (34.1%)
	SARS-CoV-2	1 (3.7%)	8 (5.6%)	9 (5.3%)
	Influenza	0 (0.0%)	7 (4.9%)	7 (4.1%)
	Rhinovirus	3 (11.1%)	27 (18.9%)	30 (17.6%)
Any ARI symptom(s)h	1+ ARI symptoms	25 (92.6%)	134 (93.7%)	159 (93.5%)
	2+ ARI symptoms	23 (85.2%)	105 (73.4%)	128 (75.3%)
	Cough and/or rhinorrhea	25 (92.6%)	133 (93.0%)	158 (92.9%)
Any care seeking during illnessi		7 (25.9%)	20 (14.0%)	27 (15.9%)
Any behavior change to reduce transmission in the air (e.g., masking)j		8 (29.6%)	56 (39.2%)	64 (37.6%)
Any behavior change to reduce transmission on surfaces (e.g., disinfecting)k		9 (33.3%)	50 (35.0%)	59 (34.7%)
Characteristic of household contact		(n = 73)	(n = 296)	(n = 369)
Age	Median [min, max] years	37 [2, 49]	38 [0, 49]	38 [0, 49]
	6 m to 1 y	0 (0.0%)	4 (1.4%)	4 (1.1%)
	2–4 y	4 (5.5%)	14 (4.7%)	18 (4.9%)
	5–12 y	23 (31.5%)	75 (25.3%)	98 (26.6%)
	13–50 y	46 (63.0%)	203 (68.6%)	249 (67.5%)
Female sex at birth		37 (50.7%)	176 (59.5%)	213 (57.7%)
Gender	Female	36 (49.3%)	169 (57.1%)	205 (55.6%)
	Male	35 (47.9%)	119 (40.2%)	154 (41.7%)
	Other	2 (2.7%)	8 (2.7%)	10 (2.7%)
Race	American Indian/Alaska Native	0 (0.0%)	1 (0.3%)	1 (0.3%)
	Asian	1 (1.4%)	18 (6.1%)	19 (5.1%)
	Black/African American	1 (1.4%)	4 (1.4%)	5 (1.4%)
	Native Hawaiian/Pacific Islander	1 (1.4%)	0 (0.0%)	1 (0.3%)
	White	59 (80.8%)	245 (82.8%)	304 (82.4%)
	Multiracial	10 (13.7%)	24 (8.1%)	34 (9.2%)
	Other	1 (1.4%)	4 (1.4%)	5 (1.3%)
Hispanic		5 (6.8%)	16 (5.4%)	21 (5.7%)
Any smoking		2 (2.7%)	12 (4.1%)	14 (3.8%)
Any comorbiditiesl		32 (43.8%)	119 (40.2%)	151 (40.9%)
Vaccinated for influenza in 2021–2022 seasonm		55 (75.3%)	249 (84.1%)	304 (82.4%)
Masking in public	Any	56 (76.7%)	230 (77.7%)	286 (77.5%)
	Never	17 (23.3%)	63 (21.5%)	80 (21.9%)
	Missing	0 (0.0%)	3 (1.0%)	3 (0.8%)
Baseline binding antibody titers	Median [min, max]	5.1 [2.2, 6.0]	5.1 [1.5, 5.9]	5.1 [1.5, 6.0]
	≤ median (5.1 log-AU/mL)	25 (34.2%)	99 (33.4%)	124 (33.6%)

^aPotential secondary transmission defined as occurring when a secondary household member with RSV is detected 1-14 days after index case

Unlikely secondary transmission defined as occurring when there is only 1 RSV case detected in the household OR when there are 2+ RSV cases in a household but secondary case detected >14 days after index case

Household density represents the number of household members regardless of enrollment in the study (thus, this may be greater than the number of individuals enrolled in the study if not all household members are enrolled)

Daycare or school attendance among at least one child <5 years in the household as reported at enrollment

Comorbidities include asthma, chronic obstructive pulmonary disease (COPD) (including chronic bronchitis and emphysema), sleep apnea, heart disease, congenital heart disease, heart failure, down syndrome, hypertension (high blood pressure), diabetes (high blood sugar), liver condition, weak or failing kidneys, cancer or malignancy, arthritis, stroke, deep vein thrombosis (DVT) or pulmonary embolism (PE), sickle cell disease or thalassemia, weakened immune system, depression, anxiety, thyroid issues, or other health diagnosis

Vaccinated for influenza in 2021-2022 season versus no vaccination

Other viruses codetected among index cases in households with and without secondary transmission included adenovirus (n=1 vs. n=5), enterovirus (n=1 vs. n=4), human coronavirus (n=1 vs. n=5), and human parainfluenza virus (n=1 vs. n=2); no index cases had human metapneumovirus codetected

ARI, acute respiratory illness symptom reported within ±7 days of the individual RSV illness episode’s first positive specimen collection, including fever​, cough​, sore throat​, shortness of breath​, myalgia​, and/or rhinorrhea

Self-reported care seeking defined as seeking health care (from a healthcare provider) during illness

Includes masking, sleeping separately, covering cough/sneeze

Includes handwashing, cleaning, disinfecting

Comorbidities include asthma, chronic obstructive pulmonary disease (COPD) (including chronic bronchitis and emphysema), sleep apnea, heart disease, congenital heart disease, heart failure, down syndrome, hypertension (high blood pressure), diabetes (high blood sugar), liver condition, weak or failing kidneys, cancer or malignancy, arthritis, stroke, deep vein thrombosis (DVT) or pulmonary embolism (PE), sickle cell disease or thalassemia, weakened immune system, depression, anxiety, thyroid issues, or other health diagnosis

Vaccinated for influenza in 2021-2022 season versus no vaccination

RSV was detected among 310 (10%) individuals within 200 (20%) households from 5 September 2022, to 18 May 2023 (Figure 1). This included 248 participants with RSV A, 57 participants with RSV B, and 5 participants with both RSV A and RSV B (detected a median of 77 days apart; range: 26–196). We also identified eight individuals with possible repeat RSV A infection (3%; range: 21–104; Supplementary Table 3), including 6 children (age: 1–7) and 2 adults (age: 35–35), none of whom self-reported immunocompromising conditions. No cases of RSV A and B codetection or repeat RSV B infections were identified. The most common symptoms were rhinorrhea (84%) and cough (72%).

Of 200 incident household RSV illness episodes, 170 had a distinct index case and 30 had co-primary index cases (Figures 1, Supplementary Figures 1 and 2). In our primary analysis, we included 170 households with a distinct index case (Table 1). A median of 3 (range: 2–6) participants were enrolled in each household. Approximately 58% of these households had complete ascertainment. Potential secondary transmission was detected in 27 (16%) households, with a median generation interval of 7 days (IQR: 7–13) between index and secondary cases. Most distinct index cases (24/27; 89%) were children ≤12 years [median: 6 years (IQR: 3–10); 4 (15%) 6 months-1 year, 7 (26%) 2–4 years, 13 (48%) 5–12 years, 0 (0%) 13–29 years, 3 (11%) 30–50 years], while most secondary cases (21/37; 57%) were adults [median age: 35 years (IQR: 7–41); 4 (11%) 2–4 years, 12 (32%) 5–12 years, 0 (0%) 13–29 years, and 21 (57%) 30–50 years]. Approximately 37% of the households with potential secondary transmission reported childcare attendance by a child <5 years in the household (vs 28% among households with unlikely transmission). Thirty percent (8/27) of households with potential secondary transmission reported sharing a kitchen or living space with ≥5 other people, compared to 19% (27/143) households with unlikely transmission. Most distinct index cases (159/170, 94%) and secondary cases (33/37, 89%) reported ≥1 ARI symptom(s). However, few sought medical care during their illness (27 distinct index cases [24 children, 3 adults]; 4 secondary cases [2 children, 2 adults]). Among 11 asymptomatic index cases, only 2 (1 child, 1 adult) had a household contact with secondary RSV infection.

Among 369 household contacts enrolled in the study during the first household RSV illness episode, we identified 37 cases of potential secondary transmission within 14 days of a distinct index case (10.0%, 95% CI: 7.0, 14.4%). Risk of secondary transmission remained similar across sensitivity analyses, with proportions ranging from 8.2% to 10.7% (Supplementary Table 4). Comparing contacts with the same demographics, clinical, and household characteristics, contact age ≤12 years was associated with a higher estimated relative risk of secondary RSV infection; specifically, we observed a 3.5 (95% CI: 1.3, 9.4) and 2.4 (95% CI: 1.1, 5.5) times higher risk among contacts 6 months to 4 years and 5–12 years, respectively, compared to contacts 13–50 years (Table 2). Additionally, household contacts exposed to a distinct index case with viral detection ≥1 week (vs <1 week) had a 5.6 (95% CI: 2.6, 12.0) times higher risk of secondary RSV infection. There was no evidence of an association of viral load of the index case, or baseline RSV binding antibody titers of the index or the secondary case, with the risk of secondary transmission (Table 2, Figure 2). Using interaction terms yielded no evidence that baseline binding antibody titers of the household contact modified the association between viral load of the index case and household transmission. Notably the risk factors of index viral detection >1 week and contact age ≤12 years were robust to all sensitivity analyses (Supplementary Figure 1, Tables 4–6). Factors associated with lower transmission risk included index presence of viral codetection and household contact influenza vaccination. Additionally, household childcare attendance was associated with secondary transmission, with higher estimated risk among those in households with a child <5 years in childcare compared to those with no children <5 years in the household.

Table 2.

Factors Associated With Household Transmission of RSV Within ≤14 d of a Distinct Index Case, Washington and Oregon, United States, 2022–2023 (n = 369 Contacts)

Characteristic		% of Contacts Infecteda	Unadjusted Relative Risk (95% CI)	Adjusted Relative Risk (95% CI)	P Valueb
Characteristic of household
Household densityc	2–4 individuals	9% (25/271)	Ref	Ref	.920
	5 + individuals	12% (12/98)	1.33 (.61, 2.89)	0.96 (.45, 2.06)
Housing type	House, condo, or townhouse	10% (36/359)	Ref	Ref	.878
	Otherd	10% (1/10)	1.00 (.14, 7.33)	1.20 (.12, 12.46)
Childcare attended by child <5 y in householde	No child <5	7% (17/227)	Ref	Ref	.036
	Child <5, but no childcare	17% (5/30)	2.23 (.71, 6.97)	0.58 (.11, 3.00)
	Child <5 in childcare	13% (15/112)	1.79 (.82, 3.89)	2.56 (.45, 14.63)
Study sitef	KPNW	4% (7/162)	Ref	Ref	<.001
	UW	14% (30/207)	3.35 (1.39, 8.07)	5.50 (2.76, 10.96)
Characteristic of index case
Age	6 mo to 4 y	17% (16/92)	5.45 (1.65, 17.96)	1.73 (.40, 7.44)	.182
	5–12 y	10% (18/183)	3.08 (.94, 10.10)	3.24 (.90, 11.62)
	13–50 y	3% (3/94)	Ref	Ref
Vaccinated for influenza in 2021–2022 seasonh	Yes	10% (29/298)	0.86 (.34, 2.19)	1.35 (.45, 4.10)	.591
	No	11% (8/71)	Ref	Ref
Baseline binding antibody titers ≤ median (5.1 log-AU/mL)	Yes	12% (25/208)	2.07 (.82, 5.20)	2.14 (.57, 8.06)	.529
	No	6% (5/86)	Ref	Ref
	Missing	9% (7/75)	1.61 (.52, 4.99)	2.15 (.30, 15.50)
Relative cycle threshold (Crt) ≤ median (18.9 Crt)	Yes	14% (22/159)	2.25 (.97, 5.22)	1.89 (.93, 3.82)	.200
	No	6% (11/179)	Ref	Ref
	Missing	13% (4/31)	2.10 (.50, 8.75)	1.32 (.34, 5.04)
Presence of viral codetection	Yes	7% (9/133)	0.57 (.23, 1.43)	0.32 (.13, 0.79)	.013
	No	12% (28/236)	Ref	Ref
Viral detection ≥ 1 wk	Yes	29% (17/58)	4.56 (2.35, 8.83)	5.62 (2.62, 12.06)	<.001
	No	6% (20/311)	Ref	Ref
Asymptomatici	Yes	8% (2/25)	0.79 (.22, 2.75)	3.63 (.84, 15.68)	.084
	No	10% (35/344)	Ref	Ref
Any care seeking during illnessj	Yes	19% (10/53)	2.21 (.99, 4.93)	2.21 (.75, 6.48)	.150
	No	9% (27/316)	Ref	Ref
Any behavior to reduce transmission in the airk	Yes	9% (12/138)	0.80 (.36, 1.79)	0.50 (.15, 1.64)	.252
	No	11% (25/231)	Ref	Ref
Any behavior to reduce transmission on surfacesl	Yes	10% (13/125)	1.06 (.49, 2.29)	1.87 (.54, 6.50)	.326
	No	10% (24/244)	Ref	Ref
Characteristic of household contact
Age	6 mo–4 y	18% (4/22)	2.17 (.86, 5.51)	3.49 (1.30, 9.38)	.029
	5–12 y	13% (12/96)	1.49 (.83, 2.68)	2.41 (1.05, 5.50)
	13–50 y	8% (21/251)	Ref	Ref
Any comorbiditiesg	Yes	9% (14/151)	0.88 (.50, 1.54)	1.24 (.59, 2.59)	.573
	No	11% (23/218)	Ref	Ref
Vaccinated for influenza in 2021–2022 seasonh	Yes	9% (27/304)	0.58 (.28, 1.17)	0.36 (.15, .84)	.018
	No	15% (10/65)	Ref	Ref
Baseline binding antibody titers ≤ median (5.1 log-AU/mL)	Yes	13% (16/124)	1.54 (.80, 2.95)	1.45 (.77, 2.75)	.167
	No	8% (16/191)	Ref	Ref
	Missing	9% (5/54)	1.11 (.46, 2.64)	0.41 (.07, 2.45)

Abbreviations: CI, confidence interval; RSV, respiratory syncytial virus.

^a% of contacts infected = (# secondary RSV cases/# contacts exposed to a distinct index case).

^bP- value represents a grouped P- value for categorical variables with 3 levels.

^cHousehold density represents the number of household members regardless of enrollment in the study (thus, this may be greater than the number of individuals enrolled in the study if not all household members are enrolled).

^dOther housing type includes “apartment” and missing housing type.

^eDaycare or school attendance among at least 1 child <5 years in the household as reported at enrollment.

^fKPNW, Kaiser Permanente Northwest; UW, University of Washington.

^gComorbidities include asthma, chronic obstructive pulmonary disease (COPD) (including chronic bronchitis and emphysema), sleep apnea, heart disease, congenital heart disease, heart failure, down syndrome, hypertension (high blood pressure), diabetes (high blood sugar), liver condition, weak or failing kidneys, cancer or malignancy, arthritis, stroke, deep vein thrombosis (DVT) or pulmonary embolism (PE), sickle cell disease or thalassemia, weakened immune system, depression, anxiety, thyroid issues, or other health diagnosis.

^hVaccinated for influenza in 2021–2022 season versus no vaccination.

ⁱNo acute respiratory illness (ARI) symptoms within 7 days of detection.

^jSelf-reported care seeking defined as seeking health care (from a healthcare provider) during illness.

^kAny behavior change including masking, sleeping separate, covering cough/sneeze.

^lAny behavior change including handwashing, cleaning.

Figure 2.

Binding antibody titers and viral load, stratified by secondary RSV infection in household contacts (n = 369 household contacts of a distinct index case). A, Binding antibody titer of household contact at enrollment. B, Binding antibody titer of index case at enrollment. C, Relative cycle threshold (C_rt) of household contact at positive specimen collection. D, Relative cycle threshold (C_rt) of index case at positive specimen collection. Each data point (n) represents a single household contact with a distinct RSV index case in their household (ie, a contact-index pair). Multiple contacts can have the same index case in their contact-index pair, but each contact is represented by a distinct data point (ie, 4 contacts of a single index case would be represented by 4 data points). Panels (A) and (C) illustrate viral characteristics of the contact, whereas panels (B) and (D) present viral characteristics of the index case paired with each contact. Contacts with missing viral characteristics are excluded (ie, [A] 54 of 369 contacts are missing antibody titers, [B] 75 of 361 contacts of index cases are missing antibody titers, [C] 337 of 369 contacts are missing RSV C_rt, [D] 31 of 369 contacts of index cases are missing RSV C_rt). Abbreviation: RSV, respiratory syncytial virus.

Genomic Analysis

Sequence data were generated for 93 RSV A and 27 RSV B specimens collected from 27 September 2022 to 11 May 2023 (Figure 3). Main RSV A lineages represented were A.D.5.2 (n = 37), A.D.1 (n = 30), A.D.3 (n = 12), and A.D.2.1 (n = 6). Main RSV B lineages represented were B.D.E.1 (n = 24) and B.D.4.1.1 (n = 2). These lineages largely mirrored globally circulating lineages during the study period.

Figure 3.

Genetic diversity of RSV A (upper panels) and RSV B (lower panels) described by the distribution of circulating clades over the study period (panels A and C) and phylogenetic trees (panels B and D). In the trees, sequenced study specimens are indicated with green tip points, regional sequences (Washington/Oregon) are indicated with gray tip points, and other global sequences are indicated without tip points. Abbreviation: RSV, respiratory syncytial virus.

Among 10 participants with multiple sequenced viruses from unique nasal swabs, 1 individual had 2 different lineages and subtypes (B.D.E.1 and A.d.1) in specimens collected 196 days apart, indicating a new infection. One individual had 4 specimens with the same lineage (A.d.5.2) over 40 days, with 100% pairwise identity after masking low coverage regions, which suggests persistent shedding from a single infection. Of the remaining eight participants, each had two sequences of the same lineage 2–7 days apart (7 RSV A, 1 RSV B), with a median of 3 nucleotide differences (range 0–15 differences).

Of the sequenced specimens, 19 households (15 RSV A and 4 RSV B) had ≥2 RSV cases and were included in a further analysis of sequence similarity. Households included in this analysis had a median of 2 sequences (range 2–6 sequences) collected a median of 5 days apart (range 0–59 days). Within households with ≥2 RSV cases, sequenced viruses were more similar when collected with onset ≤14 days apart (mean pairwise difference 4 nucleotides [range 0–13], n = 17 households) versus >14 days after the index case (137 [37–236], n = 2) (Supplementary Figure 3). No sequenced specimens from household members with RSV detected ≤14 days apart were found to be of different clades (ie, there was no misclassification according to available genomic data), and thus we did not conduct a sensitivity analysis reclassifying the outcome based on genomics.

DISCUSSION

In this community-based prospective cohort with active in-home weekly respiratory illness surveillance, we found that 1 in 6 multi-person households with an RSV infection had secondary household transmission, with transmission occurring in 10% of household contacts. Young children 6 months to 12 years of age were the primary introducers of RSV into a household, and risk factors for secondary household transmission included prolonged viral detection of the index case, contact age ≤12 years, and childcare attendance among a child <5 years in the household. Genomic sequencing confirmed that household infections within 2 weeks were likely the result of within-household transmission.

The overall risk of secondary infection was similar to what was estimated in a South African cohort from 2017 to 2018 (11%) [3] but much lower than estimates from a household transmission study in Maryland, United States, during the post-pandemic RSV surge from 2020 to 2021 (44%) [1]. The majority of infected persons in our study reported symptoms, higher than previous reports of symptomatic infection in Kenya (58%) and South Africa (33%) [2, 3]. Over a range of time and geographic areas, studies have identified different household members most likely to introduce RSV into the household. In a recent US-based study, infants and children <5 years were most likely to introduce RSV [1], and in a pre-SARS-CoV-2 study in Kenya, school-aged children were most likely to introduce RSV [19]. We found that children ≤12 were most likely to introduce RSV into the household and spread to adult household members, consistent with the findings of both studies. In 2023, RSV imunizations were licensed for the protection of young infants and for older adults. Novel RSV prevention products in the development pipeline for children may ultimately protect these remaining vulnerable age groups, and may help decrease the risk of ongoing transmission.

Children ≤12 years were at a higher risk of secondary RSV infection compared to older household contacts when exposed to an index case, which is likely due to repeated exposures and household mixing patterns. We also found that prolonged viral detection in the index case was associated with higher risk of transmission. We did not find an association between index case viral load or serologic status of the index or secondary cases on risk of household transmission. Our weekly sampling cadence identified a distinct index case in 85% of incident household illness episodes; however, we were not able to capture the peak viral load of the illness episode to precisely measure the relationship between initial viral load and risk of onward transmission. Although daily sampling could improve the resolution of transmission dynamics, this would greatly increase study costs and may not be feasible for maintaining high household enrollment and participant retention.

The use of genomic sequencing enhances confidence in the risk of person-to-person transmission, particularly at times of high viral circulation when multiple introductions of RSV may occur in one household [20]. Genomic sequencing confirmed that secondary cases of RSV within 14 days of an index case in the same household were likely due to intrahousehold transmission. This finding provides a benchmark for interpretation for other epidemiologic studies of RSV household transmission that may not include genomic analysis. However, these findings are particularly important for designing studies of interventions, such as for antiviral therapy to prevent household transmission, because misclassification of secondary transmission has the potential to attenuate study results.

This study has several limitations. First, participant symptom and behavior data were self-reported and thus may be subject to recall or social desirability bias. Some RSV infections and thus index cases could be misclassified, because swabbing was not done in asymptomatic persons as frequently as symptomatic persons, who are likely to have had shorter periods of viral detection and test positivity [2]. In some households, not all members were enrolled leading to potential incomplete ascertainment of all household cases. Although we included all households in the primary analysis to ensure greater generalizability of our findings, we conducted sensitivity analyses including only households with complete ascertainment and found minimal impact on estimates. Future studies should prioritize data collection on non-enrolled household members (eg, ages, known ARI symptoms) where feasible. Baseline serum samples were collected at time points unrelated to the illness episode, and it is possible that the measured level was not reflective of their antibody titer at the time of infection. Additionally, the lag between specimen collection and sequencing of viral genomes prevents the use of sequencing data to disambiguate transmission patterns in real-time, but these data can provide valuable insights into geographic patterns of spread, number of introductions into a household or region, and identification of mutations and variants with transmission advantages that might drive spread. Only specimens with an adequate viral burden, represented by the RSV C_t or C_rt value, could be successfully sequenced, so our viral sequencing data represents viral genomes from persons with a higher RSV viral load. Lastly, the median household income reported by participants was above regional median estimates, which may limit generalizability of our findings.

CONCLUSIONS

This prospective community-based study provides critical data on RSV household transmission after COVID-19 pandemic mitigation measures were lifted and prior to the introduction of novel RSV vaccines and monoclonal antibodies. We observed that most index cases originated from children ≤12 years. Risk factors for secondary RSV transmission included index case with prolonged viral detection and/or contact age ≤12 years. Genomic sequencing demonstrated that multiple household infections within two weeks are likely due to within-household transmission. These findings can be used to identify target populations for novel RSV prevention measures and establish baseline RSV household transmission dynamics prior to the introduction of RSV immunizations.

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.

Notes

Present address. S. N. C. affiliation at the time this work was completed with the University of Washington. Current affiliation is Pfizer, Inc., New York, NY, USA.

Author Contributions. S. N. C. and H. Y. C. had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All co-authors reviewed and approved the manuscript prior to submission. Concept and design: S. N. C., P. R., T. M. B., M. C., J. A. E., J. L. K., A. L. N., A. A. W., H. Y. C. Acquisition, analysis, or interpretation of data: All. Drafting of the manuscript: S. N. C., P. R., M. C., A. A. W., H. N. C. Critical review of the manuscript for important intellectual content: All. Statistical analysis: S. N. C., C. F., K. L. H., M. C. RT-qPCR testing: L. M. S., C. M. L., P. D. H., S. G., A. A. W.

Acknowledgments. The authors would like to thank all CASCADIA study households for their time and active participation. University of Washington: Ariana Magedson, Denise McCulloch, Kyle Luiten, Devon McDonald, Jenni Logue, Jean Mernaugh, Melissa MacMillan, Mark Drummond, Grace Marshall, Anna Elias-Warren, Daniel Nguyen, Zarna Marfatia, Chidozie Iwu, Julia Bennett, Erica Clark, Jordan Opsahl, Jeremy Stone, David Reinhart, Ben Cappodano, Sarah Heidl, Alex Harteloo, Lani Regelbrugge, Madison Hollcroft, Leslie Rodriguez-Salas, Alex Harteloo; UW Virology team—Ailyn Perez, Alex Graninger, Hanna Edgar. Seattle Children's Hospital: Dallas Haws, Josh Sanders, Irem Onalan, Laura Ostrina Restrepo. Kaiser Permanente Center for Health Research: Michael Allison, Deralyn Almaguer, David Amy, Britt Ash, Kristi Bays, Tara Beatty, Kristin Bialobok, Allison Bianchi, Cathleen Bourdoin, Delanie Brown, Stacy Bunnell, Joseph Cerizo, Evelin Coto, Phil Crawford, Robin Daily, Lantoria Davis, Kendall Frimodig, Stephen Fortmann, Lisa Fox, Kenni Graham, Tarika Holness, Matt Hornbrook, Emily Jubitz, Terry Kimes, Keelee Kloer, Dorothy Kurdyla, Isaiah Lankham, Teri Lawer, Caroline Lee, Max Lin, Richard Martin, Bryony Melcher, John Ogden, Aaron Piepert, Joanne Price, Angela Reyes-Ochoa, Jennifer Rivelli, Sperry Robinson, Katrina Schell, Emily Schield, Meagan Shaw, Anna Shivinsky, Nina Shockman, Ellen Sullivan, Martin Simer, Valencia Smith, Senait Tadesse, Britta Torgrimson-Ojerio, Meredith Vandermeer, Brooke Wainwright, Mica Werner, Danika Whitcomb. Kaiser Permanente Washington Heath Research Institute: Serah Kimachia, Rebecca Ziebell, Chester Pabiniak, Kristin Delaney. Vanderbilt University Medical Center: Sydney (Afan) Swan, Onika Abrams, Theresa (Terri) Scott.

Data availability. Data available on NCBI Genbank (Bioprojects PRJNA988152: RSV-A, PRJNA988154: RSV-B).

Financial support. This work was supported by the Centers for Disease Control and Prevention (contract number 75D30121C12297 to Kaiser Foundation Hospitals). The funders were not involved in the design of the study and do not have any ownership over the management and conduct of the study, the data, or the rights to publish. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.