Respiratory Syncytial Virus–Associated Hospitalizations Among Children Less Than 24 Months of Age
BACKGROUND: Respiratory syncytial virus (RSV) infection is a leading cause of hospitalization among infants. However, estimates of the RSV hospitalization burden have varied, and precision has been limited by the use of age strata grouped in blocks of 6 to ≥12 months.
METHODS: We analyzed data from a 5-year, prospective, population-based surveillance for young children who were hospitalized with laboratory-confirmed (reverse-transcriptase polymerase chain reaction) RSV acute respiratory illness (ARI) during October through March 2000–2005. The total population at risk was stratified by month of age by birth certificate information to yield hospitalization rates.
RESULTS: There were 559 (26%) RSV-infected children among the 2149 enrolled children hospitalized with ARI (85% of all eligible children with ARI). The average RSV hospitalization rate was 5.2 per 1000 children <24 months old. The highest age-specific rate was in infants 1 month old (25.9 per 1000 children). Infants ≤2 months of age, who comprised 44% of RSV-hospitalized children, had a hospitalization rate of 17.9 per 1000 children. Most children (79%) were previously healthy. Very preterm infants (<30 weeks’ gestation) accounted for only 3% of RSV cases but had RSV hospitalization rates 3 times that of term infants.
CONCLUSIONS: Young infants, especially those who were 1 month old, were at greatest risk of RSV hospitalization. Four-fifths of RSV-hospitalized infants were previously healthy. To substantially reduce the burden of RSV hospitalizations, effective general preventive strategies will be required for all young infants, not just those with risk factors.
- ARI —
- acute respiratory illness
- CI —
- confidence interval
- ICD-9-CM —
- International Classification of Diseases, Ninth Revision, Clinical Modification
- IRR —
- incidence rate ratio
- RSV —
- respiratory syncytial virus
What’s Known on This Subject:
Respiratory syncytial virus (RSV) infection is a leading cause of hospitalization among infants. Most estimates of RSV hospitalization rates are imprecise, having been calculated by using retrospective discharge diagnosis data and stratified age groups no narrower than 6 to 12 months.
What This Study Adds:
Prospective, population-based surveillance data for infants hospitalized with laboratory-confirmed RSV infection were combined with birth certificate information to yield more precise age-specific hospitalization rates. These data should help determine priorities for the use of existing and future RSV prophylaxis strategies.
Respiratory syncytial virus (RSV) infection is recognized as a leading cause of hospitalization among infants.1–6 Accurate assessment of the burden of RSV hospitalizations, therefore, is essential in determining priorities for the use of existing RSV prophylaxis and future vaccines. However, estimates of the burden of RSV hospitalizations in the United States have varied.
Retrospective analyses using national databases and International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM)–coded discharge diagnoses to determine RSV hospitalizations have estimated annual hospitalization rates of 30 to 50 per 1000 infants <6 months old, 11 to 15 per 1000 infants 6 to <12 months old, and 15 to 26 per 1000 infants <12 months old.1–6 However, substantially lower rates, approximately one-half lower, are reported by the few available prospective, population-based studies of laboratory-confirmed cases.7,8
Our goal, therefore, was to more accurately define and characterize RSV hospitalizations among children <24 months old, the group at greatest risk of severe RSV illness. We used prospective, population-based surveillance data for laboratory-confirmed RSV hospitalizations over a 5-year period and verified the denominator-based data by reviewing birth certificates to provide monthly age strata.
We previously described the clinical burden of inpatient and outpatient RSV infections among children <5 years old during the 4 respiratory seasons of 2000–2001 through 2003–2004 using census data for the denominators of the at-risk population.7 The current 5-year study includes the subset of inpatients <24 months old from the respiratory seasons (October through March each year) reported in that study but includes additional inpatients <24 months old enrolled during the 2004–2005 respiratory season. Importantly, the denominators of the at-risk population in this report were defined more precisely (by each month of age) by reviewing birth certificate data for the studied cohorts. Children with RSV infection were prospectively identified by surveillance for laboratory-confirmed RSV infections through the New Vaccine Surveillance Network, as previously described.7–10
Viral surveillance was conducted from October through March during the years 2000 through 2005 in counties surrounding Nashville, Tennessee (Davidson County), and Rochester, New York (Monroe County), and during 2003–2005 in Cincinnati, Ohio (Hamilton County). During the respiratory seasons, children with acute respiratory illness (ARI) were enrolled 4 to 7 days per week, as previously described.7–10 Surveillance hospitals admitted >95% of each county’s children.
Eligible children were county residents <24 months old with an ARI defined as acute illness onset with ≥1 of the following: fever or respiratory tract signs, including cough, earache, nasal congestion, rhinorrhea, wheezing, or rapid or labored respirations.8
Institutional review board approvals were obtained from the Centers for Disease Control and Prevention and from each study site’s institution; informed consent for enrollment of all children was obtained from a parent or legal guardian.
Patient and Laboratory Data
The child’s gender, race, gestational age at birth, presence of other children in the home, and palivizumab administration were obtained from parental or guardian interviews. Comorbid conditions (chronic pulmonary, cardiac, kidney, or neurologic diseases or immunodeficiency, cancer, or sickle cell anemia) and insurance information were determined by medical record review. Children were defined as having RSV infection if the nasal/throat sample performed in duplicate by reverse-transcriptase polymerase chain reaction was RSV-positive.7–10
Birth Certificate Data
The monthly denominator of live births for chronologic and gestational ages, maternal race, and gender was calculated from each county’s birth certificates, and age cohorts were determined. Children with gestational ages of <30 weeks were defined as very preterm, 30 to 33 weeks as early preterm, 34 to 36 weeks as late preterm, and ≥37 weeks as term.11
χ2 Tests compared the frequencies of characteristics of inpatients with RSV-positive and RSV-negative ARIs.
Monthly incidence rates of RSV hospitalizations were calculated by dividing the weighted number of RSV cases, accounting for both the number of eligible children enrolled and surveillance days, by the total number of children within each birth cohort. Seasonal average monthly incidences were calculated by combining the birth cohorts using direct standardization. The cumulative seasonal incidence rate was the summed monthly rates.
Population-based hospitalization rates were calculated for subgroups as defined by birth certificate information, as well as rates by county, age, and month of admission. Subgroups were compared by using incidence rate ratios (IRRs). To describe the variability around the rates and rate ratios, bias-corrected bootstrap 95% confidence intervals (CIs) were calculated from 10 000 replications, and bootstrapped SEs were used to determine P values.
Analyses were performed with Stata/MP version 12.1 (StataCorp LP, College Station, TX).
Of 132 085 children born in the surveillance area during the study period, 2539 had an ARI hospitalization before reaching 24 months of age during the surveillance period and were eligible for inclusion. Of these, 2149 (85%) were enrolled, and 559 (26%) enrolled children had laboratory-confirmed RSV infection. The major reasons for nonenrollment were parental refusal (46%) or unavailability (30%) or discharge before enrollment (15%). Less than 1% of the children’s physicians declined participation.
The characteristics of children hospitalized with RSV-positive ARIs were similar to those with ARIs from other causes, except a higher proportions of RSV-infected children were <12 months old and were white (Table 1).
The mean overall RSV season duration from the first to the last RSV-positive sample for all years and sites was 28 weeks. For Davidson, Monroe, and Hamilton Counties, it was 30, 26, and 28 weeks, respectively. However, nearly 80% of RSV infections occurred over 14 weeks beginning in early December.
Average Seasonal RSV Hospitalization Rates Among Subsets of Young Children
The average rate of RSV admissions over the 6-month respiratory season was 5.2 per 1000 (95% CI: 4.8–5.7) children <24 months old, with significant seasonal variation from 2.5 to 9.9 per 1000 children (P < .01; data not shown). One-month-old infants had the highest average hospitalization rate, 25.9 per 1000 (95% CI: 21.3–30.8) infants (Table 2), with rates ranging by season from 8.6 to 37.0 per 1000 infants. Infants who were 2 months old and infants <1 month old had the next highest average hospitalization rates, with wide seasonal ranges: 14.3 (95% CI: 11.1–17.8) and 13.5 (95% CI: 10.3–17.1) per 1000 infants, respectively (Table 2). Overall, infants ≤2 months old were the most likely to be hospitalized, with an average rate of 17.9 (95% CI: 15.7–20.1) per 1000 infants.
Over the next 3 months of life the average seasonal rates of RSV hospitalization consistently declined from 10.3 per 1000 infants who were 3 months old to 4.8 per 1000 infants who were 5 months old (Table 2). The average hospitalization rate for all 3- to 5-month-old infants was 8.0 (95% CI: 6.6–9.5), and for all 6- to 11-month-old infants was 3.9 (95% CI: 3.2–4.7), which was not significantly different than any of the individual rates for each month of age between 6 and 11 months of age.
Infants who were ≤2 months old were hospitalized more than twice as often as infants who were 3 to 5 months old (IRR: 2.2; 95% CI: 1.8–2.8) and 4.5 times more often than infants who were 6 to 11 months old (IRR: 4.5; 95% CI: 3.6–5.8). Of all children hospitalized with RSV, 44% were ≤2 months old, 64% were <6 months old, and 83% were <12 months old.
County and Calendar Month
The average seasonal hospitalization rates per 1000 children <24 months old varied among the 3 counties. The rates of Monroe County (6.5; 95% CI: 5.8–7.3) and Hamilton County (6.2; 95% CI: 5.2–7.2) were significantly higher than Davidson County’s rate (4.0; 95% CI: 3.4–4.6) (IRR: 1.6; 95% CI: 1.2–2.0) (Supplemental Fig 2)
The calendar month also significantly influenced RSV hospitalizations. Average monthly hospitalization rates per 1000 children <24 months old were highest in December (1.0; 95% CI: 0.8–1.2), January (1.7; 95% CI: 1.5–1.9), and February (1.2; 95% CI: 1.0–1.4). During these 3 months, 73% of RSV hospitalizations for all children <24 months old occurred, 19% in December, 31% in January, and 23% in February. Most RSV hospitalizations occurred in December in Hamilton County (27%) and in January in Davidson and Monroe Counties (27% and 37%, respectively).
Of RSV-hospitalized children, 10% were preterm, of whom 38% had an additional comorbid condition beyond prematurity. For all preterm infants (<37 weeks’ gestation), the average seasonal RSV hospitalization rate was 4.6 (95% CI: 3.4–5.8) per 1000 children, which was not significantly different from that for term infants, which was 5.3 (95% CI: 4.9–5.8) per 1000 children (IRR: 0.9; 95% CI: 0.6–1.1). However, the few very preterm infants born at <30 weeks gestation (3%) had the highest RSV hospitalization rate (18.7 per 1000 children), which was significantly greater than that for both early-preterm (30–33 weeks’ gestation) and late-preterm (34–36 weeks’ gestation) infants, as well as for full-term infants (IRR for very preterm and early-preterm infants: 2.7; 95% CI: 1.3–5.8) (Fig 1). Late-preterm infants, however, were hospitalized significantly less often than term infants (IRR: 0.5; 95% CI: 0.3–0.7).
Among all preterm infants hospitalized with RSV, 23% received palivizumab (Table 1). When analyzed according to the gestational age groups used for palivizumab recommendations by the American Academy of Pediatrics12 (Supplemental Table 3), the hospitalization rate of all infants of 32 to 34 weeks’ gestation (regardless of palivizumab usage) was 6.9 (95% CI: 4.3–10.1) per 1000 compared with that of all infants ≥35 weeks’ gestation, which was 5.1 (95% CI: 4.7–5.5) per 1000 (IRR: 1.4; 95% CI: 0.8–2.0).
Gender and Race
The overall seasonal rates of RSV hospitalization among all children <24 months old were not significantly affected by gender or race (Supplemental Fig 3). Hospitalization rates per 1000 children <24 months old were 5.4 (95% CI: 4.6–6.3) for black children and 4.7 (95% CI: 4.2–5.1) for white children (IRR: 1.2; 95% CI: 1.0–1.4). Among infants <6 months old, the hospitalization rates for black and white children were not significantly different, but among children ≥6 months old black children were hospitalized significantly more often (IRR: 1.9; 95% CI: 1.4–2.6).
Overall, 21% of children who were <24 months old hospitalized with RSV infection had a comorbid condition. The most frequent comorbid condition among RSV-positive children was cardiopulmonary disease. The proportion of RSV-hospitalized children with comorbid conditions varied with age. Coexisting medical conditions were present more frequently among older children. Among 12- to 23-month-old RSV-hospitalized children, 53% had a comorbid condition, whereas among infants ages <6 months and 6 to 11 months, the proportions were only 7% and 39%, respectively. Thus, the proportion of the 12- to 23-month-old children with comorbid conditions was >7 times greater than that of infants <6 months old and 1.4 times greater than that of infants 6 to 11 months old. Similarly, 12- to 23-month-old RSV-hospitalized children had 9- and 1.4-fold greater proportions of cardiopulmonary comorbidities than did infants who were <6 months old and those who were 6 to 11 months old, respectively.
Other Children in Home
Among all children <24 months old who were hospitalized with RSV, 76% lived in homes with ≥1 other children of any age; 57% lived with another child <5 years old, and 19% lived with another child 5 to <18 years old. Among RSV-hospitalized infants who were <12 months old, 58% lived with another child <5 years old, and 20% lived with another child 5 to <18 years old.
The average seasonal RSV hospitalization rate in this study was 5.2 per 1000 children who were <24 months old, but the rate varied by season as much as fourfold. Nevertheless, 1-month-old infants consistently were most likely to be hospitalized, almost twice as often as the next 2 most at-risk groups: infants <1 month old and infants 2 months old. These youngest infants accounted for an important proportion of all children admitted with RSV infection in the first 2 years of life: 11% were infants <1 month old, 44% were ≤2 months old, and only 36% were >5 months old.
The appreciable, but generally unrecognized, impact of neonatal RSV infection is important in planning strategies to control RSV infection because of both the financial health care burden and the hurdles inherent in developing safe and effective vaccines and therapeutic measures that must be administered within the first few weeks of life. Although recent studies, most of which were retrospective assessments, have consistently shown RSV hospitalizations to be the highest in the first year of life, very few have dissected the exact age of RSV-specific hospitalizations more precisely.4–6,13–20 For example, a retrospective, population-based study from Spain found that among children <24 months old hospitalized with laboratory-identified RSV, 9% were <1 month of age, similar to our findings; however, their rate of RSV hospitalizations among infants <1 month of age was more than twice ours (31 per 1000 vs 13.5 per 1000, respectively).19 Another study estimated a notably high annual rate of RSV-associated hospitalizations among infants ≤3 months old, 48.9 per 1000 children, which is more than twice the rate we found.6 The differences between studies could result from actual differences in disease burden or from the use in database studies of ICD-9-CM–coded discharge diagnoses for RSV-associated syndromes such as bronchiolitis, because they are inaccurate indicators of RSV infection among young infants.
Race has been correlated with more severe viral infections, especially influenza.21,22 For RSV infection, information is limited, but the reported severity of disease among black versus white children has been variable and inconclusive.14,22,23 The rates of RSV hospitalization among black children <24 months old or <6 months old were not significantly different from those for white children. However, after 6 months of age, black children were hospitalized nearly twice as often. This finding suggests that the decision to hospitalize young infants is strongly affected by their age, and perhaps also by other unidentified factors (eg, socioeconomic factors). We found no differences in RSV hospitalization rates by gender in this study or in previous work.7,22
The reported proportions of children with chronic conditions hospitalized with RSV have ranged widely, possibly reflecting the variable underlying populations and study designs.6,15,17,24 Stockman et al6 recently found that 12% of RSV-hospitalized children <12 months old and 2.4% of those 12 to 23 months old had heart disease, chronic neonatal respiratory distress, or prematurity, as determined by RSV ICD-9-CM–coded discharge diagnoses. In comparison, we found that 21% of RSV-hospitalized children <12 months old and 53% of children 12 to 23 months old had chronic conditions, as determined by medical chart review. Again, this difference could be explained by the inaccuracy of RSV-specific diagnostic codes, the increased accuracy of our laboratory-confirmed diagnoses, or by our chart review identifying comorbidities not captured by ICD-9-CM codes. Increased palivizumab use early in life could reduce the proportion of children with comorbidities among all RSV hospitalizations during the first year of life compared with the second year of life, as found by Stockman et al6 and in our study.
Prematurity has been consistently associated with severe RSV infection, but the degree of risk has varied.14,15,17,25 Of our children hospitalized with RSV, 10% were preterm, but their risk of hospitalization was not significantly different from that for term infants. We did find lower rates of hospitalization among late-preterm infants, and higher rates among very premature infants, in contrast to other studies.12,15 Variability in hospitalization rates among early-preterm, late-preterm, and term infants may primarily result from premature infants in certain gestational age ranges being too few to reliably calculate hospitalization rates by gestational weeks. Perhaps the care patterns or environmental risks are different for late-preterm infants as well. Nevertheless, the effect of preterm infants on our overall RSV hospitalization rates and their impact on general health care issues are limited by prematurity being relatively infrequent among the general population. The current national population estimate for premature birth at <37 weeks’ gestation is 12%, which is similar to what we found among children hospitalized with ARI (11%); the national estimate for premature birth at <30 weeks’ gestation is 1%.26
Many studies have examined environmental factors associated with an increased risk of children acquiring RSV disease, especially contact with other children.12,14,16,17,27,28 A recent review examining the effect of residential crowding on RSV-associated hospitalizations among young children found significant correlations existed with severe disease in studies reporting laboratory-confirmed RSV infections.27 However, the effect varied according to age and multiple factors associated with crowding.14,17,29 In our study, the proportion of hospitalized children residing in homes with other children <5 years old was more than twice that of children living with older children.
Our study has several potential limitations. The 3 counties, although geographically diverse, may not be representative of the entire United States. In addition, our data do not account for local differences in medical practices that may affect RSV hospitalization rates. In addition, we did not study the effect of the recommended and nonrecommended use of palivizumab on our RSV hospitalization rates.12,30 However, palivizumab’s effect on our overall rates of RSV hospitalization was unlikely to be appreciable, because only a small proportion (<5%) of our study population was eligible for palivizumab, and its use was variable (30% to 70%) during most study years.30 Day care attendance has also been described as a risk factor for RSV infection,12 although inconsistently7,22; the current work is limited by not analyzing this variable.
Strengths of our study include the population-based, prospective, laboratory-confirmed surveillance, medical chart review, and unlike our previous study,7 the use of birth certificates to better quantify the population at risk by individual months of age rather than as blocks of age groups. Taken together, our findings suggest that the children most likely to be hospitalized with RSV infection are those infants ≤2 months of age (especially those 1 month of age), those living with other young children, and those born during the peak months of RSV circulation, which in our 3 counties were December and January.
These findings indicate that strategies for diminishing the health care burden from RSV infections should include appropriate prophylaxis and the development of vaccines that are effective in very young infants, even those within the first month of life. In addition, general infection control practices such as restrictions of visits from ill individuals and careful hand washing should be emphasized, especially during the peak months of RSV circulation. In addition, these preventive measures should not be restricted to high-risk infants, but rather targeted to all infants, because most of the affected infants were previously healthy and born at full term. Until such time that universal RSV immunization becomes possible, our prospective, population-based surveillance data for laboratory-confirmed RSV hospitalizations stratified by month of age may assist in prioritizing preventive measures.
We thank the children and their parents who participated in this study; all the members of the New Vaccine Surveillance Network, including Geraldine Lofthus, Kenneth Schnabel, Lynne Shelley, Jennifer Carnahan, Linda Anderson, and Christina Albertin at the University of Rochester; Diane Kent, Erin Keckley, and Ann Clay at Vanderbilt University; Ardythe Morrow, Marilyn Rice, Linda Jamison, Vanessa Florian, Emily Grube, Joel Mortensen, David Witte, Cindy Ventrola, and Pam Groen at Cincinnati Children’s Hospital Medical Center; and Aaron Curns, Ranee Seither, Benjamin Schwartz, Frances Walker, John Copeland, Dean Erdman, and James Alexander at the Centers for Disease Control and Prevention.
This article is dedicated to the memory of Caroline Breese Hall, MD, a true leader in RSV clinical research, a respected mentor for many trainees and faculty alike, and a valued colleague to the authors.
- Accepted May 16, 2013.
- Address correspondence to Geoffrey A. Weinberg, MD, Department of Pediatrics, University of Rochester School of Medicine and Dentistry, PO Box 690, 601 Elmwood Ave, Rochester, NY 14642. E-mail:
Dr Hall is deceased.
Drs Hall, Weinberg, Szilagyi, and Iwane conceptualized and designed the study; acquired, analyzed, and interpreted the data; drafted the manuscript and critically revised it for important intellectual content; obtained funding; provided administrative, technical, and material support; and supervised the study; Mr Blumkin acquired, analyzed, and interpreted the data; critically revised the manuscript for important intellectual content; and performed critical statistical analyses of the data; Drs Edwards, Staat, and Griffin conceptualized and designed the study; acquired, analyzed, and interpreted the data; drafted the manuscript and critically revised it for important intellectual content; obtained funding; and supervised the study; Mr. Schultz acquired, analyzed, and interpreted the data; critically revised the manuscript for important intellectual content; performed critical statistical analyses of the data; and provided administrative, technical, and material support; Dr Poehling conceptualized and designed the study; acquired, analyzed, and interpreted the data; drafted the manuscript and critically revised it for important intellectual content; and supervised the study; Dr Williams conceptualized and designed the study; acquired, analyzed, and interpreted the data; critically revised the manuscript for important intellectual content; and supervised the study; Drs Zhu and Grijalva acquired, analyzed, and interpreted the data; performed critical statistical analyses of the data; and critically revised the manuscript for important intellectual content; and Ms Prill acquired, analyzed, and interpreted the data; performed critical statistical analyses of the data; critically revised the manuscript for important intellectual content; and provided administrative, technical, and material support. All authors 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.
The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the Centers for Disease Control and Prevention.
FINANCIAL DISCLOSURE: Dr Hall has consulted for GlaxoSmithKline (GSK) and MedImmune and has received research support from MedImmune. Dr Weinberg has served on speakers’ bureaus for GSK, Merck, and Sanofi Pasteur. Dr Edwards has received research funding from Novartis for nonrelated vaccine trials. Dr Staat has received research support from GlaxoSmithKline, Merck, and MedImmune; has consulted for GSK, Merck, and MedImmune; and has been on the speakers’ bureaus for GSK and Merck. Dr Poehling has received research support from BD Diagnostics. Dr Griffin has consulted for Novavax and MedImmune. Dr Williams serves on the Scientific Advisory Board for Quidel. The other authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Supported by the Centers for Disease Control and Prevention (cooperative agreements U38 CCU217969, U01 IP000017, U38 CCU417958, U01 IP000022, U38 CCU522352, and U01 IP000147).
- Iwane MK,
- Edwards KM,
- Szilagyi PG,
- et al.,
- New Vaccine Surveillance Network
- Engle WA,
- Tomashek KM,
- Wallman C,
- Committee on Fetus and Newborn, American Academy of Pediatrics
- Committee on Infectious Diseases
- ↵Zhou H, Thompson WW, Viboud CG, et al. Hospitalizations associated with influenza and respiratory syncytial virus in the United States, 1993-2008. Clin Infect Dis. 2012;54(10):1427–1436
- ↵Sommer C, Resch B, Simoes EA. Risk factors for severe respiratory syncytial virus lower respiratory tract infection. Open Microbiol J. 2011;5(Suppl 2):144–154
- Iwane MK,
- Chaves SS,
- Szilagyi PG,
- et al
- ↵Kristensen K, Hjuler T, Ravn H, Simoes EA, Stensballe LG. Chronic diseases, chromosomal abnormalities, and congenital malformations as risk factors for respiratory syncytial virus hospitalization: a population-based cohort study. Clin Infect Dis. 2012;54(6):810–817
- ↵Rossi GA, Silvestri M, Lanari M. Should the American Academy of Pediatrics respiratory syncytial virus guidelines be modified? Pediatrics. 2010;125(4):e1021; author reply e1022
- ↵Centers for Disease Control and Prevention, National Center for Health Statistics, Division of Vital Statistics. Natality public-use data 2007-2009, on CDC WONDER online database, March 2012. Available at: http://wonder.cdc.gov/natality-current.html. Accessed August 7, 2012
- Lanari M,
- Giovannini M,
- Giuffré L,
- et al.,
- Investigators R.A.DA.R. Study Group
- Law BJ,
- Langley JM,
- Allen U,
- et al
- Copyright © 2013 by the American Academy of Pediatrics