- Email: [email protected]
Serious Early Childhood Wheezing After Respiratory Syncytial Virus Lower Respiratory Tract Illness in Preterm Infants
Clinical Therapeutics/Volume 32, Number 14, 2010
Serious Early Childhood Wheezing After Respiratory Syncytial Virus Lower Respiratory Tract Illness in Preterm Infants José R. Romero, MD1; Dan L. Stewart, MD2,3; Erin K. Buysman, MS4; Ancilla W. Fernandes, PhD5; Hasan S. Jafri, MD6; and Parthiv J. Mahadevia, MD, MPH5 1
Department of Pediatrics, Arkansas Children’s Hospital and University of Arkansas for Medical Sciences, Little Rock, Arkansas; 2Department of Pediatrics, University of Louisville Hospital, Louisville, Kentucky; 3 Kosair Children’s Hospital, Louisville, Kentucky; 4Health Economics and Outcomes Research, i3 Innovus, an Ingenix Company, Eden Prairie, Minnesota; 5Health Outcomes and Pharmacoeconomics, MedImmune, LLC, Gaithersburg, Maryland; and 6Clinical Development, MedImmune, LLC, Gaithersburg, Maryland ABSTRACT Background: Respiratory syncytial virus (RSV) lower respiratory tract infection (LRI) in early life has been associated with sustained airway hyperreactivity during childhood; however, corresponding data in premature infants are sparse. Objective: The objective of this study was to determine whether RSV-LRI during early infancy of preterm infants was associated with an increased risk for serious early childhood wheezing (SECW) by age 3 years. Methods: A retrospective cohort study was conducted using data from a large (⬃14 million members) US health plan database. The study population included infants ⱕ6 months of age born at ⱕ36 weeks’ gestational age or weighing ⬍2500 g, or both. Preterm infants with any medically attended RSV-LRI from May 2001 through April 2004 with 3 years of continuous eligibility were selected and propensity matched with ⱕ3 control infants. SECW was defined as ⬎3 office, outpatient, or emergency department (ED) visits with asthma or wheezing; ⱖ1 office, outpatient, or ED visit with asthma or wheezing plus treatment with systemic corticosteroids within 7 days; ⱖ1 inpatient stay with asthma or wheezing; or ⱖ150 days’ supply of asthma-control medications. The presence of SECW between ages 2 and 3 years was compared between infants with and without RSV-LRI using univariate and multivariate methods. Health care costs for patients with SECW were explored. Results: A total of 378 infants with RSV were matched to 606 controls. The prevalence of SECW between ages 2 and 3 years was 16.7% in the RSV-LRI group versus 8.6% in the control group (P ⬍ 0.001).
2422
Logistic regression showed that preterm infants with RSV in early life were 2.52-fold (95% CI,1.65–3.85) more likely to present with SECW between ages 2 and 3 years (P ⬍ 0.001). Patients with SECW had a mean SECW-related cost of US $1378 (95% CI, $939 –$1816) and total health care cost of $7138 (95% CI, $5087– $9189) compared with $37 (95% CI, $24 –$51) and $2521 (95% CI, $1789 –$3253), respectively, for patients without SECW. After adjusting for possible confounders, patients with SECW had a significantly higher total health care cost than did patients without evidence of SECW (P ⬍ 0.001). Conclusions: The development of RSV-LRI in infancy in preterm infants was associated with an increased prevalence of SECW between ages 2 and 3 years. Patients with SECW had higher total health care costs than those who did not have SECW. (Clin Ther. 2010;32:2422–2432) © 2010 Elsevier HS Journals, Inc. Key words: infants, outcome measures, premature, RSV, wheezing.
INTRODUCTION Recognized as the leading cause of significant lower respiratory tract disease in infants and children, respiThis study was presented as a poster at the American Academy of Pediatrics National Conference and Exhibition, October 17–20, 2009, Washington, DC. Accepted for publication November 4, 2010. doi:10.1016/j.clinthera.2011.01.007 0149-2918/$ - see front matter © 2010 Elsevier HS Journals, Inc. All rights reserved.
Volume 32 Number 14
J.R. Romero et al. ratory syncytial virus (RSV) accounts for 50% to 80% of cases of bronchiolitis and 30% to 60% of pneumonia admissions in children ⬍5 years of age during the winter season in the United States.1-6 RSV bronchiolitis was the leading cause of hospital admissions of infants aged ⬍1 year between 1997 and 1999 in the United States.7 Studies have reported that acute RSV-lower respiratory infection (LRI) in infancy may lead to long-term airway morbidity in the form of recurrent wheezing or asthma,8,9 and that airway reactivity rates can exceed 50% after an RSV infection.10 Gestational age (GA) may affect the development of airway reactivity,10 –12 and premature infants have been documented to be at high risk for serious or fatal RSV infection.13 However, limited studies have focused on the incidence of early childhood wheezing in premature infants infected with RSV.14,15 The reported prevalence and duration of the wheezing vary widely; epidemiologic studies have reported that wheezing after RSV-LRI occurs in 16% to 71% of patients.8,14,16-18 In their population-based birth cohort study, Henderson et al16 reported a cumulative prevalence of wheezing of 28.1% in the RSV group and 13.1% in the control group at 30 to 42 months. At 69 to 81 months, the prevalence of wheezing remained 22.6% in the RSV group and 9.6% in the control group. At 91 months, this study found a cumulative prevalence of asthma of 38.4% in the RSV group and 20.1% in the control group. A prospective study by Sigurs et al17 reported that the prevalence of asthma (30%) and wheezing (38%) was significantly higher in children aged ⱕ7 years who had been hospitalized as infants for RSV bronchiolitis than in a control population (3% and 2%, respectively). The data from patients aged 13 years18 supported RSV bronchiolitis in infancy as an independent risk factor for asthma and recurrent wheezing, with 43% of the RSV group reporting asthma or recurrent wheezing versus 8% of the control group. Stein et al19 found that outpatient-diagnosed RSV-LRI was associated with an increased risk for infrequent wheeze (odds ratio [OR] ⫽ 3.2) and frequent wheeze (OR ⫽ 4.3) at age 6 to 11 years. The risk decreased markedly with age and was no longer significant at age 13 years. Recurrent or frequent wheezing has been used as an outcome in several prospective studies14,17–20 as a way to capture airway disease that has a greater risk for adverse effects on a child’s health. A diagnosis of
December 2010
asthma is recognized to be clinically important, as its associated symptoms and effects on a child’s health are better understood. However, diagnosing asthma in children ⬍5 years of age is difficult because there is not a consensus on objective measures for asthma diagnosis and management guidelines.21 However, outcomes relying only on the frequency of symptoms may miss less frequent but severe or life-threatening events that require long-term control medications or hospitalization. Therefore, serious early childhood wheezing (SECW) has been proposed as an outcome that captures not only frequency but also severity of disease by linking wheezing to the use of specific medications, such as systemic steroids or asthma-control medication, or both. The objective of this study was to assess the association of SECW occurring between ages 2 and 3 years after RSV-LRI in early infancy in a cohort of preterm infants infected during their first RSV season. In a subset of infants in whom GA codes were available, it was determined whether GA modified this association between RSV-LRI and SECW. Similarly, in a subset of infants in whom a family history of atopy or asthma could be established, it was determined whether these factors modified the association between RSV-LRI and SECW. The financial impact of SECW in terms of direct costs, including pharmacy costs, was explored.
PATIENTS AND METHODS Study Design A retrospective cohort study was conducted using eligibility, medical, and pharmacy administrative claims data from a large US managed care organization (MCO). Eligible enrollees were from a commercially insured preferred-provider organization model of a national MCO consisting of ⬃14 million members with medical and pharmacy benefits during 2007. The MCO provided full insurance coverage for physician, hospital, and pharmacy services. Individuals covered by the health plan were from geographically diverse regions of the United States.
Study Population Infants born May 2001 through April 2004 were identified from the MCO claims data. Patients were included in this study if they met the following criteria: (1) had at least one medical claim with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM)22 diagnostic code for pre-
2423
Clinical Therapeutics maturity, defined as ⬍36 weeks GA (codes 765.21– 765.28) or low birth weight, defined as ⬍2500 g (codes 765.0x, 765.1x, and V21.30 –V21.35) in either the primary or secondary position on physician or hospital claims; (2) were commercially insured with a medical and pharmacy benefit; and (3) had ⱖ36 months of continuous enrollment in the health plan after birth. Infants with claims designated for light for date (ICD9-CM code 764.XX) or heavy for date (code 766.XX) were excluded from the study to decrease heterogeneity in the selected cohort. Patients were followed up from their date of birth through their third birthday. Each infant was assigned to the seasonal year of birth based on date of birth. A seasonal year of birth was defined as May 1 through April 30 of the following year, allowing each study year to capture an entire RSV season. This is in contrast with using the calendar year, which would have captured half of 2 different RSV seasons. Study patients were classified as with or without evidence of an RSV event during their first RSV season (ie, November 1–April 30 of the infant’s seasonal year of birth; peak RSV activity normally occurs between November and March in the temperate climates of North America).23 An RSV event was defined as either (1) a claim for an RSV condition (ICD-9-CM code 079.6, 466.11, or 480.1 in either the primary or secondary position) during an inpatient, emergency department (ED), or ambulatory visit; or (2) a claim for an RSV-like condition (code 466.0, 466.19, 480.9, 485, or 486 in either the primary or secondary position) during an inpatient, ED, or ambulatory visit, with no evidence of claims for influenza or bacterial pneumonia (code 481, 482.XX, or 487.X) within 3 days of the RSV-like claim.
Study Outcomes The primary objective of this study was to determine the rate of clinically significant childhood wheezing in premature or low-birth-weight patients with and without evidence of RSV infection during their first RSV season. After consultation with physicians and researchers with expertise in pulmonology, infectious diseases, immunology, allergy, neonatology, general pediatrics, and health outcomes, SECW between the second and third birthdays was proposed as a measure of clinically meaningful childhood wheezing. To be classified as having SECW, patients were required to meet at least one of the following criteria between their
2424
second and third birthdays: (1) ⱖ3 office, outpatient, or ED visits with a diagnosis of asthma or wheezing (ICD-9-CM code 493.XX or 786.07); (2) ⱖ1 office, outpatient, or ED visit with a diagnosis of asthma or wheezing with a claim for systemic (oral or intravenous) corticosteroids within 7 days of the asthma or wheezing claim; (3) ⱖ1 inpatient stays with a diagnosis of asthma or wheezing; or (4) ⱖ150 total days’ supply of asthma-control medications, including oral or inhaled long-acting -agonists (LABAs), oral or inhaled corticosteroids (ICSs), methylxanthines, mast-cell stabilizers, leukotriene modifiers, anticholinergics, and combination medications (LABA ⫹ ICS). Short-acting -agonists were not included in the definition of asthma-control medication. Of particular interest was the possible relationship between GA and parental medical history of asthma or atopy with SECW in the study subjects. Therefore, the study patients were classified as having known GA ⱕ32 or 33 to 36 weeks, and SECW was assessed in these subgroups. In addition, parents’ medical claims during the 3 years before the child’s birth were assessed to determine parental history of atopy (defined as asthma, eczema, or allergic rhinitis based on presence of ICD-9-CM codes 493.XX, 691.8X, 692.9X, and 477.XX). In addition to SECW, the direct medical costs incurred between the second and third birthdays were calculated as the cost to the health plan as well as to the health plan member (ie, cost in this study included the total amount paid by the health plan as well as any copayment, coinsurance, and deductible amounts paid by the health plan member). Costs related to SECW (ie, office or outpatient, ED, or inpatient visits with a diagnosis of asthma or wheezing plus claims for systemic corticosteroids or asthma control medications, or both) were also calculated separately for pharmacyrelated and medical-related services. It is possible that, with this approach, any patient who did not meet the full definition of SECW (eg, patients with 1 outpatient visit for asthma or wheezing but no claim for systemic corticosteroids) would still incur “SECW-related” costs. All cost variables were consumer price index– adjusted to 2006 US dollars.24
Statistical Analysis To identify a control group of infants comparable to those with an RSV-LRI, we created a matched sample using a 2-step process. Each infant with RSV was
Volume 32 Number 14
J.R. Romero et al. matched to 5 control infants using either GA or birth weight (based on ICD-9-CM coding). Propensity scores were generated for this subpopulation using logistic regression based on seasonal year of birth, quarter of birth (May–July, August–October, November– January, and February–April), sex, health plan region, neonatal intensive care unit (NICU) use at birth, and length of birth hospitalization. Up to 3 control infants whose propensity scores were within ⫾0.01 of the propensity score of each RSV-LRI infant were selected. Propensity matching included NICU use and length of birth hospitalization, because these parameters serve as indicators of neonatal disease severity. Demographic and clinical characteristics were compared between cohorts using t tests for continuous variables and 2 tests for categoric variables. Such characteristics included seasonal year and quarter of birth, sex, health plan region, GA, birth weight, presence of select comorbid conditions following birth (Appendix), use of RSV prophylaxis, parental characteristics, and characteristics of the initial hospitalization for the infant’s birth. Logistic regression models were used to estimate the relationship between presenting with RSV in the first RSV season and SECW between the second and third birthdays. The primary independent variable in the model was the presence of RSV or an RSV-like condition. In a subanalysis, infants with known GA were stratified by GA group (GA ⱕ 32 or 33–36 weeks) and the relationship of RSV-LRI and SECW was explored within each stratum using 2 tests. Similarly, patients were stratified by parental history of atopy or asthma, and the relationship of RSV-LRI and SECW was examined within each stratum. Total health care costs and SECW costs were compared between subjects with and without SECW using t tests. A generalized linear model specified with a ␥ error distribution and a log link was used to compare total health care costs between patients with and without SECW, controlling for RSV cohort, seasonal year and quarter of birth, sex, health plan region, birth weight, postnatal comorbid conditions, RSV prophylaxis use, NICU use during birth hospitalization, and length of birth hospitalization. This method accounts for the skewed nature of cost data while avoiding the potential difficulties introduced by transformation and retransformation.25 All descriptive and logistic analyses were conducted using SAS version 9.1 (SAS Institute Inc., Cary, North Carolina). Generalized linear
December 2010
modeling was performed using Stata version 10.0 (StataCorp LP, College Station, Texas).
RESULTS Before application of the exclusion criteria, 39,812 patients with claims for prematurity or low birth weight were identified. Infants with claims for light for date or heavy for date were excluded from the study (n ⫽ 8405). A total of 7401 patients with missing GA and birth weight information and 17,299 patients who were not continuously enrolled in the health plan for 36 months after birth were excluded. Of the 6707 patients eligible for this study, 1024 (15%) were identified as having RSV or RSV-like conditions during the first RSV season, and 5683 (85%) lacked evidence of RSV infection. After the propensity score–matching procedure, 378 RSV patients and 606 matched controls were included in the analytic data set. The goal of matching was to create 2 cohorts that were balanced using observable demographics and clinical characteristics. Table I shows the descriptive characteristics of the 2 groups, both before and after propensity score matching. Before matching, RSV patients were significantly more likely to be male (P ⬍ 0.001), have comorbid conditions after birth (P ⫽ 0.007), have parents with a history of asthma (P ⫽ 0.045) or atopy (P ⫽ 0.009), and require oxygen at birth (P ⫽ 0.006). In addition, patients with RSV and controls had significantly different distributions of quarter of birth (P ⬍ 0.001), GA (P ⬍ 0.001), and birth weight (P ⫽ 0.001). After the propensity score–matching process, RSV patients were significantly less likely to be male (57% vs 67%; P ⫽ 0.002) and have a slightly different distribution of birth weight than controls (P ⫽ 0.048). Patients with RSV were significantly less likely to have a history of prophylaxis (P ⫽ 0.016) and were more likely to have parents with a history of atopy (P ⫽ 0.001) and a shorter initial hospitalization at birth (P ⫽ 0.005). Although propensity score matching created more balanced cohorts, there were still some differences among the groups. Logistic regression models were used to control for remaining differences. The prevalence of SECW between the second and third birthdays was significantly higher among infants with RSV in the matched population (Table II). After propensity score matching, 16.7% of RSV patients experienced SECW between their second and third birthdays, whereas 8.6% of matched controls experienced
2425
Clinical Therapeutics
Table I. Descriptive characteristics of the study population. Prematch
Characteristic
RSV Cohort Control Cohort (n ⫽ 5683), (n ⫽ 1024), No. (%) No. (%)
Seasonal year of birth 2001 2002 2003
299 (29.2) 354 (34.6) 371 (36.2)
1569 (27.6) 1970 (34.7) 2144 (37.7)
Quarter of birth May–July August–October November–January February–April
381 (37.2) 350 (34.2) 237 (23.1) 56 (5.5)
1323 (23.3) 1406 (24.7) 1416 (24.9) 1538 (27.1)
Postmatch
P
RSV Cohort Control Cohort (n ⫽ 606), (n ⫽ 378), No. (%) No. (%)
0.520
0.079 104 (27.5) 132 (34.9) 142 (37.6)
171 (28.2) 247 (40.8) 188 (31.0)
145 (38.4) 111 (29.4) 83 (22.0) 39 (10.3)
219 (36.1) 199 (32.8) 116 (19.1) 72 (11.9)
215 (56.9)
404 (66.7)
148 (39.2) 137 (36.2) 35 (9.3) 58 (15.3)
255 (42.1) 222 (36.6) 48 (7.9) 81 (13.4)
236 (62.4) 9 (2.4) 11 (2.9) 118 (31.2) 4 (1.1)
392 (64.7) 19 (3.1) 31 (5.1) 158 (26.1) 6 (1.0)
15 (4.0) 32 (8.5) 163 (43.1) 168 (44.4)
20 (3.3) 83 (13.7) 270 (44.6) 233 (38.5)
⬍0.001
⬍0.001 0.072
P
0.454
Male sex Health plan geographic region South Midwest Northeast West
617 (60.3)
3015 (53.1)
404 (39.5) 367 (35.8) 89 (8.7) 164 (16.0)
2206 (38.8) 1890 (33.3) 634 (11.2) 953 (16.8)
Gestational age, wk Unknown/Unspecified ⱕ28 ⬎28–⬍33 33–⬍37 ⱖ37
630 (61.5) 37 (3.6) 57 (5.6) 291 (28.4) 9 (0.9)
3035 (53.4) 178 (3.1) 426 (7.5) 1988 (35.0) 56 (1.0)
Birth weight, g Unknown/Unspecified ⬍1500 1500–⬍2500 ⱖ2500
37 (3.6) 132 (12.9) 438 (42.8) 417 (40.7)
213 (3.7) 789 (13.9) 2736 (48.1) 1945 (34.2)
125 (12.2) 258 (25.2)
538 (9.5) 1445 (25.4)
0.007 0.876
34 (9.0) 88 (23.3)
66 (10.9) 184 (30.4)
0.338 0.016
80 (9.0) 261 (29.4)
336 (7.1) 1194 (25.2)
0.045 0.009
34 (10.5) 107 (32.9)
34 (6.8) 111 (22.1)
0.058 0.001
405 (39.6) 46 (4.5) 16.2 (27.4) 6.7 (12.9)
2144 (37.7) 163 (2.9) 15.1 (23.5) 7.4 (12.6)
0.268 0.006 0.257 0.252
132 (34.9) 7 (1.9) 11.3 (16.8) 5.5 (8.2)
205 (33.8) 16 (2.6) 14.9 (22.9) 7.8 (14.8)
0.725 0.426 0.005 0.010
Any postnatal conditions* Use of RSV prophylaxis Characteristics of parents History of asthma History of atopy Characteristics of initial hospitalization NICU use Oxygen use Length of birth hospitalization, mean (SD) Length of initial hospitalization in those who did not have a NICU stay, mean (SD)
⬍0.001
0.002 0.652
0.239
0.001
0.048
RSV ⫽ respiratory syncytial virus; NICU ⫽ neonatal intensive care unit. * Comorbid postnatal conditions include congenital heart disease, chronic lung disease, HIV, trisomy 21, intraventricular hemorrhage, necrotizing enterocolitis, hydrocephalus, periventricular leukomalacia, retinopathy of prematurity, sensorineural hearing loss, and cerebral palsy (codes in Appendix).
2426
Volume 32 Number 14
J.R. Romero et al.
Table II. Postmatch serious early childhood wheezing between second and third birthday.
Parameter SECW Presence of ⱖ1 of the following* ⱖ3 Office, outpatient, or ED visits Office, outpatient, or ED visit ⫹ systemic corticosteroids within 7 days Inpatient stay ⱖ150-Day supply of asthma controller medications
RSV Cohort (n ⫽ 378), No. (%)
Control Cohort (n ⫽ 606), No. (%)
P
63 (16.7)
52 (8.6)
⬍0.001
33 (8.7) 51 (13.5)
20 (3.3) 42 (6.9)
⬍0.001 ⬍0.001
4 (1.1) 14 (3.7)
12 (2.0) 14 (2.3)
0.266 0.201
RSV ⫽ respiratory syncytial virus; SECW ⫽ serious early childhood wheezing; ED ⫽ emergency department. * Visits or inpatient stays with a diagnosis of asthma or wheezing (International Classification of Diseases, Ninth Revision, Clinical Modification22 codes 493.XX or 786.07) were included. Asthma controller medications included oral or inhaled long-acting beta agonists, oral or inhaled corticosteroids, methylxanthines, mast cell stabilizers, leukotriene modifiers, anticholinergics, and combination medications.
SECW (P ⬍ 0.001) during the same period. As shown in Table II, subgroup analysis indicated a significantly greater proportion of patients with 3 or more office, outpatient, or ED visits for wheezing among the RSV cohort (8.7%) than among the control cohort (3.3%) (P ⬍ 0.001). Similarly, the RSV cohort had significantly higher use of systemic steroids for wheezing events (13.5%) than did the control group (6.9%) (P ⬍ 0.001). Therefore, using SECW criteria allowed the study to capture an additional 8% wheezing-related events in the RSV cohort compared with recurrent wheezing (ⱖ3 events) and an additional 5.3% wheezing-related events in the control group. Table III presents the results of the logistic regression model for SECW. Patients with known birth weight were included in the multivariate model (n ⫽ 949). After adjusting for possible confounding variables, patients presenting with RSV infection in the first RSV season had a significant positive association (OR ⫽ 2.52) with SECW between the second and third birthdays (P ⬍ 0.001). In addition, males had higher odds of SECW than females (OR ⫽ 1.80; P ⫽ 0.011), and each additional inpatient day during the birth hospitalization was significantly associated with higher odds of SECW (OR ⫽ 1.01; P ⫽ 0.036). Table IV provides the results of the substudies of patients with known GA ⱕ36 weeks and patients with and without parental atopy or asthma. After stratification of patients based on GA, RSV patients with GA 33
December 2010
to 36 weeks had a significantly higher rate of SECW than matched controls (P ⫽ 0.009). The prevalence of SECW was significantly higher (P ⬍ 0.001) among patients with RSV compared with controls in patients whose parents did not have a history of atopy or asthma. Among patients whose parents had a history of atopy or asthma, the rate of SECW was not significantly different between patients with RSV and controls. Finally, SECW-related and total health care costs (costs for pharmacy- and medical-related services) were evaluated in patients who experienced SECW between their second and third birthdays. During this interval, patients with SECW had a mean SECW-related cost of $1378 (95% CI, $939 –$1816). Of this total, ⬃46% of costs were for pharmacy-filled prescriptions (mean, $630; 95% CI, $471–$789), and the remaining 54% of costs were for medical services (mean, $748; 95% CI, $375–$1121). In contrast, patients without evidence of SECW had a mean SECWrelated cost of $37 (95% CI, $24 –$51). Of the total costs, a mean of $32 was pharmacy related (95% CI, $19 –$45) and $6 was for medical services (95% CI, $2–$10). SECW-related total costs among patients with SECW accounted for nearly 20% of the $7138 (95% CI, $5087–$9189) in total health care costs during this time period. The mean SECW-related pharmacy cost among patients with SECW accounted for 60% of the $1051 (95% CI, $823–$1279) in total
2427
Clinical Therapeutics
Table III. Logistic model of serious early childhood wheezing between second and third birthdays (N ⫽ 949). Variables
Odds Ratio
95% CI
P
2.522
1.652–3.849
⬍0.001
0.903 1.083
0.534–1.529 0.643–1.822
0.704 0.765
0.878 0.751 0.754
0.533–1.445 0.419–1.344 0.357–1.592
0.609 0.334 0.460
1.799
1.142–2.834
0.011
0.822 1.362 0.703
0.507–1.333 0.683–2.718 0.356–1.391
0.427 0.381 0.312
Birth weight, g (⬍1500 g as reference) 1500–⬍2500 ⱖ2500
1.090 0.901
0.505–2.353 0.384–2.110
0.827 0.810
Any postnatal condition* RSV prophylaxis use NICU use during birth hospitalization Length of birth hospitalization
0.764 1.026 1.171 1.013
0.363–1.608 0.626–1.681 0.727–1.885 1.001–1.024
0.479 0.919 0.517 0.036
RSV cohort Seasonal year of birth (2001 as reference) 2002 2003 Quarter of birth (May–July as reference) August–October November–January February–April Male sex Health plan geographic region (South as reference) Midwest Northeast West
RSV ⫽ respiratory syncytial virus; NICU ⫽ neonatal intensive care unit. * Comorbid postnatal conditions included congenital heart disease, chronic lung disease, HIV, trisomy 21, intraventricular hemorrhage, necrotizing enterocolitis, hydrocephalus, periventricular leukomalacia, retinopathy of prematurity, sensorineural hearing loss, and cerebral palsy (codes in the Appendix).
pharmacy-related costs, and SECW-related medical costs accounted for 12% of the $6087 (95% CI, $4165–$8009) in total medical costs. In contrast, patients without evidence of SECW had a mean total health care cost of $2521 (95% CI, $1789 –$3253), with $164 in pharmacy costs and $2357 in medical costs. After adjusting for possible confounders, patients with SECW had a significantly higher total health care cost than patients without evidence of SECW (P ⬍ 0.001). Patients with SECW had adjusted total costs that were 2.67-fold higher than did patients without evidence of SECW (data not shown).
DISCUSSION The primary end point, SECW, was developed to capture clinically meaningful wheezing in addition to determine the frequency of wheezing illnesses. Perhaps
2428
more important than the additional number of events were the types of events that this definition of SECW captured. These additional events consisted of those requiring the use of asthma-control medications for ⱖ5 months, those requiring systemic steroids for acute exacerbations, and those needing hospitalization. These additional events are considered clinically meaningful because of the inherent severity of wheezing disease that necessitates such clinical management. Therefore, in this study, the authors conclude that SECW is a more clinically relevant and accurate end point that represents airway morbidity during early childhood years and measures not only the frequency but also the severity of wheezing. Using this definition, an association was found between RSV-LRI occurring in infancy and airway morbidity between 2 and 3 years of age. Even after adjust-
Volume 32 Number 14
J.R. Romero et al.
Table IV. Rates of serious early childhood wheezing between second and third birthdays stratified by other variables of interest. Variable
RSV Cohort, No. (%)
Control Cohort, No. (%)
P
Gestational age ⱕ32 wk* 33–36 wk
4/20 (20.0) 22/118 (18.6)
4/50 (8.0) 12/158 (7.6)
Parental atopy Yes* No
16/107 (15.0) 40/218 (18.3)
9/111 (8.1) 32/392 (8.2)
0.113 ⬍0.001
Parental asthma Yes* No
4/34 (11.8) 52/291 (17.9)
3/34 (8.8) 38/469 (8.1)
1.000 ⬍0.001
0.213 0.009*
RSV ⫽ respiratory syncytial virus. * RSV and control patients were matched overall and matches may not appear in the same stratum.
ing for different covariates using a logistic model, this association was found to be significant (P ⬍ 0.001). Stratification of the analysis for preterm infants with known GA showed differences similar to those observed for the overall group. However, the number of children analyzed was too small to achieve statistical significance. Similarly, the number of evaluable patients decreased substantially when the data were stratified to explore the role of parental history of atopy or asthma on the association of RSV-LRI and presence of wheezing. An association between RSV-LRI in infancy and later wheezing remained significant in patients with no family history of asthma, atopy, or both. Thus, the current data did not further the understanding of the role of parental atopy–asthma and its association with RSV-LRI in infancy and subsequent SECW. It is expected that a larger sample size would allow the stratification necessary to address the role of family history of asthma or atopy in development of SECW even beyond the 3 years investigated in this analysis. This study did, however, provide some insight into the effects of SECW on direct costs. Patients with SECW had higher total health care costs compared with those who did not have SECW. The costs contributed by SECW were between $1300 and $1400 per study subject. Annual direct asthma-related costs in pediatric patients aged 4 to 17 years have been reported to range from $393 to $1316, based on treatments used.26,27 The SECW-related costs in children
December 2010
aged 2 to 3 years appear to be comparable, suggesting that in a younger patient cohort, SECW could place a financial burden on health plan resources similar to that associated with asthma in older children. A systematic review by Pérez-Yarza et al9 concluded that RSV-LRI was associated with an increased risk for subsequent development of asthma or recurrent wheezing and that this association became progressively smaller with increasing age. Using an end point that considers frequency and severity, the present study in preterm infants adds to this body of literature and supports the association between RSV-LRI in infancy and airway hyperreactivity in later life. The study also found associations between variables such as male sex and longer length of birth hospitalization and increased airway morbidity in later life. As summarized by Meissner and Long,11 multiple theories have been proposed in an effort to understand the association between RSV-LRI in infancy and subsequent recurrent wheezing and asthma. In the causal theory, it is believed that RSV infection occurring during the final stage of lung development (ie, the first 2–3 years of life) alters normal development and that the resulting long-term consequences on lung structure and function predispose these infants to future episodes of wheezing.11 Another possibility is that in certain infants, innate preexisting abnormalities of airway function or immune response are triggered by an early viral respiratory tract infection and lead to airway ob-
2429
Clinical Therapeutics struction.11 A complex multifactorial theory holds that no single response to a viral respiratory tract infection determines outcome.11 Instead, the long-term outcome is a product of multiple factors, such as the infant’s genetic makeup, concomitant exposure to environmental antigens, and state of maturation of the immune system and airway at the time of infection. In this theory, no single factor dominates any other in determining outcome. Although the strongest association between viral lower respiratory tract infection and subsequent development of wheezing has been noted with RSV, there have been reports of other viral infections, such as rhinovirus, influenza virus, and parainfluenza virus, that triggered subsequent wheezing.28 –30 One approach to delineating whether the pathway is causal or RSV-specific would be to demonstrate a reduction in long-term airway morbidity by reducing RSV in infancy. Although some investigations of RSV immunoprophylaxis support the hypothesis that modification of RSV infection in premature infants decreases risk for subsequent recurrent wheezing, a well-designed randomized study is needed to establish the association.14,15 The findings from this study must be considered in light of several limitations. Thirty-six months of continuous enrollment in the health plan was required. Although continuous enrollment is often required for retrospective analyses of health care outcomes, it can be problematic when including patients with potentially fatal illnesses, such as RSV. In this analysis, patients who disenrolled from the health plan before age 3 years for any reason (including death) were excluded. The analysis included infants with RSV-like illnesses for whom a diagnosis of RSV infection could be confirmed. Some of the infants in the control group may have experienced mild RSV infections that did not require medical attention and were misclassified, thereby underestimating the effect on study outcomes. The data used in this study came from a commercial MCO population. Therefore, the findings from this analysis are primarily applicable to the management of RSV infections in premature and low-birth-weight infants in similar managed care settings and may not be applicable to a Medicaid plan. Although propensity score matching and multivariate methods allowed for the selection of a comparable control group of infants, complete matching was not attained. These methods cannot adjust for imbalances due to unobserved covariates. All limitations associated with the use of medi-
2430
cal claims data (ie, coding errors, undercoding, and capture of only direct costs) are applicable.
CONCLUSIONS The development of RSV-LRI in infancy in preterm infants was associated with an increased prevalence of SECW between ages 2 and 3 years. Patients with SECW had higher total health care costs compared with those who did not have SECW.
ACKNOWLEDGMENTS Funding for this research was provided by MedImmune, LLC. The authors acknowledge Physicians World, Seacaucus, New Jersey, and Chris A. Kirk, PhD, and Gerard P. Johnson, PhD, Complete Healthcare Communications, Chadds Ford, Pennsylvania, for providing editing support funded by MedImmune. Drs. Romero and Stewart provide consultant services to MedImmune. Dr. Romero has received funding from MedImmune, has been a member of MedImmune’s speakers’ bureau, and has participated in clinical studies sponsored by Novartis Pharmaceuticals Corporation, GlaxoSmithKline, Roche, and Johnson & Johnson Services, Inc. Dr. Stewart has been a member of speakers’ bureaus for Ikaria, Inc. and Abbott Nutrition and a member of an advisory board with Abbott International. Ms. Buysman is employed by i3 Innovus, which was contracted by MedImmune to do the analysis. Drs. Fernandes, Jafri, and Mahadevia are employees of MedImmune. The authors have indicated that they have no other conflicts of interest regarding the content of this article.
REFERENCES 1. Hall CB. Respiratory syncytial virus. In: Feigen RD, Cherry JD, eds. Textbook of Pediatric Infectious Diseases. 2nd ed. Philadelphia, Pa: Saunders; 1987:1653–1675. 2. Hall CB, Long CE, Schnabel KC. Respiratory syncytial virus infections in previously healthy working adults. Clin Infect Dis. 2001;33:792–796. 3. Hall CB, Weinberg GA, Iwane MK, et al. The burden of respiratory syncytial virus infection in young children. N Engl J Med. 2009;360:588 –598. 4. Institute of Medicine. Prospects for Immunizing Against Respiratory Synctial Virus. Washington, DC: National Academy Press; 1985. 5. Ruuskanen O, Ogra PL. Respiratory syncytial virus. Curr Probl Pediatr. 1993;23:50 –79.
Volume 32 Number 14
J.R. Romero et al. 6. Shay DK, Holman RC, Newman RD, et al. Bronchiolitis-associated hospitalizations among US children, 1980 –1996. JAMA. 1999;282: 1440 –1446. 7. Leader S, Kohlhase K. Respiratory syncytial virus– coded pediatric hospitalizations, 1997 to 1999. Pediatr Infect Dis J. 2002;21:629 – 632. 8. Bont L, Aalderen WM, Kimpen JL. Long-term consequences of respiratory syncytial virus (RSV) bronchiolitis. Paediatr Respir Rev. 2000;1:221– 227. 9. Pérez-Yarza EG, Moreno A, Lázaro P, et al. The association between respiratory syncytial virus infection and the development of childhood asthma: A systematic review of the literature. Pediatr Infect Dis J. 2007;26: 733–739. 10. Mohapatra SS, Boyapalle S. Epidemiologic, experimental, and clinical links between respiratory syncytial virus infection and asthma. Clin Microbiol Rev. 2008;21:495–504. 11. Meissner HC, Long SS. Respiratory syncytial virus infection and recurrent wheezing: A complex relationship. J Pediatr. 2007;151:6 –7. 12. Simões EA. RSV disease in the pediatric population: Epidemiology, seasonal variability, and long-term outcomes. Manag Care. 2008;17:3– 6, discussion 18 –19. 13. Respiratory Syncytial Virus. In: Pickering LK, Baker CJ, Long SS, McMillan JA, eds. Red Book: 2006 Report of the Committee on Infectious Diseases. 27th ed. Elk Grove Village, Ill: American Academy of Pediatrics; 2006: 560 –566. 14. Simoes EA, Groothuis JR, CarbonellEstrany X, et al, for the Palivizumab Long-Term Respiratory Outcomes Study Group. Palivizumab prophylaxis, respiratory syncytial virus, and subsequent recurrent wheezing. J Pediatr. 2007;151:34 – 42, 42.e1. 15. Wenzel SE, Gibbs RL, Lehr MV, Simoes EA. Respiratory outcomes in high-risk children 7 to 10 years after prophylaxis with respiratory syncy-
December 2010
16.
17.
18.
19.
20.
21.
22.
tial virus immune globulin. Am J Med. 2002;112:627– 633. Henderson J, Hilliard TN, Sherriff A, et al. Hospitalization for RSV bronchiolitis before 12 months of age and subsequent asthma, atopy and wheeze: A longitudinal birth cohort study. Pediatr Allergy Immunol. 2005; 16:386 –392. Sigurs N, Bjarnason R, Sigurbergsson F, Kjellman B. Respiratory syncytial virus bronchiolitis in infancy is an important risk factor for asthma and allergy at age 7. Am J Respir Crit Care Med. 2000;161:1501–1507. Sigurs N, Gustafsson PM, Bjarnason R, et al. Severe respiratory syncytial virus bronchiolitis in infancy and asthma and allergy at age 13. Am J Respir Crit Care Med. 2005;171:137– 141. Stein RT, Sherrill D, Morgan WJ, et al. Respiratory syncytial virus in early life and risk of wheeze and allergy by age 13 years. Lancet. 1999;354:541– 545. Sigurs N, Bjarnason R, Sigurbergsson F, et al. Asthma and immunoglobulin E antibodies after respiratory syncytial virus bronchiolitis: A prospective cohort study with matched controls. Pediatrics. 1995; 95:500 –505. National Heart, Lung, and Blood Institute, US Department of Health and Human Services. National Asthma Education and Prevention Program. Expert panel report 3: Guidelines for the Diagnosis and Management of Asthma. http://www.nhlbi.nih.gov/ guidelines/asthma/asthgdln.pdf. Accessed January 20, 2011. World Health Organization. International Statistical Classification of Diseases and Related Health Problems, Ninth Revision, Clinical Modification. Geneva, Switzerland: WHO; 2003.
23. Committee on Infectious Diseases. From the American Academy of Pediatrics: Policy statements—Modified recommendations for use of palivizumab for prevention of respiratory syncytial virus infections. Pediatrics. 2009;124:1694 –1701. 24. Bureau of Labor Statistics, US Department of Labor. Consumer Price Index, Chained Consumer Price Index 2006. http://www.bls.gov/ data/. Accessed January 20, 2011. 25. Manning WG. The logged dependent variable, heteroscedasticity, and the retransformation problem. J Health Econ. 1998;17:283–295. 26. Stempel DA, Kruzikas DT, Manjunath R. Comparative efficacy and cost of asthma care in children with asthma treated with fluticasone propionate and montelukast. J Pediatr. 2007;150:162–167. 27. Stempel DA, Riedel AA, Carranza RosenzweigJR.Resourceutilizationwith fluticasone propionate and salmeterol in a single inhaler compared with other controller therapies in children with asthma. Curr Med Res Opin. 2006;22: 463–470. 28. ErikssonM,BennetR,NilssonA.Wheezing following lower respiratory tract infections with respiratory syncytial virus and influenza A in infancy. Pediatr Allergy Immunol. 2000;11:193–197. 29. Korppi M, Kotaniemi-Syrjänen A, Waris M, et al. Rhinovirus-associated wheezingininfancy:Comparisonwithrespiratory syncytial virus bronchiolitis. Pediatr InfectDisJ.2004;23:995–999. 30. Lemanske RF Jr. Viruses and asthma: Inception, exacerbation, and possible prevention. J Pediatr. 2003;142(Suppl 2):S3–S7; discussion S7–S8. 31. American Medical Association (AMA). CurrentProceduralTerminology:2001.Chicago, Ill: AMA; 2000.
Address correspondence to: José R. Romero, MD, Arkansas Children’s Hospital, 1 Children’s Way, Slot 512-11, Little Rock, AR 72202-3591. E-mail: [email protected]
2431
Clinical Therapeutics
Appendix. Codes and medications used to define comorbid conditions of interest. Comorbid Condition Congenital heart disease*
Code Type ICD-9-CM diagnosis Prescription
CPT ICD-9-CM procedure † Chronic lung disease ICD-9-CM diagnosis Prescription HIV ICD-9-CM diagnosis Trisomy 21 ICD-9-CM diagnosis Intraventricular hemorrhage ICD-9-CM diagnosis Necrotizing enterocolitis ICD-9-CM diagnosis Hydrocephalus ICD-9-CM diagnosis Periventricular leukomalacia ICD-9-CM diagnosis Retinopathy of prematurity ICD-9-CM diagnosis Sensorineural hearing loss ICD-9-CM diagnosis Cerebral palsy ICD-9-CM diagnosis
Codes 425.3, 425.4, 425.8, 745.XX, 746.XX, and 747.XX Digoxin, diuretics, ACE inhibitors, angiotensin receptor blockers, -blockers, vasodilators, combination medications 92992–92993 35.41 770.7 Bronchodilators, diuretics, diuretic combinations 042, V08, 079.51–079.53 758.0 772.1x 557.0, 777.5 741.0X, 742.3 779.7 362.21, 362.29 389.1X, 389.2 343.X, 333.71
ICD-9-CM ⫽ International Classification of Diseases, Ninth Revision, Clinical Modification22; ACE ⫽ angiotensin-converting enzyme; CPT ⫽ Current Procedural Terminology.31 * Defined as claims meeting ⱖ1 of the following criteria: (1) diagnosis code listed above and medication listed in the first 6 months of life or (2) CPT code or procedure code listed above in the first 6 months of life. † Defined as claims for 1 of the diagnostic codes listed above and claims for 1 of the medications listed in the first 30 days of life.
2432
Volume 32 Number 14
Serious Early Childhood Wheezing After Respiratory Syncytial Virus Lower Respiratory Tract Illness in Preterm Infants José R. Romero, MD1; Dan L. Stewart, MD2,3; Erin K. Buysman, MS4; Ancilla W. Fernandes, PhD5; Hasan S. Jafri, MD6; and Parthiv J. Mahadevia, MD, MPH5 1
Department of Pediatrics, Arkansas Children’s Hospital and University of Arkansas for Medical Sciences, Little Rock, Arkansas; 2Department of Pediatrics, University of Louisville Hospital, Louisville, Kentucky; 3 Kosair Children’s Hospital, Louisville, Kentucky; 4Health Economics and Outcomes Research, i3 Innovus, an Ingenix Company, Eden Prairie, Minnesota; 5Health Outcomes and Pharmacoeconomics, MedImmune, LLC, Gaithersburg, Maryland; and 6Clinical Development, MedImmune, LLC, Gaithersburg, Maryland ABSTRACT Background: Respiratory syncytial virus (RSV) lower respiratory tract infection (LRI) in early life has been associated with sustained airway hyperreactivity during childhood; however, corresponding data in premature infants are sparse. Objective: The objective of this study was to determine whether RSV-LRI during early infancy of preterm infants was associated with an increased risk for serious early childhood wheezing (SECW) by age 3 years. Methods: A retrospective cohort study was conducted using data from a large (⬃14 million members) US health plan database. The study population included infants ⱕ6 months of age born at ⱕ36 weeks’ gestational age or weighing ⬍2500 g, or both. Preterm infants with any medically attended RSV-LRI from May 2001 through April 2004 with 3 years of continuous eligibility were selected and propensity matched with ⱕ3 control infants. SECW was defined as ⬎3 office, outpatient, or emergency department (ED) visits with asthma or wheezing; ⱖ1 office, outpatient, or ED visit with asthma or wheezing plus treatment with systemic corticosteroids within 7 days; ⱖ1 inpatient stay with asthma or wheezing; or ⱖ150 days’ supply of asthma-control medications. The presence of SECW between ages 2 and 3 years was compared between infants with and without RSV-LRI using univariate and multivariate methods. Health care costs for patients with SECW were explored. Results: A total of 378 infants with RSV were matched to 606 controls. The prevalence of SECW between ages 2 and 3 years was 16.7% in the RSV-LRI group versus 8.6% in the control group (P ⬍ 0.001).
2422
Logistic regression showed that preterm infants with RSV in early life were 2.52-fold (95% CI,1.65–3.85) more likely to present with SECW between ages 2 and 3 years (P ⬍ 0.001). Patients with SECW had a mean SECW-related cost of US $1378 (95% CI, $939 –$1816) and total health care cost of $7138 (95% CI, $5087– $9189) compared with $37 (95% CI, $24 –$51) and $2521 (95% CI, $1789 –$3253), respectively, for patients without SECW. After adjusting for possible confounders, patients with SECW had a significantly higher total health care cost than did patients without evidence of SECW (P ⬍ 0.001). Conclusions: The development of RSV-LRI in infancy in preterm infants was associated with an increased prevalence of SECW between ages 2 and 3 years. Patients with SECW had higher total health care costs than those who did not have SECW. (Clin Ther. 2010;32:2422–2432) © 2010 Elsevier HS Journals, Inc. Key words: infants, outcome measures, premature, RSV, wheezing.
INTRODUCTION Recognized as the leading cause of significant lower respiratory tract disease in infants and children, respiThis study was presented as a poster at the American Academy of Pediatrics National Conference and Exhibition, October 17–20, 2009, Washington, DC. Accepted for publication November 4, 2010. doi:10.1016/j.clinthera.2011.01.007 0149-2918/$ - see front matter © 2010 Elsevier HS Journals, Inc. All rights reserved.
Volume 32 Number 14
J.R. Romero et al. ratory syncytial virus (RSV) accounts for 50% to 80% of cases of bronchiolitis and 30% to 60% of pneumonia admissions in children ⬍5 years of age during the winter season in the United States.1-6 RSV bronchiolitis was the leading cause of hospital admissions of infants aged ⬍1 year between 1997 and 1999 in the United States.7 Studies have reported that acute RSV-lower respiratory infection (LRI) in infancy may lead to long-term airway morbidity in the form of recurrent wheezing or asthma,8,9 and that airway reactivity rates can exceed 50% after an RSV infection.10 Gestational age (GA) may affect the development of airway reactivity,10 –12 and premature infants have been documented to be at high risk for serious or fatal RSV infection.13 However, limited studies have focused on the incidence of early childhood wheezing in premature infants infected with RSV.14,15 The reported prevalence and duration of the wheezing vary widely; epidemiologic studies have reported that wheezing after RSV-LRI occurs in 16% to 71% of patients.8,14,16-18 In their population-based birth cohort study, Henderson et al16 reported a cumulative prevalence of wheezing of 28.1% in the RSV group and 13.1% in the control group at 30 to 42 months. At 69 to 81 months, the prevalence of wheezing remained 22.6% in the RSV group and 9.6% in the control group. At 91 months, this study found a cumulative prevalence of asthma of 38.4% in the RSV group and 20.1% in the control group. A prospective study by Sigurs et al17 reported that the prevalence of asthma (30%) and wheezing (38%) was significantly higher in children aged ⱕ7 years who had been hospitalized as infants for RSV bronchiolitis than in a control population (3% and 2%, respectively). The data from patients aged 13 years18 supported RSV bronchiolitis in infancy as an independent risk factor for asthma and recurrent wheezing, with 43% of the RSV group reporting asthma or recurrent wheezing versus 8% of the control group. Stein et al19 found that outpatient-diagnosed RSV-LRI was associated with an increased risk for infrequent wheeze (odds ratio [OR] ⫽ 3.2) and frequent wheeze (OR ⫽ 4.3) at age 6 to 11 years. The risk decreased markedly with age and was no longer significant at age 13 years. Recurrent or frequent wheezing has been used as an outcome in several prospective studies14,17–20 as a way to capture airway disease that has a greater risk for adverse effects on a child’s health. A diagnosis of
December 2010
asthma is recognized to be clinically important, as its associated symptoms and effects on a child’s health are better understood. However, diagnosing asthma in children ⬍5 years of age is difficult because there is not a consensus on objective measures for asthma diagnosis and management guidelines.21 However, outcomes relying only on the frequency of symptoms may miss less frequent but severe or life-threatening events that require long-term control medications or hospitalization. Therefore, serious early childhood wheezing (SECW) has been proposed as an outcome that captures not only frequency but also severity of disease by linking wheezing to the use of specific medications, such as systemic steroids or asthma-control medication, or both. The objective of this study was to assess the association of SECW occurring between ages 2 and 3 years after RSV-LRI in early infancy in a cohort of preterm infants infected during their first RSV season. In a subset of infants in whom GA codes were available, it was determined whether GA modified this association between RSV-LRI and SECW. Similarly, in a subset of infants in whom a family history of atopy or asthma could be established, it was determined whether these factors modified the association between RSV-LRI and SECW. The financial impact of SECW in terms of direct costs, including pharmacy costs, was explored.
PATIENTS AND METHODS Study Design A retrospective cohort study was conducted using eligibility, medical, and pharmacy administrative claims data from a large US managed care organization (MCO). Eligible enrollees were from a commercially insured preferred-provider organization model of a national MCO consisting of ⬃14 million members with medical and pharmacy benefits during 2007. The MCO provided full insurance coverage for physician, hospital, and pharmacy services. Individuals covered by the health plan were from geographically diverse regions of the United States.
Study Population Infants born May 2001 through April 2004 were identified from the MCO claims data. Patients were included in this study if they met the following criteria: (1) had at least one medical claim with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM)22 diagnostic code for pre-
2423
Clinical Therapeutics maturity, defined as ⬍36 weeks GA (codes 765.21– 765.28) or low birth weight, defined as ⬍2500 g (codes 765.0x, 765.1x, and V21.30 –V21.35) in either the primary or secondary position on physician or hospital claims; (2) were commercially insured with a medical and pharmacy benefit; and (3) had ⱖ36 months of continuous enrollment in the health plan after birth. Infants with claims designated for light for date (ICD9-CM code 764.XX) or heavy for date (code 766.XX) were excluded from the study to decrease heterogeneity in the selected cohort. Patients were followed up from their date of birth through their third birthday. Each infant was assigned to the seasonal year of birth based on date of birth. A seasonal year of birth was defined as May 1 through April 30 of the following year, allowing each study year to capture an entire RSV season. This is in contrast with using the calendar year, which would have captured half of 2 different RSV seasons. Study patients were classified as with or without evidence of an RSV event during their first RSV season (ie, November 1–April 30 of the infant’s seasonal year of birth; peak RSV activity normally occurs between November and March in the temperate climates of North America).23 An RSV event was defined as either (1) a claim for an RSV condition (ICD-9-CM code 079.6, 466.11, or 480.1 in either the primary or secondary position) during an inpatient, emergency department (ED), or ambulatory visit; or (2) a claim for an RSV-like condition (code 466.0, 466.19, 480.9, 485, or 486 in either the primary or secondary position) during an inpatient, ED, or ambulatory visit, with no evidence of claims for influenza or bacterial pneumonia (code 481, 482.XX, or 487.X) within 3 days of the RSV-like claim.
Study Outcomes The primary objective of this study was to determine the rate of clinically significant childhood wheezing in premature or low-birth-weight patients with and without evidence of RSV infection during their first RSV season. After consultation with physicians and researchers with expertise in pulmonology, infectious diseases, immunology, allergy, neonatology, general pediatrics, and health outcomes, SECW between the second and third birthdays was proposed as a measure of clinically meaningful childhood wheezing. To be classified as having SECW, patients were required to meet at least one of the following criteria between their
2424
second and third birthdays: (1) ⱖ3 office, outpatient, or ED visits with a diagnosis of asthma or wheezing (ICD-9-CM code 493.XX or 786.07); (2) ⱖ1 office, outpatient, or ED visit with a diagnosis of asthma or wheezing with a claim for systemic (oral or intravenous) corticosteroids within 7 days of the asthma or wheezing claim; (3) ⱖ1 inpatient stays with a diagnosis of asthma or wheezing; or (4) ⱖ150 total days’ supply of asthma-control medications, including oral or inhaled long-acting -agonists (LABAs), oral or inhaled corticosteroids (ICSs), methylxanthines, mast-cell stabilizers, leukotriene modifiers, anticholinergics, and combination medications (LABA ⫹ ICS). Short-acting -agonists were not included in the definition of asthma-control medication. Of particular interest was the possible relationship between GA and parental medical history of asthma or atopy with SECW in the study subjects. Therefore, the study patients were classified as having known GA ⱕ32 or 33 to 36 weeks, and SECW was assessed in these subgroups. In addition, parents’ medical claims during the 3 years before the child’s birth were assessed to determine parental history of atopy (defined as asthma, eczema, or allergic rhinitis based on presence of ICD-9-CM codes 493.XX, 691.8X, 692.9X, and 477.XX). In addition to SECW, the direct medical costs incurred between the second and third birthdays were calculated as the cost to the health plan as well as to the health plan member (ie, cost in this study included the total amount paid by the health plan as well as any copayment, coinsurance, and deductible amounts paid by the health plan member). Costs related to SECW (ie, office or outpatient, ED, or inpatient visits with a diagnosis of asthma or wheezing plus claims for systemic corticosteroids or asthma control medications, or both) were also calculated separately for pharmacyrelated and medical-related services. It is possible that, with this approach, any patient who did not meet the full definition of SECW (eg, patients with 1 outpatient visit for asthma or wheezing but no claim for systemic corticosteroids) would still incur “SECW-related” costs. All cost variables were consumer price index– adjusted to 2006 US dollars.24
Statistical Analysis To identify a control group of infants comparable to those with an RSV-LRI, we created a matched sample using a 2-step process. Each infant with RSV was
Volume 32 Number 14
J.R. Romero et al. matched to 5 control infants using either GA or birth weight (based on ICD-9-CM coding). Propensity scores were generated for this subpopulation using logistic regression based on seasonal year of birth, quarter of birth (May–July, August–October, November– January, and February–April), sex, health plan region, neonatal intensive care unit (NICU) use at birth, and length of birth hospitalization. Up to 3 control infants whose propensity scores were within ⫾0.01 of the propensity score of each RSV-LRI infant were selected. Propensity matching included NICU use and length of birth hospitalization, because these parameters serve as indicators of neonatal disease severity. Demographic and clinical characteristics were compared between cohorts using t tests for continuous variables and 2 tests for categoric variables. Such characteristics included seasonal year and quarter of birth, sex, health plan region, GA, birth weight, presence of select comorbid conditions following birth (Appendix), use of RSV prophylaxis, parental characteristics, and characteristics of the initial hospitalization for the infant’s birth. Logistic regression models were used to estimate the relationship between presenting with RSV in the first RSV season and SECW between the second and third birthdays. The primary independent variable in the model was the presence of RSV or an RSV-like condition. In a subanalysis, infants with known GA were stratified by GA group (GA ⱕ 32 or 33–36 weeks) and the relationship of RSV-LRI and SECW was explored within each stratum using 2 tests. Similarly, patients were stratified by parental history of atopy or asthma, and the relationship of RSV-LRI and SECW was examined within each stratum. Total health care costs and SECW costs were compared between subjects with and without SECW using t tests. A generalized linear model specified with a ␥ error distribution and a log link was used to compare total health care costs between patients with and without SECW, controlling for RSV cohort, seasonal year and quarter of birth, sex, health plan region, birth weight, postnatal comorbid conditions, RSV prophylaxis use, NICU use during birth hospitalization, and length of birth hospitalization. This method accounts for the skewed nature of cost data while avoiding the potential difficulties introduced by transformation and retransformation.25 All descriptive and logistic analyses were conducted using SAS version 9.1 (SAS Institute Inc., Cary, North Carolina). Generalized linear
December 2010
modeling was performed using Stata version 10.0 (StataCorp LP, College Station, Texas).
RESULTS Before application of the exclusion criteria, 39,812 patients with claims for prematurity or low birth weight were identified. Infants with claims for light for date or heavy for date were excluded from the study (n ⫽ 8405). A total of 7401 patients with missing GA and birth weight information and 17,299 patients who were not continuously enrolled in the health plan for 36 months after birth were excluded. Of the 6707 patients eligible for this study, 1024 (15%) were identified as having RSV or RSV-like conditions during the first RSV season, and 5683 (85%) lacked evidence of RSV infection. After the propensity score–matching procedure, 378 RSV patients and 606 matched controls were included in the analytic data set. The goal of matching was to create 2 cohorts that were balanced using observable demographics and clinical characteristics. Table I shows the descriptive characteristics of the 2 groups, both before and after propensity score matching. Before matching, RSV patients were significantly more likely to be male (P ⬍ 0.001), have comorbid conditions after birth (P ⫽ 0.007), have parents with a history of asthma (P ⫽ 0.045) or atopy (P ⫽ 0.009), and require oxygen at birth (P ⫽ 0.006). In addition, patients with RSV and controls had significantly different distributions of quarter of birth (P ⬍ 0.001), GA (P ⬍ 0.001), and birth weight (P ⫽ 0.001). After the propensity score–matching process, RSV patients were significantly less likely to be male (57% vs 67%; P ⫽ 0.002) and have a slightly different distribution of birth weight than controls (P ⫽ 0.048). Patients with RSV were significantly less likely to have a history of prophylaxis (P ⫽ 0.016) and were more likely to have parents with a history of atopy (P ⫽ 0.001) and a shorter initial hospitalization at birth (P ⫽ 0.005). Although propensity score matching created more balanced cohorts, there were still some differences among the groups. Logistic regression models were used to control for remaining differences. The prevalence of SECW between the second and third birthdays was significantly higher among infants with RSV in the matched population (Table II). After propensity score matching, 16.7% of RSV patients experienced SECW between their second and third birthdays, whereas 8.6% of matched controls experienced
2425
Clinical Therapeutics
Table I. Descriptive characteristics of the study population. Prematch
Characteristic
RSV Cohort Control Cohort (n ⫽ 5683), (n ⫽ 1024), No. (%) No. (%)
Seasonal year of birth 2001 2002 2003
299 (29.2) 354 (34.6) 371 (36.2)
1569 (27.6) 1970 (34.7) 2144 (37.7)
Quarter of birth May–July August–October November–January February–April
381 (37.2) 350 (34.2) 237 (23.1) 56 (5.5)
1323 (23.3) 1406 (24.7) 1416 (24.9) 1538 (27.1)
Postmatch
P
RSV Cohort Control Cohort (n ⫽ 606), (n ⫽ 378), No. (%) No. (%)
0.520
0.079 104 (27.5) 132 (34.9) 142 (37.6)
171 (28.2) 247 (40.8) 188 (31.0)
145 (38.4) 111 (29.4) 83 (22.0) 39 (10.3)
219 (36.1) 199 (32.8) 116 (19.1) 72 (11.9)
215 (56.9)
404 (66.7)
148 (39.2) 137 (36.2) 35 (9.3) 58 (15.3)
255 (42.1) 222 (36.6) 48 (7.9) 81 (13.4)
236 (62.4) 9 (2.4) 11 (2.9) 118 (31.2) 4 (1.1)
392 (64.7) 19 (3.1) 31 (5.1) 158 (26.1) 6 (1.0)
15 (4.0) 32 (8.5) 163 (43.1) 168 (44.4)
20 (3.3) 83 (13.7) 270 (44.6) 233 (38.5)
⬍0.001
⬍0.001 0.072
P
0.454
Male sex Health plan geographic region South Midwest Northeast West
617 (60.3)
3015 (53.1)
404 (39.5) 367 (35.8) 89 (8.7) 164 (16.0)
2206 (38.8) 1890 (33.3) 634 (11.2) 953 (16.8)
Gestational age, wk Unknown/Unspecified ⱕ28 ⬎28–⬍33 33–⬍37 ⱖ37
630 (61.5) 37 (3.6) 57 (5.6) 291 (28.4) 9 (0.9)
3035 (53.4) 178 (3.1) 426 (7.5) 1988 (35.0) 56 (1.0)
Birth weight, g Unknown/Unspecified ⬍1500 1500–⬍2500 ⱖ2500
37 (3.6) 132 (12.9) 438 (42.8) 417 (40.7)
213 (3.7) 789 (13.9) 2736 (48.1) 1945 (34.2)
125 (12.2) 258 (25.2)
538 (9.5) 1445 (25.4)
0.007 0.876
34 (9.0) 88 (23.3)
66 (10.9) 184 (30.4)
0.338 0.016
80 (9.0) 261 (29.4)
336 (7.1) 1194 (25.2)
0.045 0.009
34 (10.5) 107 (32.9)
34 (6.8) 111 (22.1)
0.058 0.001
405 (39.6) 46 (4.5) 16.2 (27.4) 6.7 (12.9)
2144 (37.7) 163 (2.9) 15.1 (23.5) 7.4 (12.6)
0.268 0.006 0.257 0.252
132 (34.9) 7 (1.9) 11.3 (16.8) 5.5 (8.2)
205 (33.8) 16 (2.6) 14.9 (22.9) 7.8 (14.8)
0.725 0.426 0.005 0.010
Any postnatal conditions* Use of RSV prophylaxis Characteristics of parents History of asthma History of atopy Characteristics of initial hospitalization NICU use Oxygen use Length of birth hospitalization, mean (SD) Length of initial hospitalization in those who did not have a NICU stay, mean (SD)
⬍0.001
0.002 0.652
0.239
0.001
0.048
RSV ⫽ respiratory syncytial virus; NICU ⫽ neonatal intensive care unit. * Comorbid postnatal conditions include congenital heart disease, chronic lung disease, HIV, trisomy 21, intraventricular hemorrhage, necrotizing enterocolitis, hydrocephalus, periventricular leukomalacia, retinopathy of prematurity, sensorineural hearing loss, and cerebral palsy (codes in Appendix).
2426
Volume 32 Number 14
J.R. Romero et al.
Table II. Postmatch serious early childhood wheezing between second and third birthday.
Parameter SECW Presence of ⱖ1 of the following* ⱖ3 Office, outpatient, or ED visits Office, outpatient, or ED visit ⫹ systemic corticosteroids within 7 days Inpatient stay ⱖ150-Day supply of asthma controller medications
RSV Cohort (n ⫽ 378), No. (%)
Control Cohort (n ⫽ 606), No. (%)
P
63 (16.7)
52 (8.6)
⬍0.001
33 (8.7) 51 (13.5)
20 (3.3) 42 (6.9)
⬍0.001 ⬍0.001
4 (1.1) 14 (3.7)
12 (2.0) 14 (2.3)
0.266 0.201
RSV ⫽ respiratory syncytial virus; SECW ⫽ serious early childhood wheezing; ED ⫽ emergency department. * Visits or inpatient stays with a diagnosis of asthma or wheezing (International Classification of Diseases, Ninth Revision, Clinical Modification22 codes 493.XX or 786.07) were included. Asthma controller medications included oral or inhaled long-acting beta agonists, oral or inhaled corticosteroids, methylxanthines, mast cell stabilizers, leukotriene modifiers, anticholinergics, and combination medications.
SECW (P ⬍ 0.001) during the same period. As shown in Table II, subgroup analysis indicated a significantly greater proportion of patients with 3 or more office, outpatient, or ED visits for wheezing among the RSV cohort (8.7%) than among the control cohort (3.3%) (P ⬍ 0.001). Similarly, the RSV cohort had significantly higher use of systemic steroids for wheezing events (13.5%) than did the control group (6.9%) (P ⬍ 0.001). Therefore, using SECW criteria allowed the study to capture an additional 8% wheezing-related events in the RSV cohort compared with recurrent wheezing (ⱖ3 events) and an additional 5.3% wheezing-related events in the control group. Table III presents the results of the logistic regression model for SECW. Patients with known birth weight were included in the multivariate model (n ⫽ 949). After adjusting for possible confounding variables, patients presenting with RSV infection in the first RSV season had a significant positive association (OR ⫽ 2.52) with SECW between the second and third birthdays (P ⬍ 0.001). In addition, males had higher odds of SECW than females (OR ⫽ 1.80; P ⫽ 0.011), and each additional inpatient day during the birth hospitalization was significantly associated with higher odds of SECW (OR ⫽ 1.01; P ⫽ 0.036). Table IV provides the results of the substudies of patients with known GA ⱕ36 weeks and patients with and without parental atopy or asthma. After stratification of patients based on GA, RSV patients with GA 33
December 2010
to 36 weeks had a significantly higher rate of SECW than matched controls (P ⫽ 0.009). The prevalence of SECW was significantly higher (P ⬍ 0.001) among patients with RSV compared with controls in patients whose parents did not have a history of atopy or asthma. Among patients whose parents had a history of atopy or asthma, the rate of SECW was not significantly different between patients with RSV and controls. Finally, SECW-related and total health care costs (costs for pharmacy- and medical-related services) were evaluated in patients who experienced SECW between their second and third birthdays. During this interval, patients with SECW had a mean SECW-related cost of $1378 (95% CI, $939 –$1816). Of this total, ⬃46% of costs were for pharmacy-filled prescriptions (mean, $630; 95% CI, $471–$789), and the remaining 54% of costs were for medical services (mean, $748; 95% CI, $375–$1121). In contrast, patients without evidence of SECW had a mean SECWrelated cost of $37 (95% CI, $24 –$51). Of the total costs, a mean of $32 was pharmacy related (95% CI, $19 –$45) and $6 was for medical services (95% CI, $2–$10). SECW-related total costs among patients with SECW accounted for nearly 20% of the $7138 (95% CI, $5087–$9189) in total health care costs during this time period. The mean SECW-related pharmacy cost among patients with SECW accounted for 60% of the $1051 (95% CI, $823–$1279) in total
2427
Clinical Therapeutics
Table III. Logistic model of serious early childhood wheezing between second and third birthdays (N ⫽ 949). Variables
Odds Ratio
95% CI
P
2.522
1.652–3.849
⬍0.001
0.903 1.083
0.534–1.529 0.643–1.822
0.704 0.765
0.878 0.751 0.754
0.533–1.445 0.419–1.344 0.357–1.592
0.609 0.334 0.460
1.799
1.142–2.834
0.011
0.822 1.362 0.703
0.507–1.333 0.683–2.718 0.356–1.391
0.427 0.381 0.312
Birth weight, g (⬍1500 g as reference) 1500–⬍2500 ⱖ2500
1.090 0.901
0.505–2.353 0.384–2.110
0.827 0.810
Any postnatal condition* RSV prophylaxis use NICU use during birth hospitalization Length of birth hospitalization
0.764 1.026 1.171 1.013
0.363–1.608 0.626–1.681 0.727–1.885 1.001–1.024
0.479 0.919 0.517 0.036
RSV cohort Seasonal year of birth (2001 as reference) 2002 2003 Quarter of birth (May–July as reference) August–October November–January February–April Male sex Health plan geographic region (South as reference) Midwest Northeast West
RSV ⫽ respiratory syncytial virus; NICU ⫽ neonatal intensive care unit. * Comorbid postnatal conditions included congenital heart disease, chronic lung disease, HIV, trisomy 21, intraventricular hemorrhage, necrotizing enterocolitis, hydrocephalus, periventricular leukomalacia, retinopathy of prematurity, sensorineural hearing loss, and cerebral palsy (codes in the Appendix).
pharmacy-related costs, and SECW-related medical costs accounted for 12% of the $6087 (95% CI, $4165–$8009) in total medical costs. In contrast, patients without evidence of SECW had a mean total health care cost of $2521 (95% CI, $1789 –$3253), with $164 in pharmacy costs and $2357 in medical costs. After adjusting for possible confounders, patients with SECW had a significantly higher total health care cost than patients without evidence of SECW (P ⬍ 0.001). Patients with SECW had adjusted total costs that were 2.67-fold higher than did patients without evidence of SECW (data not shown).
DISCUSSION The primary end point, SECW, was developed to capture clinically meaningful wheezing in addition to determine the frequency of wheezing illnesses. Perhaps
2428
more important than the additional number of events were the types of events that this definition of SECW captured. These additional events consisted of those requiring the use of asthma-control medications for ⱖ5 months, those requiring systemic steroids for acute exacerbations, and those needing hospitalization. These additional events are considered clinically meaningful because of the inherent severity of wheezing disease that necessitates such clinical management. Therefore, in this study, the authors conclude that SECW is a more clinically relevant and accurate end point that represents airway morbidity during early childhood years and measures not only the frequency but also the severity of wheezing. Using this definition, an association was found between RSV-LRI occurring in infancy and airway morbidity between 2 and 3 years of age. Even after adjust-
Volume 32 Number 14
J.R. Romero et al.
Table IV. Rates of serious early childhood wheezing between second and third birthdays stratified by other variables of interest. Variable
RSV Cohort, No. (%)
Control Cohort, No. (%)
P
Gestational age ⱕ32 wk* 33–36 wk
4/20 (20.0) 22/118 (18.6)
4/50 (8.0) 12/158 (7.6)
Parental atopy Yes* No
16/107 (15.0) 40/218 (18.3)
9/111 (8.1) 32/392 (8.2)
0.113 ⬍0.001
Parental asthma Yes* No
4/34 (11.8) 52/291 (17.9)
3/34 (8.8) 38/469 (8.1)
1.000 ⬍0.001
0.213 0.009*
RSV ⫽ respiratory syncytial virus. * RSV and control patients were matched overall and matches may not appear in the same stratum.
ing for different covariates using a logistic model, this association was found to be significant (P ⬍ 0.001). Stratification of the analysis for preterm infants with known GA showed differences similar to those observed for the overall group. However, the number of children analyzed was too small to achieve statistical significance. Similarly, the number of evaluable patients decreased substantially when the data were stratified to explore the role of parental history of atopy or asthma on the association of RSV-LRI and presence of wheezing. An association between RSV-LRI in infancy and later wheezing remained significant in patients with no family history of asthma, atopy, or both. Thus, the current data did not further the understanding of the role of parental atopy–asthma and its association with RSV-LRI in infancy and subsequent SECW. It is expected that a larger sample size would allow the stratification necessary to address the role of family history of asthma or atopy in development of SECW even beyond the 3 years investigated in this analysis. This study did, however, provide some insight into the effects of SECW on direct costs. Patients with SECW had higher total health care costs compared with those who did not have SECW. The costs contributed by SECW were between $1300 and $1400 per study subject. Annual direct asthma-related costs in pediatric patients aged 4 to 17 years have been reported to range from $393 to $1316, based on treatments used.26,27 The SECW-related costs in children
December 2010
aged 2 to 3 years appear to be comparable, suggesting that in a younger patient cohort, SECW could place a financial burden on health plan resources similar to that associated with asthma in older children. A systematic review by Pérez-Yarza et al9 concluded that RSV-LRI was associated with an increased risk for subsequent development of asthma or recurrent wheezing and that this association became progressively smaller with increasing age. Using an end point that considers frequency and severity, the present study in preterm infants adds to this body of literature and supports the association between RSV-LRI in infancy and airway hyperreactivity in later life. The study also found associations between variables such as male sex and longer length of birth hospitalization and increased airway morbidity in later life. As summarized by Meissner and Long,11 multiple theories have been proposed in an effort to understand the association between RSV-LRI in infancy and subsequent recurrent wheezing and asthma. In the causal theory, it is believed that RSV infection occurring during the final stage of lung development (ie, the first 2–3 years of life) alters normal development and that the resulting long-term consequences on lung structure and function predispose these infants to future episodes of wheezing.11 Another possibility is that in certain infants, innate preexisting abnormalities of airway function or immune response are triggered by an early viral respiratory tract infection and lead to airway ob-
2429
Clinical Therapeutics struction.11 A complex multifactorial theory holds that no single response to a viral respiratory tract infection determines outcome.11 Instead, the long-term outcome is a product of multiple factors, such as the infant’s genetic makeup, concomitant exposure to environmental antigens, and state of maturation of the immune system and airway at the time of infection. In this theory, no single factor dominates any other in determining outcome. Although the strongest association between viral lower respiratory tract infection and subsequent development of wheezing has been noted with RSV, there have been reports of other viral infections, such as rhinovirus, influenza virus, and parainfluenza virus, that triggered subsequent wheezing.28 –30 One approach to delineating whether the pathway is causal or RSV-specific would be to demonstrate a reduction in long-term airway morbidity by reducing RSV in infancy. Although some investigations of RSV immunoprophylaxis support the hypothesis that modification of RSV infection in premature infants decreases risk for subsequent recurrent wheezing, a well-designed randomized study is needed to establish the association.14,15 The findings from this study must be considered in light of several limitations. Thirty-six months of continuous enrollment in the health plan was required. Although continuous enrollment is often required for retrospective analyses of health care outcomes, it can be problematic when including patients with potentially fatal illnesses, such as RSV. In this analysis, patients who disenrolled from the health plan before age 3 years for any reason (including death) were excluded. The analysis included infants with RSV-like illnesses for whom a diagnosis of RSV infection could be confirmed. Some of the infants in the control group may have experienced mild RSV infections that did not require medical attention and were misclassified, thereby underestimating the effect on study outcomes. The data used in this study came from a commercial MCO population. Therefore, the findings from this analysis are primarily applicable to the management of RSV infections in premature and low-birth-weight infants in similar managed care settings and may not be applicable to a Medicaid plan. Although propensity score matching and multivariate methods allowed for the selection of a comparable control group of infants, complete matching was not attained. These methods cannot adjust for imbalances due to unobserved covariates. All limitations associated with the use of medi-
2430
cal claims data (ie, coding errors, undercoding, and capture of only direct costs) are applicable.
CONCLUSIONS The development of RSV-LRI in infancy in preterm infants was associated with an increased prevalence of SECW between ages 2 and 3 years. Patients with SECW had higher total health care costs compared with those who did not have SECW.
ACKNOWLEDGMENTS Funding for this research was provided by MedImmune, LLC. The authors acknowledge Physicians World, Seacaucus, New Jersey, and Chris A. Kirk, PhD, and Gerard P. Johnson, PhD, Complete Healthcare Communications, Chadds Ford, Pennsylvania, for providing editing support funded by MedImmune. Drs. Romero and Stewart provide consultant services to MedImmune. Dr. Romero has received funding from MedImmune, has been a member of MedImmune’s speakers’ bureau, and has participated in clinical studies sponsored by Novartis Pharmaceuticals Corporation, GlaxoSmithKline, Roche, and Johnson & Johnson Services, Inc. Dr. Stewart has been a member of speakers’ bureaus for Ikaria, Inc. and Abbott Nutrition and a member of an advisory board with Abbott International. Ms. Buysman is employed by i3 Innovus, which was contracted by MedImmune to do the analysis. Drs. Fernandes, Jafri, and Mahadevia are employees of MedImmune. The authors have indicated that they have no other conflicts of interest regarding the content of this article.
REFERENCES 1. Hall CB. Respiratory syncytial virus. In: Feigen RD, Cherry JD, eds. Textbook of Pediatric Infectious Diseases. 2nd ed. Philadelphia, Pa: Saunders; 1987:1653–1675. 2. Hall CB, Long CE, Schnabel KC. Respiratory syncytial virus infections in previously healthy working adults. Clin Infect Dis. 2001;33:792–796. 3. Hall CB, Weinberg GA, Iwane MK, et al. The burden of respiratory syncytial virus infection in young children. N Engl J Med. 2009;360:588 –598. 4. Institute of Medicine. Prospects for Immunizing Against Respiratory Synctial Virus. Washington, DC: National Academy Press; 1985. 5. Ruuskanen O, Ogra PL. Respiratory syncytial virus. Curr Probl Pediatr. 1993;23:50 –79.
Volume 32 Number 14
J.R. Romero et al. 6. Shay DK, Holman RC, Newman RD, et al. Bronchiolitis-associated hospitalizations among US children, 1980 –1996. JAMA. 1999;282: 1440 –1446. 7. Leader S, Kohlhase K. Respiratory syncytial virus– coded pediatric hospitalizations, 1997 to 1999. Pediatr Infect Dis J. 2002;21:629 – 632. 8. Bont L, Aalderen WM, Kimpen JL. Long-term consequences of respiratory syncytial virus (RSV) bronchiolitis. Paediatr Respir Rev. 2000;1:221– 227. 9. Pérez-Yarza EG, Moreno A, Lázaro P, et al. The association between respiratory syncytial virus infection and the development of childhood asthma: A systematic review of the literature. Pediatr Infect Dis J. 2007;26: 733–739. 10. Mohapatra SS, Boyapalle S. Epidemiologic, experimental, and clinical links between respiratory syncytial virus infection and asthma. Clin Microbiol Rev. 2008;21:495–504. 11. Meissner HC, Long SS. Respiratory syncytial virus infection and recurrent wheezing: A complex relationship. J Pediatr. 2007;151:6 –7. 12. Simões EA. RSV disease in the pediatric population: Epidemiology, seasonal variability, and long-term outcomes. Manag Care. 2008;17:3– 6, discussion 18 –19. 13. Respiratory Syncytial Virus. In: Pickering LK, Baker CJ, Long SS, McMillan JA, eds. Red Book: 2006 Report of the Committee on Infectious Diseases. 27th ed. Elk Grove Village, Ill: American Academy of Pediatrics; 2006: 560 –566. 14. Simoes EA, Groothuis JR, CarbonellEstrany X, et al, for the Palivizumab Long-Term Respiratory Outcomes Study Group. Palivizumab prophylaxis, respiratory syncytial virus, and subsequent recurrent wheezing. J Pediatr. 2007;151:34 – 42, 42.e1. 15. Wenzel SE, Gibbs RL, Lehr MV, Simoes EA. Respiratory outcomes in high-risk children 7 to 10 years after prophylaxis with respiratory syncy-
December 2010
16.
17.
18.
19.
20.
21.
22.
tial virus immune globulin. Am J Med. 2002;112:627– 633. Henderson J, Hilliard TN, Sherriff A, et al. Hospitalization for RSV bronchiolitis before 12 months of age and subsequent asthma, atopy and wheeze: A longitudinal birth cohort study. Pediatr Allergy Immunol. 2005; 16:386 –392. Sigurs N, Bjarnason R, Sigurbergsson F, Kjellman B. Respiratory syncytial virus bronchiolitis in infancy is an important risk factor for asthma and allergy at age 7. Am J Respir Crit Care Med. 2000;161:1501–1507. Sigurs N, Gustafsson PM, Bjarnason R, et al. Severe respiratory syncytial virus bronchiolitis in infancy and asthma and allergy at age 13. Am J Respir Crit Care Med. 2005;171:137– 141. Stein RT, Sherrill D, Morgan WJ, et al. Respiratory syncytial virus in early life and risk of wheeze and allergy by age 13 years. Lancet. 1999;354:541– 545. Sigurs N, Bjarnason R, Sigurbergsson F, et al. Asthma and immunoglobulin E antibodies after respiratory syncytial virus bronchiolitis: A prospective cohort study with matched controls. Pediatrics. 1995; 95:500 –505. National Heart, Lung, and Blood Institute, US Department of Health and Human Services. National Asthma Education and Prevention Program. Expert panel report 3: Guidelines for the Diagnosis and Management of Asthma. http://www.nhlbi.nih.gov/ guidelines/asthma/asthgdln.pdf. Accessed January 20, 2011. World Health Organization. International Statistical Classification of Diseases and Related Health Problems, Ninth Revision, Clinical Modification. Geneva, Switzerland: WHO; 2003.
23. Committee on Infectious Diseases. From the American Academy of Pediatrics: Policy statements—Modified recommendations for use of palivizumab for prevention of respiratory syncytial virus infections. Pediatrics. 2009;124:1694 –1701. 24. Bureau of Labor Statistics, US Department of Labor. Consumer Price Index, Chained Consumer Price Index 2006. http://www.bls.gov/ data/. Accessed January 20, 2011. 25. Manning WG. The logged dependent variable, heteroscedasticity, and the retransformation problem. J Health Econ. 1998;17:283–295. 26. Stempel DA, Kruzikas DT, Manjunath R. Comparative efficacy and cost of asthma care in children with asthma treated with fluticasone propionate and montelukast. J Pediatr. 2007;150:162–167. 27. Stempel DA, Riedel AA, Carranza RosenzweigJR.Resourceutilizationwith fluticasone propionate and salmeterol in a single inhaler compared with other controller therapies in children with asthma. Curr Med Res Opin. 2006;22: 463–470. 28. ErikssonM,BennetR,NilssonA.Wheezing following lower respiratory tract infections with respiratory syncytial virus and influenza A in infancy. Pediatr Allergy Immunol. 2000;11:193–197. 29. Korppi M, Kotaniemi-Syrjänen A, Waris M, et al. Rhinovirus-associated wheezingininfancy:Comparisonwithrespiratory syncytial virus bronchiolitis. Pediatr InfectDisJ.2004;23:995–999. 30. Lemanske RF Jr. Viruses and asthma: Inception, exacerbation, and possible prevention. J Pediatr. 2003;142(Suppl 2):S3–S7; discussion S7–S8. 31. American Medical Association (AMA). CurrentProceduralTerminology:2001.Chicago, Ill: AMA; 2000.
Address correspondence to: José R. Romero, MD, Arkansas Children’s Hospital, 1 Children’s Way, Slot 512-11, Little Rock, AR 72202-3591. E-mail: [email protected]
2431
Clinical Therapeutics
Appendix. Codes and medications used to define comorbid conditions of interest. Comorbid Condition Congenital heart disease*
Code Type ICD-9-CM diagnosis Prescription
CPT ICD-9-CM procedure † Chronic lung disease ICD-9-CM diagnosis Prescription HIV ICD-9-CM diagnosis Trisomy 21 ICD-9-CM diagnosis Intraventricular hemorrhage ICD-9-CM diagnosis Necrotizing enterocolitis ICD-9-CM diagnosis Hydrocephalus ICD-9-CM diagnosis Periventricular leukomalacia ICD-9-CM diagnosis Retinopathy of prematurity ICD-9-CM diagnosis Sensorineural hearing loss ICD-9-CM diagnosis Cerebral palsy ICD-9-CM diagnosis
Codes 425.3, 425.4, 425.8, 745.XX, 746.XX, and 747.XX Digoxin, diuretics, ACE inhibitors, angiotensin receptor blockers, -blockers, vasodilators, combination medications 92992–92993 35.41 770.7 Bronchodilators, diuretics, diuretic combinations 042, V08, 079.51–079.53 758.0 772.1x 557.0, 777.5 741.0X, 742.3 779.7 362.21, 362.29 389.1X, 389.2 343.X, 333.71
ICD-9-CM ⫽ International Classification of Diseases, Ninth Revision, Clinical Modification22; ACE ⫽ angiotensin-converting enzyme; CPT ⫽ Current Procedural Terminology.31 * Defined as claims meeting ⱖ1 of the following criteria: (1) diagnosis code listed above and medication listed in the first 6 months of life or (2) CPT code or procedure code listed above in the first 6 months of life. † Defined as claims for 1 of the diagnostic codes listed above and claims for 1 of the medications listed in the first 30 days of life.
2432
Volume 32 Number 14