Abstract
Objective:
Healthcare use and costs within 1 year of a respiratory syncytial virus lower respiratory tract infection (RSV-LRI) among Medicaid early-preterm and late-preterm infants compared with full-term infants were evaluated.
Methods:
Infants born during 2003–2005 were identified from the Thomson Reuters MarketScan Multi-State Medicaid Database. Infants <1 year of age were grouped based on RSV-LRI and unspecified bronchiolitis/pneumonia (UBP) diagnosis codes and stratified by inpatient or outpatient setting. Infants without RSV-LRI/UBP were selected for comparison. Economic and clinical outcomes were analyzed descriptively; the relationship between RSV-LRI/UBP and costs incurred within 1 year of infection were analyzed using logged ordinary least squares models. Results were stratified by gestational age.
Results:
Most infants were diagnosed with RSV-LRI/UBP after 90 days of chronologic age. Early-preterm infants had the greatest mean number of inpatient, outpatient, and emergency department visits after an RSV-LRI/UBP episode. The marginal costs among infants with RSV-LRI compared with controls were $34,132 (p < 0.001) and $3869 (p = 0.115) among inpatients and outpatients, respectively. Among late-preterm infants, the marginal costs were $17,465 (p < 0.001) and $2158 (p < 0.001) among inpatients and outpatients, respectively. Full-term infants had the lowest marginal costs (inpatients, $9151 [p < 0.001]; outpatients, $1428 [p < 0.001]). Overall, inpatient infants with RSV-LRI/UBP had higher costs than outpatients, suggesting that increased downstream costs are associated with severity of RSV-LRI/UBP disease.
Limitations:
Infants with unknown etiology for bronchiolitis were assigned to the UBP group, which may underestimate the costs of the comparison group.
Conclusions:
The burden of RSV-LRI was substantial among early-preterm Medicaid infants. Costs were also higher among late-preterm relative to full-term infants.
Introduction
The Institute of Medicine noted the paucity of research on the long-term consequences of premature birthCitation1. Studies have documented increased healthcare costs among premature infants compared with full-term infantsCitation2. Late-preterm infants experience a wide range of morbidity, including respiratory distressCitation3.
Late-preterm infants are at high risk for serious respiratory syncytial virus (RSV) disease, possibly attributable to underdeveloped lungs, immature immune system, and incomplete transfer of maternal antibodiesCitation4,Citation5. Late-preterm infants have higher costs associated with the initial RSV hospitalization compared with full-term infantsCitation6, but more evidence is needed in late-preterm infants regarding the healthcare use and costs during the period after the RSV hospitalization and during early childhood.
Currently, there are no published studies that examine the healthcare burden assumed after an initial RSV lower respiratory tract infection (RSV-LRI) among infants of various gestational ages in a Medicaid population. We hypothesized that infants with RSV-LRI would have higher use and costs within 12 months of an initial RSV-LRI episode relative to a comparison group of infants without RSV-LRI, and we expected to observe a greater burden in the early-preterm and late-preterm groups relative to the full-term group.
Patients and methods
We conducted a retrospective cohort study using the MarketScan Multi-State Medicaid Database (Thomson Reuters, New York, NY, USA). A total of 586,062 infants younger than 1 year were identified between January 1, 2003 and December 31, 2005. Infants were excluded if they did not have a minimum of 12 months of data available (n = 377,084), a birth record (n = 8098), or had a capitated health plan (n = 153,539). The final sample included 47,341 infants who were classified as full-term (≥37 weeks’ gestational age [wGA]), late-preterm (33–36 wGA), and early-preterm (<33 wGA) using diagnostic related groups and International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnoses codesCitation7. The MarketScan Research Databases include persons from all US states and are de-identified and fully compliant with the Health Insurance Portability and Accountability Act of 1996 (HIPAA). Because this study did not involve the collection, use, or transmittal of individually identifiable data, institutional review board (IRB) review or approval was not requiredCitation8,Citation9.
Infants were assigned to the RSV-LRI group if they had an ICD-9-CM diagnosis for RSV, acute bronchiolitis due to RSV, or pneumonia due to RSV. Infants were assigned to the unspecified bronchiolitis/pneumonia (UBP) group if they had acute bronchitis, acute bronchiolitis, and pneumonia/bronchopneumonia due to other infectious or unspecified organisms during a typical RSV season (November through April). The RSV-LRI/UBP groups were further stratified based on the healthcare setting of the index RSV-LRI/UBP episode, resulting in mutually exclusive subgroups: RSV-LRI Inpatient, UBP Inpatient, RSV-LRI Outpatient (diagnosis in physician’s office or emergency department [ED]) and UBP Outpatient (). Infants without RSV-LRI/UBP were selected as the comparison group.
Table 1. Healthcare use* after RSV-LRI/UBP in full-term, late-preterm, and early-preterm Medicaid infants.
The study periods included a variable length pre-period (between birth and the index date) and an observational 12-month period starting with the RSV-LRI/UBP event. Demographic and clinical characteristics were evaluated at birth, including gender, birth quarter, race, population density, presence of a neonatal intensive care unit (NICU) admission during birth hospitalization, length of stay (LOS) of birth hospitalization, birth weight, presence of chronic lung disease of prematurity (CLD), congenital heart disease (CHD) or other congenital conditions, including expensive conditions such as bronchopulmonary dysplasia, patent ductus arteriosus, and necrotizing enterocolitis. Because comparison infants did not have an RSV-LRI/UBP event, an index date was assigned based on the mean length of time between the date of the birth hospitalization and the date of the initial episode among children in the RSV-LRI/UBP cohort. Outcomes among the comparison infants were observed within 12 months of this date.
The primary outcomes included healthcare costs and mean number of all-cause inpatient admissions, ED visits, and outpatient physician office visits during the 12 months after the index event. Healthcare costs included costs for inpatient services, outpatient services, and outpatient prescriptions. Outpatient services consisted of ED and physician office visits and laboratory or other outpatient services. Full-year costs were defined as total costs during the 12-month period, including the cost of the initial RSV-LRI/UBP episode and follow-up costs were defined as total costs during the 12-month period, excluding the cost of the initial RSV-LRI/UBP episode as part of sensitivity analyses. Marginal costs represent the incremental difference in full-year or follow-up costs between the RSV-LRI/UBP infants and the comparison infants. Costs were based on paid healthcare claims and were adjusted to 2006 dollars using the Consumer Price Index.
Propensity score weighting was employed to control for confounding. Multinomial logistic regression modeling was performed to calculate a generalized propensity score which indicated the conditional probability of being diagnosed as RSV-LRI inpatient or outpatient, UBP inpatient or outpatient, or comparison. Gender, birth month, birth weight, state, presence of CLD or CHD, NICU admission during the birth hospitalization, and the costs of the birth hospitalization were included as covariates. The propensity score weight – the inverse probability that an infant was assigned to a specific subgroup – was calculated for each infant and then normalized to reflect the sample size of the respective condition/gestational age subgroup. Propensity score weights were used to adjust the descriptive and multivariate analysesCitation10–12.
Logged ordinary least squares models were conducted to examine the healthcare costs. Separate models were performed for each gestational age cohort. The choice of the log normal model versus generalized linear models was based on examination of skewness/kurtosis for each outcome. Gender, race, birth year, birth quarter, birth weight, population density, presence of CLD or CHD, presence of other congenital conditions, NICU admission during the birth hospitalization, length of birth hospitalization stay, and palivizumab use between the birth hospitalization discharge date and the day before the infants’ first birthday were included. In all models, the infants without RSV-LRI/UBP served as the reference group. No adjustment was made for multiple hypothesis testing and statistical significance was declared if the 2-sided p-value was <0.05.
Results
The RSV-LRI/UBP infants were mostly older than 90 days chronologic age at the time of their initial RSV-LRI/UBP diagnoses (). In the late-preterm cohort, 44.2%, 71.9%, 70.5%, and 81.1% were older than 90 days chronologic age at the time of the RSV-LRI inpatient, RSV-LRI outpatient, UBP inpatient, and UBP outpatient visit, respectively. Full-term RSV LRI/UBP infants were approximately 3 months of age at index (range: 2.25–4.00 months) while late-preterm and early-preterm infants were slightly older (approximately 4 and 5 months of age, respectively) at index. Males were slightly overrepresented in the study sample. Infants were predominantly Caucasian (60%); approximately 18% of the infants were African-American in the preterm cohorts.
Preterm infants with RSV-LRI/UBP had significantly higher all-cause inpatient admissions, ED visits, and outpatient office visits relative to full-term infants. Approximately 71% of early-preterm infants had an ED visit for any reason (mean visit range: 5.3–8.3 days across subgroups) compared with 49% of comparison infants (mean =3.1 days; p < 0.001; ). Inpatient and outpatient late-preterm infants with RSV-LRI/UBP also had significantly more ED visits (range: 3.6–6.2 visits) and outpatient physician office visits (range: 7.6–9.6 visits) vs. comparison infants (p < 0.001). The all-cause hospitalization rate for RSV-LRI/UBP outpatient late-preterm infants ranged from 13% to 18% vs. 8% in the comparison cohort (p < 0.001). Full-term RSV-LRI/UBP infants had significantly greater healthcare use than those without RSV-LRI/UBP; however, healthcare use among full-term infants was generally reduced in comparison with early-preterm and late-preterm infants.
Stratified analyses by gestational age revealed that, while the early-preterm infants had the highest full-year costs across subgroups, late-preterm infants also had higher healthcare costs compared with full-term infants (). Among RSV-LRI inpatient late-preterm infants, full-year costs were $22,288 vs. 4823 among comparison infants (marginal burden =$17,465; p < 0.001). In contrast, full-year costs among full-term infants hospitalized with RSV-LRI were $11,007 vs. 1856 among comparison infants, (marginal burden =$9151; p < 0.001).
Table 2. Healthcare costs after RSV-LRI/UBP in full-term, late-preterm, and early-preterm Medicaid infants.
A similar trend of higher full-year costs among late-preterm versus full-term infants was also observed in the outpatient setting. Late-preterm infants with outpatient RSV-LRI had full-year costs of $6981, which was significantly higher than those of comparison infants by $2158 (p < 0.001). In contrast, full-term infants with outpatient RSV-LRI had full-year costs of $3285, which was significantly higher than those of comparison infants by $1428 (p < 0.001).
The mean follow-up costs for the sensitivity analyses remained significantly higher among infants with RSV-LRI/UBP than comparison infants across the various gestational age cohorts. However, there were no significant differences in mean follow-up costs among early-preterm infants with RSV-LRI/UBP in the outpatient setting compared with costs among comparison infants.
Conclusions
Our study found substantial healthcare resource use and cost across Medicaid infants of various gestational ages with RSV-LRI/UBP. Among all infants, early-preterm infants had the highest mean number of inpatient, outpatient, and ED visits and costs after an RSV-LRI/UBP episode. Our finding among early-preterm infants is consistent with previous work in infants <32 wGA with CLD who were hospitalized with RSV and had considerable subsequent healthcare use and costs during early childhoodCitation13,Citation14.
Representing another high-risk group for severe RSV disease, late-preterm infants in our study were also associated with significant healthcare costs. A previous study reported a significantly higher cost of RSV hospitalization among late-preterm versus full-term infantsCitation6. Our study further found differences in healthcare use and costs between late-preterm infants and full-term infants within 12 months of an initial RSV-LRI/UBP episode.
Our study also contributes to our understanding of the consequences of RSV infection by addressing the burden of RSV disease by treatment settingCitation15. Regardless of gestational age, infants with RSV-LRI/UBP in the inpatient setting had higher costs than those diagnosed in the outpatient setting, suggesting that higher downstream costs are associated with increased severity of RSV-LRI/UBP disease. However, the burden of outpatient RSV-LRI/UBP should also be recognized because the majority of infants with RSV-LRI/UBP infection are cared for in the outpatient settingCitation15.
RSV-LRI/UBP disease was associated with follow-up costs in which we excluded the costs of the initial RSV-LRI/UBP episode. In this sensitivity analyses, we found significantly higher costs during the 12-month observational period among infants with RSV-LRI/UBP compared with infants without RSV-LRI/UBP. This finding is consistent with previous studies that found increased use among infants 32–35 wGA with RSV up to 2 years after the initial hospital admissionCitation16.
Our study describes the economic burden beyond the initial RSV-LRI/UBP episode in a Medicaid population across early-preterm, late-preterm, and full-term infants. Our analyses included a cohort of infants with UBP because up to 80% of bronchiolitis cases can be caused by infection with RSVCitation17. Results from the comparison group may be underestimated because infants with unknown etiology for bronchiolitis were placed in a separate group, resulting in a potentially healthier comparison group. Study findings did not control for potential confounders such as environmental risk factors associated with RSV and patient biometric characteristics because they are not readily available from medical claims databases. Nevertheless, our study used propensity score weighting in an attempt to control for selection bias and found that late-preterm infants have higher all-cause use and costs within 12 months after the initial RSV-LRI/UBP episode relative to full-term infants with RSV-LRI. Because late-preterm infants represent approximately 72% of all premature infantsCitation18, our findings have important implications for understanding the economic burden associated with RSV disease in this particular high-risk group.
Transparency
Declaration of funding
This study was funded by MedImmune, LLC, Gaithersburg, MD, USA.
Declaration of financial/other relationships
N.S., L.P., and B.-C. C. have disclosed that they are employees of Thomson Reuters and provide custom consulting services to all major pharmaceutical companies as a condition of employment. As part of a consulting agreement with Thomson Reuters, MedImmune provided funding to Thomson Reuters to support the data collection, analysis, and manuscript development activities associated with this manuscript. C.B.H. and J.P.K. have served as consultants for MedImmune. A.S.M. and P.J.M. have disclosed that they are employees of MedImmune.
Acknowledgments
The authors wish to acknowledge the contributions of Robert Sedgley, who served as the primary SAS programmer. We would also like to thank Kimmie McLaurin for her assistance during the study design phase. Formatting and editorial assistance was provided by Complete Healthcare Communications, Inc. (Chadds Ford, PA, USA) and funded by MedImmune.
References
- Institute of Medicine. Preterm birth: Causes, consequences, and prevention. Washington, DC: National Academies Press, 2006
- McLaurin KK, Hall CB, Jackson EA, et al. Persistence of morbidity and cost differences between late-preterm and term infants during the first year of life. Pediatrics 2009;123:653-9
- Engle WA, Tomashek KM, Wallman C. “Late-preterm” infants: a population at risk. Pediatrics 2007;120:1390-401
- Boyce TG, Mellen BG, Mitchel EF, Jr et al. Rates of hospitalization for respiratory syncytial virus infection among children in Medicaid. J Pediatr 2000;137:865-70
- Coffman S. Late preterm infants and risk for RSV. MCN Am J Matern Child Nurs 2009;34:378-84; quiz 85-6
- Forbes ML, Hall CB, Jackson A, et al. Comparative costs of hospitalisation among infants at high risk for respiratory syncytial virus lower respiratory tract infection during the first year of life. J Med Econ 2010;13:136-41
- International Classification of Diseases, 9th Revision, Clinical Modification, 4th edn. Practice Management Information Corporation, 2006
- HIPAA privacy rule and public health. Guidance from CDC and the U.S. Department of Health and Human Services. MMWR Morb Mortal Wkly Rep 2003;52(Suppl):1-17, 9-20
- Department of Health and Human Services. Code of Federal Regulations. Title 45 – Public Welfare. Part 46 Protection of Human Subjects. Subpart A – Basic HHS Policy for Protection of Human Research Subjects. Available at: http://ohsr.od.nih.gov/guidelines/45cfr46.html. [Accessed January 25, 2011]
- Imai K, van Dyk DA. Causal inference with general treatment regimes. J Am Stat Assoc 2004;99:854-66
- Imbens G. The role of the propensity score in estimating dose-response functions. Biometrika 2000;87:706-10
- Imbens GW. Nonparametric estimation of average treatment effects under exogeneity: a review. Rev Econ Stat 2004;86:4-29
- Greenough A, Alexander J, Burgess S, et al. Health care utilisation of prematurely born, preschool children related to hospitalisation for RSV infection. Arch Dis Child 2004;89:673-8
- Greenough A, Cox S, Alexander J, et al. Health care utilisation of infants with chronic lung disease, related to hospitalisation for RSV infection. Arch Dis Child 2001;85:463-8
- Paramore LC, Ciuryla V, Ciesla G, et al. Economic impact of respiratory syncytial virus-related illness in the US: an analysis of national databases. Pharmacoeconomics 2004;22:275-84
- Sampalis JS. Morbidity and mortality after RSV-associated hospitalizations among premature Canadian infants. J Pediatr 2003;143:S150-6
- Hall CB. Respiratory syncytial virus. In Feigin RD, Cherry JD, eds. Textbook of Pediatric Infectious Diseases, 4th edn. Philadelphia, PA: WB Saunders, 1998 2084-110
- Martin JA, Hamilton BE, Sutton PD, et al. Births: final data for 2006. Natl Vital Stat Rep 2009;57:1-104