ARTICLE |
a Department of Emergency Medicine, Massachusetts General Hospital, Boston, Massachusetts
b Department of Medicine, Boston Children's Hospital, Boston, Massachusetts
| ABSTRACT |
|---|
|
|
|---|
PATIENTS AND METHODS. We analyzed the 2002 Health Care Utilization Project-National Inpatient Sample, a federal, stratified random survey of hospital discharges. For admissions age <2 years with a discharge diagnosis of bronchiolitis (International Classification of Diseases, Ninth Revision, Clinical Modification, code 466.1), we used nationally representative weighted estimates to determine frequency and total hospital charges. Costs were estimated from reported charges by applying hospital-specific cost/charge ratios based on all-payer inpatient cost.
RESULTS. In 2002, an estimated 149000 patients were hospitalized with bronchiolitis. Frequency of hospitalizations was higher among children age <1 year of age, male gender, and nonwhite race. Mean length of stay was 3.3 days. Total annual costs for bronchiolitis-related hospitalizations were $543 million, with a mean cost of $3799 per hospitalization. Mean cost of bronchiolitis with a codiagnosis of pneumonia was $6191. In a multivariate analysis controlling for 3 confounding factors (including length of stay), cost per hospitalization was higher for children
1 year and lower for those in the South versus Northeast.
CONCLUSIONS. Bronchiolitis admissions cost more than $500 million annually. A codiagnosis of bronchiolitis and pneumonia almost doubles the cost of the hospitalization. Inpatient health care costs of bronchiolitis are higher than estimated previously and highlight the need for initiatives to safely reduce bronchiolitis hospitalizations and thereby decrease health care costs.
Key Words: bronchiolitis cost HCUP
Abbreviations: RSVrespiratory syncytial virus HCUPHealth Care Utilization Project NISNational Inpatient Sample ICD-9International Classification of Diseases, Ninth Revision ICD-9-CMInternational Classification of Diseases, Ninth Revision, Clinical Modification CCSClinical Classifications Software CCRcost/charge ratio CIconfidence interval LOSlength of stay
Bronchiolitis is the leading cause of hospitalization for infants <1 year of age.15 Respiratory syncytial virus (RSV), the most common cause of bronchiolitis, affects almost all children by the age of 2 years.6 Although hospitalization costs for bronchiolitis are presumed to be high, previous research has focused on RSV bronchiolitis instead of bronchiolitis as a clinical diagnosis. In addition, there are no recent epidemiological data on bronchiolitis hospitalizations among children <2 years.5
Estimating bronchiolitis hospitalization costs from RSV-related bronchiolitis may underestimate the true cost of bronchiolitis hospitalizations. Although RSV infection is the most common cause of bronchiolitis,24 many other infectious agents cause bronchiolitis, including parainfluenza viruses, influenza virus, adenoviruses, and
4 other infectious agents.717 In addition, previous RSV hospitalization expenditure estimates included all of the RSV-related diagnoses, including RSV-associated pneumonia, and these studies reported hospital charges as opposed to costs based on hospital-specific cost/charge ratios.2,18,19 Lastly, there are several articles reporting hospitalization expenditures because of bronchiolitis hospitalizations in other countries and some single-site studies but none that are representative of all US hospitalizations.2028
Using the Health Care Utilization Project (HCUP), Nationwide Inpatient Sample (NIS) database, we investigated the number of US hospitalizations and associated costs for bronchiolitis from 2002. Estimates were calculated for all of the visits in which bronchiolitis was one of the diagnoses, bronchiolitis was a primary diagnosis only, and then subsets of bronchiolitis hospitalizations according to the presence (or absence) of pneumonia or asthma. Finally, we calculated age-specific estimates of bronchiolitis cost.
| PATIENTS AND METHODS |
|---|
|
|
|---|
Study Protocol
The 2002 NIS data were drawn from 35 participating states. A sampling frame of 3570 hospitals was stratified by geographic region, ownership control, teaching status, and bed size. Up to 20% of hospitals were randomly selected from within each stratum. Patient level clinical and resource use information typically available in a discharge abstract were included for all of the discharges from selected hospitals for the sample year. The final sample included 7853982 discharges from 995 hospitals.
Measurements and Patients
Each discharge abstract included age, gender, race/ethnicity, primary payer, admission source, length of stay, disposition, and
15 procedures and
15 discharge diagnoses. Procedures and diagnoses were available using both International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes and the Clinical Classifications Software (CCS), which is a methodology developed by the Agency for Healthcare Research and Quality to group International Classification of Diseases, Ninth Revision (ICD-9) codes into clinically sensible and mutually exclusive categories. Because the CCS system grouped bronchiolitis and bronchitis into one code (CCS code 125), we used ICD-9-CM code 466.1 in any of the 15 diagnoses fields to identify cases. We examined the most common procedures performed using the ICD-9-CM procedure codes. A portion of the analysis excluded hospitalizations because of routine births. These were identified using CCS code 218 ("liveborn") in the first field. We also identified bronchiolitis cases with a codiagnosis of pneumonia (CCS code 122) and asthma (CCS code 128). When bronchiolitis was not the primary diagnosis, pneumonia or asthma was the primary diagnosis for
70% of cases. Bronchiolitis is a clinical diagnosis, and as a result, there will be practice variation diagnosing bronchiolitis, potential difficulty distinguishing clinically between bronchiolitis and asthma in children < 2 years, and an overlap with pneumonia. Case subjects were identified when bronchiolitis was the primary diagnosis or, for the purposes of analysis, when bronchiolitis was included in any of the 15 diagnoses fields. Only children < 2 years were eligible for inclusion.
Total costs were estimated by applying the HCUP cost/charge ratio (CCR) files. Each file contains hospital-specific CCRs based on all-payer inpatient cost for nearly every hospital in the corresponding NIS or State Inpatient Database databases. Cost information was obtained from the hospital accounting reports collected by the Centers for Medicare and Medicaid Services. Because of incomplete hospital CCR data, cases without hospital-specific CCRs were excluded, and data were appropriately reweighted to analyze costs.
Data Analysis
We determined point estimates and 95% confidence intervals (CIs) for patient characteristics, charges, and costs for US bronchiolitis-related discharges. We also examined the frequency of procedures and associated diagnoses. Nationally representative estimates were determined using NIS-assigned discharge weights appropriate for the 20% subsample, which adjusted for the probability of selection. Data are presented with number of observations for the 20% subsample and the corresponding weighted totals representative of all discharges in the United States.
Age was studied as a dichotomous variable: <1 and
1 to < 2 years (hereafter referred to as age 1 year). Binomial and categorical variables were analyzed using weighted
2. Differences in means were evaluated using a t statistic appropriate for survey data. A multivariate linear regression of total costs was performed with predefined demographic, administrative, and treatment characteristics. All of the analyses were performed using appropriate survey commands to account for the stratified, 2-stage sampling frame. Although cost data were skewed, we did not transform the data because of the very large size of the data set. Based on the central limit theorem, in a very large data set, the distribution of an average tends to be normal, even when the distribution from which the average is computed is decidedly nonnormal.31 All of the P values are 2-sided, with P < .05 considered statistically significant. Data analysis was performed using Stata 9.0. (Stata Corp, College Station, TX).
| RESULTS |
|---|
|
|
|---|
|
Table 2 shows demographic characteristics for bronchiolitis patients. Approximately 85% of patients were children <1 year. Patients were more likely to be boys than girls (rate per 1000 US population age <2 years: 21 vs 16; P < .05). Almost 40% of all children with bronchiolitis were nonwhite, with a population rate higher than that of white non-Hispanics (17 vs 11; P < .05). There were no significant regional differences. More than 78% of hospitalizations were in urban hospitals. Mean length of stay (LOS) was 3.3 days overall and differed significantly between children <1 and age 1 year (3.4 vs 2.6 days, respectively; P < .01).
|
The 3 most common procedures recorded were respiratory medication given by a nebulizer (6.4%; 95% CI: 2.0% to 8.2%), lumbar puncture (4.0%; 95% CI: 3.2% to 4.9%), and oxygen enrichment (2.4%; 95% CI: 0.2% to 4.6%). The top 3 procedures were the same whether or not a codiagnosis of pneumonia or asthma was present (data not shown).
Table 3 shows multivariate regression results for predictors of total costs. Model 1, which does not include LOS, demonstrates that the mean cost per hospitalization was higher for younger children, boys, and patients in urban teaching hospitals. Patients who received lumbar puncture or mechanical ventilation had a higher mean cost than those not receiving those procedures ($2117 and $21469 higher, respectively). Having a codiagnosis of pneumonia raised the mean cost by $1781 (P < .001). In model 2, controlling for LOS, being older was more expensive, but neither having procedures nor being female predicted cost when LOS was included in the model. Having a codiagnosis of pneumonia was only a borderline significant predictor of higher cost (P = .07).
|
| DISCUSSION |
|---|
|
|
|---|
Leader et al2 used the 19972000 National Hospital Discharge Survey database linked with the HCUP NIS database to estimate hospital charges among children <1 year because of a primary diagnosis of RSV (including RSV pneumonia without actual bronchiolitis). The total reported annual charges were $650 million. If these charges are converted to actual costs (using a 2002 ratio provided by the HCUP office, written communication, 2005), the total cost would be approximately $300 million. Stang et al22 also looked at charges and used the NIS database from 1993. Their criteria for inclusion were children <1 year with an ICD-9 code of bronchiolitis (466.1) and hospitalization from November through April. Their estimated total hospital charges from the year 1993 (adjusted to 1998 dollars) were between $365 million and $585 million. The actual costs would be considerably smaller. Although we used the same ICD-9 code criteria as Stang et al22, they focused on RSV-related bronchiolitis infection only. Comparing estimates from these previous studies with the current study demonstrates that limiting cases to RSV-related diagnoses underestimates the true, total cost of bronchiolitis hospitalizations. Even when we applied the same criteria used previously, and adjusted for inflation, the bronchiolitis costs in our analysis are significantly higher than previous estimates (data not shown).
Mean hospitalization costs were considerably different based on whether a codiagnosis of pneumonia was present. Mean cost per hospitalization for those with a codiagnosis of pneumonia was significantly higher, $6191 vs $3376, for those without a codiagnosis of pneumonia, although this difference seems to be attributable to longer LOS. However, health care providers may unnecessarily order chest radiographs3,32 and may overdiagnose pneumonia.33 Future studies should identify best practices for the identification of children with the codiagnoses of bronchiolitis and pneumonia.
To put the national cost of bronchiolitis in perspective, we examined previous studies on hospitalization costs for other respiratory diseases, such as asthma and cystic fibrosis. One study reported annual costs of asthma hospitalizations in 1990 to be $1.6 billion34 and another in 1994 to be $2.5 billion.35 Although these costs are higher than those estimated for bronchiolitis hospitalizations in this study, the asthma costs include all ages, not just those <2 years. Lieu et al36 estimated the total costs for cystic fibrosis to be approximately $314 million in 1996. They report that 47% of cystic fibrosis costs were from hospitalizations, making $147 million attributable to inpatient costs. This is much lower than the bronchiolitis costs found in this study.
Younger age, male gender, urban hospital/teaching status, receiving a lumbar puncture or mechanical ventilation, and having a codiagnosis of pneumonia all predicted higher hospitalization cost based on the longer LOS. Younger age and being male are both associated with more severe bronchiolitis.5,37,38 However, when we controlled for LOS, older age became a significant predictor of higher cost. A possible explanation is that older children who require hospitalization for bronchiolitis may have chronic conditions resulting in higher costs, compared with frequently hospitalized, but otherwise healthy <1-year-olds. However, we looked at all of the codiagnoses for children <1 year and those 1 year of age, and they were similar, concluding that there may be another reason for the higher costs, a finding that merits additional study.
The study has a few potential limitations. First, the diagnosis of bronchiolitis is clinical. Accordingly, the data are subject to different diagnostic standards across the many US hospitals. We believe that including an inpatient diagnosis of bronchiolitis in any of the 15 fields reduces chances that we are missing true bronchiolitis cases. Second, the HCUP database does not collect race and ethnicity separately, making it difficult to interpret these demographic characteristics for bronchiolitis. Lastly, the HCUP database is useful for overall counts and cost calculations, but medication and imaging data are not available. Therefore, we were unable to analyze usage of antibiotics and chest radiographs and to determine their contribution to the overall cost. Because these represent likely targets for cost reduction, an itemized accounting of inpatient costs merits future study.
| CONCLUSIONS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Address correspondence to Andrea J. Pelletier, MS, MPH, EMNet Coordinating Center, Department of Emergency Medicine, Massachusetts General Hospital, 326 Cambridge St, 4th Floor, Boston, MA 02114. E-mail: apelletier1{at}partners.org
The authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. M. Mansbach and C. A. Camargo Jr Bronchiolitis: Lingering Questions About Its Definition and the Potential Role of Vitamin D Pediatrics, July 1, 2008; 122(1): 177 - 179. [Full Text] [PDF] |
||||
![]() |
J. M. Mansbach, S. Clark, N. C. Christopher, F. LoVecchio, S. Kunz, U. Acholonu, and C. A. Camargo Jr Prospective Multicenter Study of Bronchiolitis: Predicting Safe Discharges From the Emergency Department Pediatrics, April 1, 2008; 121(4): 680 - 688. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Unger and S. Cunningham Effect of Oxygen Supplementation on Length of Stay for Infants Hospitalized With Acute Viral Bronchiolitis Pediatrics, March 1, 2008; 121(3): 470 - 475. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-A. S. Harris, W. C. Huskins, J. M. Langley, J. D. Siegel, and for the Pediatric Special Interest Group of the So Health Care Epidemiology Perspective on the October 2006 Recommendations of the Subcommittee on Diagnosis and Management of Bronchiolitis Pediatrics, October 1, 2007; 120(4): 890 - 892. [Full Text] [PDF] |
||||
![]() |
M. Weinberger Corticosteroids Revisited for Acute Bronchiolitis in Hospitalized Children AAP Grand Rounds, August 1, 2007; 18(2): 13 - 13. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||