INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Despite the progress that has been made in the treatment of asthma, the prevalence and burden of this disease has continued to increase: currently, it is estimated that asthma affects over 14-million Americans (1). In the United States in 1993, asthma was responsible for an estimated 198,000 hospitalizations and 342 deaths among persons under 25 yr of age; this hospitalization rate represents an increase of 28% from 1980 (2). Furthermore, the overall death rate from asthma increased by 40% between 1982 and 1991, from 13.4 per million population to 18.8 per million (3).

The economic impact of asthma in the United States has been estimated to be 5.8 billion (4). Direct costs represent the majority of this total, accounting for 88% of it (4). Hospitalization for asthma represents the greatest single element of direct costs, accounting for over 50% of such costs (4, 5). Emergency department (ED) visits also represent a significant cost category, with previous reports estimating ED visits to incur 7 to 18% of direct costs (4, 5). Although these two categories are considered to be major contributors to the cost of asthma, little is known about the specific resources consumed and associated costs of an ED visit or hospital stay for asthma. The work by Weiss and colleagues (5) used charges rather than costs, which may overestimate the true cost of treatment, and Smith and associates (4) based their estimates on the 1987 National Medical Expenditure Survey (NMES), which may not reflect changes in treatment patterns during the 1990s. Furthermore, previous research has not been able to provide detailed reports of the resources or treatments required by asthma patients while in the hospital. Therefore, the purpose of this study was to provide a more accurate estimate of the resources consumed and costs for treating an asthma exacerbation in the ED and hospital in the United States during 1996 and 1997.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Data Source

Patient-level clinic and cost data were quantitated with Premier's Perspective Comparative Database (PCD). This database is the largest cost-based, test-level comparative database providing severity adjusted detailed, resource utilization information in the United States. For the present study, 27 hospitals across the United States were included. The hospitals included represented a broad spectrum of sites and bed sizes, comprising 30% rural hospitals with bed sizes ranging from 85 to 388, 30% suburban hospitals with bed sizes ranging from 150 to 805, and 40% urban hospitals with bed sizes ranging from 224 to 744 beds.

Sample Selection

The study population was identified from the 27 hospitals in the PCD database between October 1, 1996 and September 30, 1997. Occurrences of asthma were identified with the International Classification of Diseases, Ninth RevisionClinical Modification Codes (ICD-9 CM code 493). An occurrence of asthma was considered to be a visit to any emergency department (ED) for which the primary diagnosis code was asthma. To be eligible for the study, subjects had to be 18 yr of age or older at the time of the ED visit.

Data Collection

Following an ED visit, a patient could either be treated and released or admitted to the hospital. The costs and resource utilization data for these two possibilities were collected. Resource use information included medications and procedures performed in the ED; and nursing time, bed type (intensive care unit [ICU], coronary care unit [CCU], Telemetry, regular care), medications, and procedures performed by department during an inpatient hospital stay. Costs associated with each resource were recorded. In addition, the costs and resource utilization associated with rehospitalization for asthma was assessed. The number of patients requiring a subsequent hospital stay for asthma within 30 d of their original discharge date was determined, and the cost and resource utilization for each visit was calculated. Costs captured were the costs to the hospital for treatment of an asthma patient. All dollar amounts are recorded in 1997 U.S. dollars.

Along with resource and cost data, patient and hospital demographic data were collected for each occurrence. Patient demographics included age, gender, race, and insurance status. Because of concerns about misclassification of asthma in patients older than 45 yr of age, the costs of an ED visit and hospital stay for this group were analyzed to determine whether they differed significantly from those observed for patients 45 yr of age or younger.

For hospitalized patients, a severity-of-illness level was assigned, using the 3M All Patient Refined-Diagnosis-Related Groups (APR- DRG) severity measurement system (3M Health Information Systems, Wallingford, CT). The APR-DRG software is a patient classification system that uses hospital patient discharge data and computer-based logic to assign patients to severity-of-illness and risk-of-mortality classes so they can be accurately compared in terms of length of stay, resource consumption, and outcome. This classification system is an enhancement of the basic DRG structure, which includes four severity-of-illness and risk-of-mortality subclasses within each DRG. The four subclasses indicate minor, moderate, major, or extreme severity of illness or risk of mortality. Patients in each of the four clinically meaningful severity-of-illness and risk-of-mortality subclasses have similar resource utilization and outcome data (6).

Analyses

Data were summarized as the mean ± SEM. Multivariate analysis of variance (ANOVA) was performed to determine whether observed differences in mean costs were statistically significant; all tests were performed at a significance level of p = 0.05. Multivariate analysis allowed determination of the level of significance of the variable in question with control for potential confounders or interaction between variables. For variables with more than two levels of significance, Scheffe's test was used for testing differences between means. All analyses were done with SAS software, version 6.12 (SAS Institute, Cary, NC) (7).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sample Characteristics

A total of 3,223 patients who met the inclusion criteria were identified at 27 hospitals during the study period. The demographic characteristics of these subjects are given in Table 1. Thirty percent of the sample was male, 55.9% was white, and the average age was 40.9 ± 0.3 yr (mean ± SEM). Female subjects were slightly older than male subjects. The average age of females was 41.6 ± 0.4 yr and that of males was 39.3 ± 0.5 yr.

The most common secondary diagnosis observed in the patient sample was essential hypertension, with 328 patients (10.2%) having this diagnosis. Hospitalized patients were much more likely to have a secondary diagnosis of hypertension; 24% of hospitalized patients versus only 3.2% of ED patients had essential hypertension. Other common secondary diagnoses included electrolyte/mineral disorders (6.1%), bronchitis, (6.0%), upper respiratory infections (2.9%), sinusitis (2.3%), and pneumonia (1.6%).

Of the 3,223 patients included in the study, 1,074 (33.3%) required an inpatient hospital stay for asthma. Hospitalization for asthma occurred more often for females, with 35.5% of females hospitalized, whereas 28.3% of males were hospitalized.

The rate of hospitalization increased with age; the average age of a hospitalized patient was 44.8 ± 0.5 yr, whereas the average age for a patient visiting the ED was 38.9 ± 0.4 yr. For patients over 45 yr of age, the rate of hospitalization was compared with that of patients 45 yr of age or younger. Forty-two percent of patients over 45 yr of age were hospitalized for asthma, as compared with a hospitalization rate of 28.8% for patients 18 to 45 yr of age.

ED Visits

The resources and costs associated with an ED visit for asthma are given in Table 2. The average cost of an ED visit for asthma was $234.48 ± 4.7 (mean ± SEM). Costs associated with ED-specific supplies, equipment use, and physician fees accounted for 53.1% of the total. Other resources included respiratory therapy, which accounted for 11.0% of the total costs, and medication costs, which accounted for only 5.5% of the total.

The cost of an ED visit increased significantly (p < 0.0001) with age. Patients over 45 yr of age had an average cost of $294.08 ± 10.5, whereas patients 45 yr of age or younger had an average ED cost of $210.04 ± 5.0.

Hospitalization

Costs for patients requiring an inpatient stay for asthma were substantially higher than those for an ED visit. Table 2 provides the resource utilization and costs for an average hospital stay. The average cost of a hospital stay was $3,102.53 ± 151.9, with nursing care accounting for 44% of the cost, respiratory therapy accounting for 14%, and medications accounting for 10% (Table 2). The average cost of a hospital stay was significantly greater (p = 0.0015) for patients over 45 yr of age, averaging $3,601.95 ± 294.5, as compared with an average cost of $2,731.21 ± 147.4 for patients under 45 yr of age.

The average length of stay (LOS) was 3.8 d ± 0.1, with a range of 1 to 88 d; however, 93% of patients required 7 d of hospitalization or less. Removal of outliers (i.e., patients with extremely long stays) did not significantly affect the mean or SEM of any reported costs. Therefore, these patients were not removed from the analysis. Routine nursing care accounted for the majority of overall nursing costs, averaging 2.9 ± 0.1 d (mean ± SEM) and $976.9 ± 28.6 per patient, whereas Telemetry/Stepdown care accounted for 0.6 d (0.07) and $224.54 ($24.91) per patient. LOS also increased significantly with age; patients over 45 yr of age had an average LOS of 4.5 d as compared with 3.3 d for patients  45 yr of age.

The distribution of APR-DRG severity levels for the hospitalized patients, as well as associated costs and length of hospital stay for these patients, are provided in Table 3. Most of these patients (88%) were either in APR-DRG Subclass 1 or 2 (minor or moderate severity of illness or risk of mortality). The average cost of a hospital stay increased with increasing severity, with the mean cost for each level being significantly different from that of the others (p  0.05). Patients in APR- DRG Subclass 1 had a cost of approximately $2,195.00 ± 54.8, whereas patients in APR-DRG Subclass 4 (extreme severity) had a cost of approximately $15,131.82 ± 5,137.6. LOS also increased significantly (p  0.05) with increasing severity, beginning at 2.9 ± 0.1 d for patients in APR-DRG Subclass 1 and increasing to 12.3 ± 3.4 d for those in APR-DRG Subclass 4. The observed differences between severity subclasses remained statistically significant after adjustment for other variables such as age, gender, race, comorbidities, and payer type.

Of the 1,074 patients who were hospitalized, 968 (90.1%) were discharged to home or self care following their hospital stay, with another 81 patients (7.5%) being discharged to a home health organization, skilled nursing home, intermediate care facility, or another hospital. Two patients (0.2%) died in the hospital, and 23 (2.2%) of the patients either left against medical advice or were discharged to an unknown destination.

Rehospitalization

Twenty-three patients (2.1%) were rehospitalized for asthma within 30 d of discharge from their original inpatient stay. The number of subsequent inpatient stays ranged from one to three. Because of the small sample size, the values for rehospitalization are an average of all subsequent stays. The initial inpatient stay for these patients was shorter than the mean LOS for all hospitalized patients, at 2.4 ± 1.5 d for the rehospitalized group versus 3.80 ± 0.1 d for the overall group. The LOS increased for subsequent hospitalizations to 3.2 ± 2.3 d, but was still below the overall mean. The mean costs for both the initial and subsequent inpatient stays for the rehospitalized group were lower than the mean costs of these variables for the overall group. The mean cost of the initial hospitalization was $1,995.83 ± 1,045.13, and the mean cost for subsequent hospitalizations was $2,364.92 ± 1,953.84.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The costs reported in this study are the costs to the hospital for treating asthma in the ED or inpatient setting. This is one of the few asthma studies to measure costs in such a manner. Smith and associates (4), using 1987 NMES data, found the average cost of an ED visit for asthma to be $290 and that for a hospitalization $6,292, both in 1994 U.S. dollars. The costs reported in NMES data correspond to the total payments, including consumer out-of-pocket and third-party payer expenditures, for treatment of asthma. Expenditure data represent the amount paid for treatment, and include extraneous factors such as allocated overhead or profit to the hospital; this explains the higher estimates in the study by Smith and associates.

Smith and associates (4) estimated that approximately 1.2 million people visited the ED for asthma in 1987. Based on this estimate, the cost of treating asthma in the ED during 1997 was $281.4 million. Approximately 443,000 hospital stays related to asthma occurred in 1994. If a similar number of hospital stays occurred in 1997, the total cost of hospitalization for asthma in the United States was $1.4 billion (8). Asthma prevalence is thought to be increasing in the United States, these values are, therefore, conservative and likely underestimate the true costs.

Rates of hospitalization and cost of treatment increased with age, with 42.3% of patients over 45 yr of age being hospitalized. Each hospital stay for this age group had an average cost of $3,601.95 ± 294.52. Possible reasons for this increase include differences in disease severity or the presence of comorbid conditions. The average age of patients in APR-DRG Subclass 4 was 55 yr, as compared with an average age of 40 yr for those in APR-DRG Subclass 1, which supports the possibility of a difference in disease severity between older and younger patients. Although older patients are more likely to suffer from airway diseases other than asthma, secondary diagnoses such as emphysema or chronic bronchitis were not prevalent in this study.

Twenty-three patients who were hospitalized required a subsequent hospital admission within 30 d of discharge from their initial stay. The average LOS for the initial stay for these patients was significantly less than for the overall group. Although it is not possible to identify a causal relationship between shorter LOS and rehospitalization, these patients' shorter initial LOS may suggest that if they had remained in the hospital for a longer period initially, the second (or third) visit(s) would not have been necessary. The average number of hospitalizations for these 23 patients was 2.35 ± 0.67. The average cost for all hospitalization for these patients was $5,188.47, which is substantially higher than the average cost for the patients requiring only one hospital stay.

For this study, identification of subjects was based on the presence of an ICD-9 code for asthma. It is possible that patients seen at an ED for asthma received a different primary diagnosis and were therefore not included in our study; the reverse situation could also be true (i.e., a patient without asthma incorrectly received an asthma diagnosis). However, diagnoses in the PCD are put through an internal validation and editing system, which assesses the "correctness" of a diagnosis on the basis of resource utilization and procedure codes. If a diagnosis does not appear to match the treatment the patient received, that diagnosis is reassessed to determine whether another diagnosis code is more appropriate; as a result, the risk of incorrect coding of diagnoses is minimized. A previous study found underdiagnosis of asthma to be much more of a concern than incorrect assignment of an asthma diagnosis to someone who does not have the condition (9). This would suggest that our cost estimate for patients treated for asthma is a conservative one, since it is more likely that some asthmatic individuals were missed, as opposed to the inclusion of patients with other conditions.

Previous studies have suggested that ED utilization and hospitalization rates for asthma vary across racial groups: Hispanics subjects were twice as likely as non-Hispanic white subjects to report use of the emergency room as a source of primary medical care (10). Hispanic subjects also tended to have longer and more expensive hospitalizations (10). In this study, approximately 60% of Hispanic patients visiting the ED for asthma were hospitalized, as compared with 30% of white and 35% of African-American patients (Table 1). Hospitalization rates for African-American patients were not significantly different from those for white patients.

Coultas and coworkers (11) concluded that the respiratory health of Hispanic individuals varies across subgroups: specifically, rates of morbidity and mortality from asthma are highest among Puerto Rican as compared with Cuban-American or Mexican-American individuals (11). Unfortunately, it was not possible to identify subgroups among the Hispanic subjects in our study, and it is therefore unknown whether hospitalization rates varied across different Hispanic groups.

Subjects from 27 hospitals across the United States were identified, with most located in the southeastern United States, and relatively few in the northeast. Given the geographic limitations of the study data, the estimates in our study may differ from the true national averages. However, comparing the results of this study with those for the Healthcare Cost and Utilization Project's Nationwide Inpatient Sample (NIS), LOS values for hospitalized asthmatic patients were similar in the two studies (8).

Descriptive studies do not lend themselves to causal inferences. This should be kept in mind when evaluating the results of this study. Although certain patient groups in this study were found to have higher rates of hospitalization and/or treatment costs, further research is required to predict which patients are more likely to require hospitalization or more intensive treatment for asthma.

Previous research has suggested that asthma management before an exacerbation is often inadequate (12). It was beyond the scope of this study to determine whether patients in the study could have avoided an ED visit if they had received appropriate maintenance care. It is likely that steps toward improved education of asthmatic individuals about maintenance therapy could decrease the likelihood of their experiencing an exacerbation that would require emergency care. However, this study did not measure appropriateness or intensity of care. An important next step is to better define the resources consumed (i.e., types of medications used) for treating asthma and to determine the relationship between treatment choices and outcomes, so that optimal treatments can be identified.

Although our results tend to agree with those found in previous studies, the present study is the first to provide test-level resource utilization data for patients visiting the hospital for asthma treatment. The results presented here suggest that unavoidably used resources such as nursing care, respiratory therapy, or ED supplies account for the majority of hospital costs for asthma. Therefore, there may not be much room for cost savings once patients are hospitalized. However, previous research suggests that 20% of asthma patients account for 80% of direct costs (4), and in our study of adult asthma patients, hospitalized patients, who accounted for 33% of the sample, accounted for 86.9% of treatment costs. Thus, a continuum of care aimed at avoiding hospitalization for an exacerbation of asthma, especially in the elderly, could result in more substantial cost savings than could efforts aimed at reducing inpatient utilization.