- Email: [email protected]
Costs of asthma in the United States: 2002-2007
Costs of asthma in the United States: 2002-2007 Sarah Beth L. Barnett, MA, and Tursynbek A. Nurmagambetov, PhD Background: The economic burden of asthma is an important measure of the effect of asthma on society. Although asthma is a costly illness, the total cost of asthma to society has not been estimated in more than a decade. Objective: The purpose of this study is to provide the public with current estimates of the incremental direct medical costs and productivity losses due to morbidity and mortality from asthma at both the individual and national levels for the years 2002-2007. Methods: Data came from the Medical Expenditure Panel Survey. Two-part models were used to estimate the incremental direct costs of asthma. The incremental number of days lost from work and school was estimated by negative binomial regressions and valued following the human capital approach. Published data were used to value lives lost with an underlying cause of asthma. Results: Over the years 2002-2007, the incremental direct cost of asthma was $3,259 (2009 dollars) per person per year. The value of additional days lost attributable to asthma per year was approximately $301 for each worker and $93 for each student. For the most recent year available, 2007, the total incremental cost of asthma to society was $56 billion, with productivity losses due to morbidity accounting for $3.8 billion and productivity losses due to mortality accounting for $2.1 billion. Conclusion: The current study finds that the estimated costs of asthma are substantial, which stresses the necessity for research and policy to work toward reducing the economic burden of asthma. (J Allergy Clin Immunol 2011;127:145-52.) Key Words: Asthma, expenditures, Two-part model, direct cost, productivity losses, mortality losses
Asthma is a chronic, sometimes debilitating disorder of the airways characterized by recurrent symptoms of coughing (particularly at night), wheezing, chest tightness, and difficulty
From the Air Pollution and Respiratory Health Branch, Centers for Disease Control and Prevention. Supported by the Air Pollution and Respiratory Health Branch, Centers for Disease Control and Prevention and supported in part by an appointment to the Research Participation Program at the Centers for Disease Control and Prevention, administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the US Department of Energy and the Centers for Disease Control and Prevention. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention. Disclosure of potential conflict of interest: S. B. L. Barnett has received research support from the OakRidge Institute for Science and Education. T. A. Nurmagambetov has declared no conflict of interest. Received for publication November 10, 2009; revised October 16, 2010; accepted for publication October 20, 2010. Reprint requests: Sarah Beth L. Barnett, MA, Air Pollution and Respiratory Health Branch, Centers for Disease Control and Prevention, 4770 Buford Hwy NE, Atlanta, GA 30341. E-mail: [email protected]. 0091-6749/$36.00 Ó 2010 American Academy of Allergy, Asthma & Immunology doi:10.1016/j.jaci.2010.10.020
Atlanta, Ga
Abbreviations used GLM: Generalized linear model ICD-9-CM: International Classification of Diseases, Ninth Revision, Clinical Modification MEPS: Medical Expenditure Panel Survey
breathing. Although asthma affects persons of all ages, there is a higher prevalence of asthma among children younger than 18 years (9.3%) than among adults (7.3%).1 Asthma affects patient’s lives and those of their families, including their quality of life, productivity at work and school, and health care use, and can even result in death. The economic burden of the disease is a crucial measure of the societal effect of the disease for policy makers in setting priorities for public health programs. There have been numerous studies on the costs of asthma internationally2-12 and in the United States on the costs of adult13-16 and pediatric17-23 asthma. However, there have been few studies and reports24-28 that have estimated direct costs and productivity losses due to asthma for persons of all ages from a societal perspective. A frequently cited study, conducted by Weiss et al25 in 1992, estimated the cost of asthma to society, including direct costs and productivity losses due to morbidity and mortality, as $6.2 billion in 1990 (1990 dollars). This study used an analytic method sometimes referred to as the bottom-up method.29 Although the bottom-up method for estimating direct costs is straightforward, this method likely underestimates direct costs when only primary diagnostic events are summed. On the other hand, if secondary or further diagnostic codes are included in addition to primary diagnosis, the estimation might result in double counting of the costs of diseases.30 Regression-based methods have emerged as a standard for estimating the incremental costs attributable to a disease or risk factor.17,18,21,22,31-36 In the most recent study on the direct costs of asthma, Kamble and Bharmal34 used generalized linear regression models to estimate the incremental direct costs of asthma for adults and children separately using data from the 2004 Medical Expenditure Panel Survey (MEPS). The study finds that the 2004 incremental direct costs of asthma per person were $2,077.50 for adults and $1,004.60 for children, amounting to an estimated $37.17 billion (2007 dollars) in total direct annual medical expenditures associated with asthma. The purpose of this study is to provide a current estimate of the incremental direct medical costs and productivity losses due to morbidity and mortality from asthma at both the individual and national levels for the years 2002-2007. This study contributes to the literature on asthma because it is, to our knowledge, the first study in more than a decade to estimate total costs (incremental direct costs and productivity losses) due to asthma for persons of all ages. Furthermore, we use nationally representative personlevel data, which are the most comprehensive single source of medical expenditures. We use regression-based techniques that take into account the distribution of medical costs and productivity days lost to estimate the incremental direct costs of asthma and 145
146 BARNETT AND NURMAGAMBETOV
productivity days lost because of asthma for the years 2002-2007. We estimate the value of mortality by using vital statistics reports and published estimates of the value of productivity losses.
METHODS Data Data for calendar years 2002-2007 came from the MEPS, a large-scale nationally representative set of surveys of families and individuals and their medical providers and employers. The surveys collect detailed data on health care use, expenditures, sources of payment, and health insurance coverage. The MEPS has been conducted annually since 1996; it is cosponsored by the Agency for Healthcare Research and Quality and the National Center for Health Statistics. The survey uses a complex design and provides population weights to create nationally representative estimates of the US noninstitutionalized population.37 The publicly available Household Component of the survey, used in the current study, uses an overlapping panel design in which participants are interviewed in person for 5 rounds over 2 full calendar years. Households and individuals provide information on their demographic and socioeconomic characteristics, employment, days disrupted by injury or illness, health care and medication use, medical conditions, and health status. The Medical Provider Component of the survey samples participants’ medical providers and pharmacies and is used primarily as an imputation source to supplement household-reported expenditure information.37 The full sample during the years 2002-2007 ranged from 30,964 to 39,163, and the response rate ranged from 56.9% to 65.9%.37-42
Sample and identification of persons with asthma For each year, data from the Household Component of the survey, including the medical conditions file and event files (office-based medical provider visits, hospital outpatient visits and special clinic visits, emergency department visits, hospital inpatient stays, and prescribed medicines), were merged with the full-year consolidated file by using the unique identification variables, thus creating a merged file of person-level data for each of the years 20022007. The data files from the 6 years were also pooled to produce estimates with greater reliability, with a total sample size of 206,851 persons. The complex survey design of MEPS was accounted for by use of person-level weights and survey commands within Stata 11 software43 for all analyses and descriptive statistics, as recommended by the Agency for Healthcare Research and Quality.44 All expenditures and wages were adjusted to 2009 dollars by use of the Medical Care Consumer Price Index and the Consumer Price Index.45 Individuals were identified as persons with asthma if an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code of 493 was associated with an office-based medical provider visit, a hospital outpatient visit, an emergency department visit, a hospital inpatient stay, or a prescription medication. Self-reported medical events and prescription medications that were said to be related to asthma were recorded as verbatim text and then converted by professional coders to an ICD-9-CM code of 493 within the data.46 This case definition is similar to one version of the modified Council of State and Territorial Epidemiologists definition for probable asthma that performs well in identifying subjects with asthma in claims data.47 The definition for asthma used by this study is utilization based, and hence it does not include persons who reported having asthma as a condition but did not have any asthma-related events or prescription medications during the calendar year. Persons with asthma consisted of 4.2% of the pooled sample, representing 11.6 to 13.2 million persons during the years 2002-2007. Although not directly comparable, this percentage is close to the average annual prevalence percentage of self-reported asthma episodes, which ranged from 3.9% to 4.3% during the years 2002-2007.48
Dependent variables The exclusive major categories of expenditures, which were dependent variables for the analyses of direct cost, are office-based medical provider
J ALLERGY CLIN IMMUNOL JANUARY 2011
expenditures, hospital outpatient expenditures, emergency department expenditures, hospital inpatient expenditures, and prescribed medicine expenditures. The sum of these 5 major components of expenditures for the year, called total expenditures, was a dependent variable created for additional analyses of direct costs. The MEPS definition of expenditures is the sum of direct payments for care during the year, including out-of-pocket payments and payments by private insurance, Medicaid, Medicare, and other sources. Expenditures were chosen as the measurements of cost instead of charges because charges are often discounted and include uncollected liability, bad debt, and charitable care.37 The number of school days lost and the number of work days lost per calendar year* were the dependent variables in estimating productivity losses attributable to asthma. During each survey round, respondents reported the number of days during which they missed a half day or more of work or school because of illness, injury, or mental or emotional problems.37 Because partial and full days lost were indistinguishable in the data, all days lost were treated as full days in this study, which is consistent with previous studies using MEPS.22,24 The incremental school days lost because of asthma were estimated for students aged 3 to 19 years, and the incremental work days lost because of asthma were estimated for persons who had worked during at least 1 survey round in a year.
Independent variables The independent variable of interest is asthma, a binary variable, with 1 indicating that the person had an asthma-related medical visit or prescription medication during the calendar year and 0 indicating otherwise. Independent binary variables included in the analyses were for age groups, marital status, minority races, educational levels, sex, income less than 200% of the poverty line, uninsured status (not having private or public insurance), and calendar year (only for the pooled sample). Estimates of the incremental effect of asthma might be biased if asthma is correlated with other conditions, and therefore comorbidities were adjusted for by including the D’Hoore adaptation49 of the Charlson comorbidity index50 as a continuous independent variable. The D’Hoore version allows for the index’s use with ICD-9-CM codes, and it includes 17 conditions ; it has had good predictive power in an evaluation of published comorbidity indices for administrative databases.51
Modeling the incremental direct costs of asthma To select an appropriate econometric model, we followed the criteria outlined by Manning and Mullahy52 and Buntin and Zaslavsky.53 On the basis of their criteria, generalized linear models (GLMs) with log-link functions, which are differentially weighted nonlinear least-squares estimators, were selected; they are efficient with the correct choice of a variance function. To choose a specification among GLMs, we used the modified Park test52 for each of the expenditure models on both the full and positive portion of the distribution of the dependent variables. The results of the modified Park tests indicated that a gamma distribution with a log link was the best-fitting GLM model for estimating incremental expenditures. Although the choice of the GLM addressed the positive skewedness of the dependent variables, many of the dependent variables had a high
*Some of the disability days reported by participants in panels 6 and 7 of the survey pertain to both 2002 and 2003. Therefore we multiplied the number of days lost by the percentage of days in each person’s reference period pertaining to the year 2002. From 2003 onward, the disability days variables reflect only the data pertinent to the calendar year. The conditions include myocardial infarct, congestive heart failure, peripheral vascular disease, dementia, cerebrovascular disease, chronic pulmonary disease, connective tissue disease, ulcer disease, mild liver disease, hemiplegia, moderate-to-severe renal disease, diabetes, any tumor, leukemia, lymphoma, moderate or severe liver disease, and metastatic solid tumor. Although usually included in the index, ICD-9-CM codes related to asthma were not included in the computation of the index for this study because the condition is accounted for with a separate binary variable.
BARNETT AND NURMAGAMBETOV 147
J ALLERGY CLIN IMMUNOL VOLUME 127, NUMBER 1
concentration of observations with zero medical expenditure. To improve the precision of the estimators, we used the Two-part model.54,55 The first part of the Two-part model is a logistic that predicts the probability of positive medical expenditures, and the second part of the model is a GLM with a gamma distribution and a log link to estimate medical expenditures; such expenditures are estimated only on observations with positive expenditures. For each subject in the sample with asthma, we predicted 2 results from the Two-part model, one based on the subject having asthma and one simulating the removal of asthma with all other characteristics of the subject remaining the same. The first predicted result is the predicted probability of positive expenditures multiplied by the predicted value of expenditures for persons with asthma. The second predicted result is the predicted probability of positive expenditures multiplied by the predicted value of expenditures simulating the removal of asthma. We report the average of the difference in those predicted results, the incremental cost of asthma, for each category of expenditure. The sampling weights provided by MEPS are used with the predicted incremental direct costs of asthma to produce nationally representative estimates of the total direct cost of asthma in the United States for the years 2002-2007.
Modeling the incremental productivity costs of asthma We used a count data model to analyze the number of days lost from work for working adults and the number of days lost from school for students in one calendar year during 2002-2007 and over the entire time period through use of the pooled sample. The negative binomial model was chosen over the Poisson model because the Poisson model assumes that events are distributed independently over time, an assumption likely inappropriate for healthrelated absences56; moreover, the violation of that assumption might lead to overdispersion. The coefficient estimates from the productivity loss models were used to predict 2 results for each subject with asthma, the number of days lost with asthma and the number of days lost simulating the removal of asthma, to estimate the number of days lost attributable to asthma. The valuation of days lost follows the human capital approach.57,58 Work days were valued at the average daily wage of workers with asthma in the sample. Regarding school days lost, we assume that a parent or caregiver would forgo their usual activities to take care of a child who misses a day of school. The parent with the lower wage or the parent outside the labor force is assumed to stay home with the child. We use the parent’s wage or a wage for household work59 to value the loss to society as a result of a child missing a day of school. For the models estimating incremental direct costs and productivity costs due to asthma, we calculated percentile CIs using bootstrapping with 1000 iterations.60 Our CIs might be too wide because bootstrapping uses random resampling of the data; however, our point estimates of incremental costs and days lost use the survey design variables and should not be biased.61
Mortality cost of asthma The number of deaths caused by asthma was taken from the National Vital Statistics System’s final data on underlying causes of death by illnesses and age group; these data are based on all death certificates from the 50 states and the District of Columbia.62-67 In an attempt to express the loss of life in monetary terms, we relied on published estimates of economic productivity over the lifetime adjusted for survival and discounting.68
RESULTS Of the 206,851 observations in the pooled sample, 8,719 persons had asthma, and 198,132 persons did not have asthma (Table I). The persons in the asthma subsample tended to be younger and unmarried. They had lower education attainment, were more likely to be in poverty, and were more likely to have comorbidities.
TABLE I. Pooled sample summary statistics by asthma status Asthma
Unweighted no. Age (y) Age 0-4 (reference) Age 5-14 Age 15-34 (reference, work days lost model) Age 35-64 _65 Age > Marital status as not married (reference) Married White race (reference) Black race American Indian/Alaska Native race Asian race Native Hawaiian/Pacific Islander race Multiple races West region (reference) Northeast region Midwest region South region Children without a degree (reference) No degree earned High school diploma or GED Bachelor’s degree Graduate degree Other degree Male sex (reference) Female sex Income >200% of the poverty line (reference) Income <200% of the poverty line Insured (reference) Uninsured Charlson index for comorbidities Year 2002 (reference) Year 2003 Year 2004 Year 2005 Year 2006 Year 2007
No asthma
8,719 198,132 Mean Mean 34.558 (0.434) 36.515 (0.170) 0.081 0.066 0.221 0.135 0.205 0.276
P value
.000 .004 .000 .000
0.356 0.129 0.673
0.393 0.122 0.583
.000 .269 .000
0.327 0.781 0.156 0.009
0.417 0.806 0.123 0.008
.000 .000 .000 .582
0.024 0.004
0.043 0.003
.000 .918
0.027 0.214 0.215 0.222 0.343 0.341
0.017 0.230 0.182 0.220 0.359 0.234
.000 .030 .000 .894 .090 .000
0.134 0.314 0.104 0.047 0.059 0.430 0.570 0.636
0.148 0.374 0.125 0.062 0.057 0.492 0.508 0.694
.013 .000 .003 .001 .640 .000 .000 .000
0.364 0.306 0.947 0.873 0.053 0.127 0.397 (0.016) 0.252 (0.005) 0.160 0.163 0.158 0.165 0.157 0.166 0.166 0.168 0.181 0.169 0.178 0.170
.000 .000 .000 .000 .540 .148 .042 .728 .016 .146
Notes: SDs are reported in parentheses for continuous variables (all others are binary). Missing values for education attainment and marital status, accounting for less than 1% of the data, are recorded as zeros. Reference groups for categorical variables are specified in parentheses.
Incremental direct costs of asthma The results of the Two-part GLMs of the incremental costs attributable to asthma for each major expenditure category are shown by year in Table II. We chose to discuss the results from the pooled sample. The total incremental cost of asthma for a subject was estimated at $3,259 per year during the years 2002-2007. The predicted incremental cost of hospital outpatient visits was $151; for emergency department visits, it was $110, and for inpatient visits, it was $446. The incremental cost of office-based visits for persons with asthma was estimated at $581 per year. Prescription medication expenditures are estimated to cost an additional $1,680 a year for a person with asthma.
148 BARNETT AND NURMAGAMBETOV
J ALLERGY CLIN IMMUNOL JANUARY 2011
TABLE II. Estimates from 2-step GLM models of the per-person incremental costs of asthma by expenditure category (2009$) Total expenditure
2002 2003 2004 2005 2006 2007 Pooled sample
4,193* 3,413* 3,623* 2,745* 2,582* 3,856* 3,259*
(0 to 5,276) (0 to 4,142) (0 to 4,880) (2,012 to 3,748) (2,012 to 3,258) (2,821 to 5,153) (2,912 to 3,676)
Prescription medication
Office-based visits
1,583* 1,867* 1,779* 1,567* 1,490* 1,861* 1,680*
646* 636* 585* 483* 515* 632* 581*
(1,314 (1,563 (1,517 (1,335 (1,295 (1,571 (1,557
to to to to to to to
1,916) 2,207) 2,113) 1,929) 1,803) 2,230) 1,827)
(485 (480 (394 (303 (381 (399 (505
to to to to to to to
814) 810) 803) 689) 696) 931) 661)
Emergency department visits
143 113 74à 155 66 124 110
(90 (67 (35 (97 (31 (64 (90
to to to to to to to
197) 166) 121) 228) 104) 195) 133)
Outpatient visits
201 220 63à 153 167 143 151
(73 to 351) (66 to 369) (235 to 178) (24 to 313) (48 to 307) (39 to 262) (98 to 211)
Inpatient visits
985 329 68à 218 349 1,137 446à
(488 to 1,573) (21 to 758) (2383 to 545) (2411 to 398) (213 to 752) (406 to 1,961) (236 to 655)
Notes: The incremental cost of asthma is estimated from the 2-step model (step 1, logit; step 2, GLM with gamma distribution and log link), with both steps including age groups, married status, minority race, region, level of education, female sex, poverty, uninsured status, and Charlson comorbidity index as covariates. Asthma was positive and significant in all logit models. Ninety-five percent bootstrap CIs are presented in parentheses. *Asthma was positive and significant at 1% in GLMs Asthma was negative and significant at 1% in GLMs. àAsthma was negative and significant at 5% in GLMs.
TABLE III. Incremental days lost 2002-2007 from negative binomial model and value of days lost Days lost
2002 2003 2004 2005 2006 2007 Pooled sample
Work School Work School Work School Work School Work School Work School Work School
Incremental days lost
2.20 0.91 2.47 2.58 1.45 0.60 3.15 0.73 2.18 0.60 4.53 1.73 2.62 0.92
Weighted N with asthma
Mean day wage
Total value (billions of 2009$)
5,756,902 3,593,892 5,105,860 3,957,528 5,102,763 3,775,023 5,599,393 4,078,877 5,885,926 4,608,083 5,710,309 4,200,346 5,500,000 4,000,000
$106 $100 $117 $105 $124 $96 $112 $96 $115 $107 $117 $101 $115 $101
$1.342 $0.328 $1.472 $1.069 $0.921 $0.216 $1.981 $0.285 $1.472 $0.296 $3.024 $0.735 $1.657 $0.371
(0.92-3.67) (-0.22-2.77) (0.75-4.58) (-0.02-7.72) (-0.04-3.50) (-0.11-1.63) (1.10-5.72) (-0.07-1.86) (0.75-4.05) (-0.33-2.12) (2.36-7.43) (-0.69-7.98) (1.88-3.48) (0.28-1.81)
Average value of work days and school days lost per year over the pooled years 2002-2007 5 $2.028; total value for all years 5 $13.140. Notes: Predicted days lost for asthma and non-asthma are estimated from negative binomial regressions. Covariates for work days lost include age groups, married status, minority races, region, level of education, female, poverty, uninsured, and Charlson comorbidity index. Covariates for school days lost are the same as above less married and level of education. Ninety-five percent bootstrap confidence intervals are presented in parentheses.
Productivity losses attributable to asthma The estimates of the incremental days lost attributable to asthma, as estimated by negative binomial regressions, are presented in Table III. According to the results from the pooled sample from 2002-2007, workers with asthma experienced an additional 2.62 days lost from work per year, and students with asthma lost an additional 0.92 days from school per year. Asthma accounted for a 66% increase in work days lost and a 98% increase in school days lost. The value of additional days lost attributable to asthma per year was approximately $301 for each worker and $93 for each student. Thus, on the national scale, society loses approximately 14.41 million work days and 3.68 million school days per year because of asthma, with a combined estimated value of $2.03 billion per year. Over the study time period, the number of reported deaths caused by asthma ranged from 3445 in 2007 to 4260 in 2002 (Table IV). The present value of lives lost because of asthma during the 6-year period was an estimated $14.25 billion, an average of approximately $2.37 billion per year.
Total costs attributable to asthma The total costs of asthma to society during the years 2002-2007 are presented in Table V. The predicted incremental direct and indirect costs for a person with asthma were multiplied by the nationally weighted number of persons with asthma per year. The total cost of asthma to society in 2007 was estimated at $56 billion. Productivity losses from work days and school days lost because of morbidity and productivity losses from mortality represent 8% to 12% of annual total costs from 2002-2007. DISCUSSION Limitations Higher income whites and Asians had the lowest response rates to MEPS.69 Although MEPS oversamples minorities, including Asians, we cannot be assured that our estimates are accurate for nonrespondents. The effect of possible underrepresentation of higher-income whites and Asians on the incremental cost of asthma is unclear because there is no conclusive evidence on
BARNETT AND NURMAGAMBETOV 149
J ALLERGY CLIN IMMUNOL VOLUME 127, NUMBER 1
TABLE IV. Mortality costs of asthma Number of deaths with asthma as underlying cause
Under 1 year 1-4 years 5-14 years 15-24 years 25-34 years 35-44 years 45-54 years 55-64 years 65-74 years 75-84 years 85 years and over Total deaths by year Present value loss per year (billions)
2002
2003
2004
2005
2006
2007
4 43 123 169 235 472 608 536 583 812 675 4,260 $2.664
7 37 110 158 227 411 632 562 532 752 671 4,099 $2.544
7 29 105 159 197 376 562 520 504 673 684 3,816 $2.333
4 37 97 131 207 369 595 520 475 709 740 3,884 $2.322
6 26 99 135 194 373 566 492 443 626 653 3,613 $2.237
4 41 107 133 201 320 538 461 412 569 659 3,445 $2.146
Total deaths by age
Present value of one life (2009$)
32 213 641 885 1,261 2,321 3,501 3,091 2,949 4,141 4,082
$1,221,771 $1,221,771 $1,420,888 $1,660,503 $1,627,306 $1,324,446 $918,243 $505,923 $242,316 $125,369 $105,048
Total present value of deaths related to asthma from 2002-2007 5 $14.247 billion; Average present value of deaths related to asthma per year over the 6-years 5 $2.374 billion. Notes: numbers of deaths with asthma as the underlying cause are reported in National Vital Statistics Reports. Estimates of the present value of one life are adapted from Grosse et al (2009).68
TABLE V. Cost of asthma to society by year (billions of 2009$)
Nationally weighted N with asthma (millions) Total direct incremental cost Total cost of work and school days lost Productivity loss due to death Total cost of asthma to society
2002
2003
2004
2005
2006
2007
11.6 48.64 1.67 2.66 52.97
11.5 39.25 2.54 2.54 44.33
11.4 41.30 1.14 2.33 44.77
12.1 33.21 2.27 2.32 37.80
13.2 34.08 1.77 2.24 38.09
13.0 50.13 3.76 2.15 56.03
Notes: Estimates of the total incremental cost of asthma per person from the 2-step regressions are multiplied by the weighted number of individuals in the sample with asthma. Estimates of the incremental cost of work and school days lost per person are multiplied by the weighted number of workers and students, respectively. Estimates of the value of productivity loss due to death are the same as reported in Table IV.
TABLE VI. Alternative estimation of the direct cost of asthma in 2006 (2009$) Weighted MEPS estimates Category
Prescription medication Office-based visits Emergency department visits Outpatient visits Inpatient visits
NCHS sources estimates
Mean price
N events
Cost in billions
N events
Cost in billions
$93 $123 $638 $450 $4,767
110,000,000 17,000,000 1,000,000 645,420 325,647 Total direct cost
$10.260 $2.095 $0.638 $0.291 $1.552 $14.836
10,590,000 1,592,880 1,088,468 444,000 Total direct cost (less prescription medication)
$1.305 $1.016 $0.490 $2.116 $4.927
Notes: The mean prices reported are the mean expenditures for medical events related to asthma in 2006 from MEPS, which is multiplied by both the number of estimated events from MEPS and the number of estimated events from NCHS sources. There is not an estimate of the prescription medication usage from NCHS sources.
the independent effect of income on asthma-related health expenditures. Persons who have not used health care goods or services related to asthma in the last year and who reported having asthma are not coded as having asthma for the purposes of this study. This population accounted for 1.1% of the pooled sample, and their mean direct medical expenditures did not significantly differ from those of persons who did not have any asthma-related events or reported having asthma. We do not assert that this group of persons does not have asthma; we merely do not believe the data suggest that they meaningfully contribute to the cost of asthma during the calendar year reported.
The present study does not attempt to estimate nonmedical direct costs, presenteeism, and the intangible costs of asthma, all of which can have a substantial economic burden and would enhance the total estimates of the cost of asthma to society. The estimates for the cost of asthma are sensitive to the technique and model used,à and thus we attempted to use the most appropriate analyses on the basis of econometric tests and to control for covariates.
àFor example, when the Duan estimator54 is used in the 2-part regression technique, the direct cost of asthma in the current study for the year 2006 is $55 billion.
150 BARNETT AND NURMAGAMBETOV
Putting the results into perspective: A brief comparison of methods and estimates The current study estimates the total direct cost of asthma at $38.1 billion in 2006, with an underlying asthmatic population totaling 13.2 million in that year. In an attempt to provide another perspective on the cost of asthma, we present an alternative estimate of the direct costs of asthma for the year 2006, the most recent year possible with these data, by using the bottom-up method (Table VI). The mean expenditures for services related to asthma from the 2006 MEPS are used as the price of services. The numbers of nationally weighted events by category were estimated by using both asthma-related events from the 2006 MEPS and events with asthma as the primary diagnosis from the 2006 Nationally Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey.70-72 The alternative direct cost of asthma for 2006, using solely MEPS data, is $14.84 billion, which is less than half (39%) of the regression-based estimate of the incremental direct cost of asthma in 2006. The difference in the regression-based and bottom-up estimates can be explained in part by studies reporting that asthmatic patients incurred higher annual medical costs than persons without asthma but that asthma-specific medical care accounted for only a fraction of those higher annual medical costs.13,18,20 Persons with asthma might incur additional costs for other illnesses or events, even when asthma is not recorded as a diagnosis in the medical records. The pathology of asthma and its relatedness to other conditions is not completely understood. We postulate that the regression-based estimates of the incremental direct cost of asthma account for costs that are associated with asthma but are not necessarily coded as asthma diagnoses.
Conclusion We have estimated that between 2002 and 2007, asthma resulted in an additional 2.62 days lost from work each year and 0.92 days from school each year per person. The estimates of incremental school days lost per student per year are reasonably consistent with previous studies estimating excess school days lost because of asthma by using MEPS, as well as other data.22,73,74 The value of productivity loss is estimated at $2.03 billion because of morbidity and $2.37 billion because of mortality per year. Although productivity losses represent a significant loss to society, they are relatively small in comparison with the incremental direct costs of asthma per year, likely because of the measurement of days lost in the MEPS survey and the conservative dollar values assigned to days lost. The incremental direct cost of asthma alone over the years 2002-2007 is estimated at $3,259 per person per year. The predicted total medical expenditure simulating the removal of asthma was $4,200; consequently, having asthma implied a 78% higher total medical cost per person per year (Fig 1). Prescription medications account for more than 50% ($1,680) of the incremental direct cost of asthma. This finding is consistent with previous studies that found prescription medications to comprise the largest percentage of total medical expenditures attributable to asthma in the adult population.14,34 Hospital costs no longer are the largest share of direct costs of asthma, as they were in the 1990s.25,75
J ALLERGY CLIN IMMUNOL JANUARY 2011
FIG 1. Individual predicted annual medical expenditures by category, 20022007, estimated from 2-step GLMs on the pooled sample. Red, Inpatient; orange, outpatient; yellow, emergency department; green, office visits; blue, prescription.
Many studies have found that asthma disproportionately affects disadvantaged subpopulations, including persons in poverty, minorities, and children living in urban areas.76-78 Recent studies have also found that US–born persons of minority race and ethnic groups were more likely to have current asthma than their foreignborn counterparts.79-81 These studies suggest that environmental factors might play a role both in the development and burden of asthma. Determining the contribution of various environmental factors in the development and burden of asthma is a necessary next step toward reducing the effect of those environmental factors on asthma and the costs associated with them. The Task Force on Community Preventative Services recently recommended home-based, multitrigger, multicomponent interventions with an environmental focus for children and adolescents with asthma.82 Environmental remediation items and trigger-reducing behaviors might be a strategy to reduce health care use and subsequently costs. This study contributes to the current cost-of-illness and asthma literature because we provide both national and individual cost estimates using several years of recent nationally representative data and regression-based methods. Our estimates suggest that the costs of asthma are substantial. Understanding the costs of asthma assists policy makers, program administrators, and individuals in evaluating options to reduce the burden of the disease and improve quality of life. It is our hope that this article will promote further discussion and research aimed at improving the lives of persons with asthma. We thank Shubhayu Saha, PhD, for his input during the development of the study; Dana Flanders, MD, MPH, DSc, for his comments; and Elizabeth Herman, MD, MPH, Maureen Wilce, MS, and Paul Garbe, DVM, MPH, for their comments and support of this project. We also thank the reviewers for their thoughtful suggestions.
Key Messages d
The estimated incremental direct cost of asthma was $3,259 (2009 dollars) per person per year during the years 2002-2007.
d
In 2007, the estimated total cost of asthma (incremental direct cost and productivity costs) was $56 billion (2009 dollars).
J ALLERGY CLIN IMMUNOL VOLUME 127, NUMBER 1
REFERENCES 1. Table 4-1 current asthma prevalence percents by age, United States: National Health Interview Survey, 2006. Available at: http://www.cdc.gov/asthma/nhis/06/ table4-1.htm. Accessed July 6, 2009. 2. Antonicelli L, Bucca C, Neri M, De Benedetto F, Sabbatani P, Bonifazi F, et al. Asthma severity and medical resource utilisation. Eur Respir J 2004;5:723-9. 3. Chew FT, Goh DY, Lee BW. The economic cost of asthma in Singapore. Aust N Z J Med 1999;29:228-33. 4. Herjavecz I, Nagy GB, Gyurkovits K, Magyar P, Dobos K, Nagy L, et al. Cost, morbidity, and control of asthma in Hungary: the Hunair Study. J Asthma 2003; 40:673-81. 5. Krahn MD, Berka C, Langlois P, Detsky AS. Direct and indirect costs of asthma in Canada, 1990. CMAJ 1996;154:821-31. 6. Lodha R, Puranik M, Kattal N, Kabra SK. Social and economic impact of childhood asthma. Indian Pediatr 2003;40:874-9. 7. Rutten-van Molken MP, Postma MJ, Joore MA, Van Genugten ML, Leidl R, Jager JC. Current and future medical costs of asthma and chronic obstructive pulmonary disease in The Netherlands. Respir Med 1999;93:779-87. 8. Schramm B, Ehlken B, Smala A, Quednau K, Berger K, Nowak D. Cost of illness of atopic asthma and seasonal allergic rhinitis in Germany: 1-yr retrospective study. Eur Respir J 2003;21:116-22. 9. Schwenkglenks M, Lowy A, Anderhub H, Szucs TD. Costs of asthma in a cohort of Swiss adults: associations with exacerbation status and severity. Value Health 2003;6:75-83. 10. Stevens CA, Turner D, Kuehni CE, Couriel JM, Silverman M. The economic impact of preschool asthma and wheeze. Eur Respir J 2003;21:1000-6. 11. Szucs TD, Anderhub H, Rutishauser M. The economic burden of asthma: direct and indirect costs in Switzerland. Eur Respir J 1999;13:281-6. 12. Van Ganse E, Laforest L, Pietri G, Boissel JP, Gormand F, Ben-Joseph R, et al. Persistent asthma: disease control, resource utilisation and direct costs. Eur Respir J 2002;20:260-7. 13. Birnbaum HG, Berger WE, Greenberg PE, Holland M, Auerbach R, Atkins KM, et al. Direct and indirect costs of asthma to an employer. J Allergy Clin Immunol 2002;109:264-70. 14. Cisternas MG, Blanc PD, Yen IH, Katz PP, Earnest G, Eisner MD, et al. A comprehensive study of the direct and indirect costs of adult asthma. J Allergy Clin Immunol 2003;111:1212-8. 15. Stanford R, McLaughlin T, Okamoto LJ. The cost of asthma in the emergency department and hospital. Am J Respir Crit Care Med 1999;160:211-5. 16. Yelin E, Henke J, Katz PP, Eisner MD, Blanc PD. Work dynamics of adults with asthma. Am J Ind Med 1999;35:472-80. 17. Gendo K, Sullivan SD, Lozano P, Finkelstein JA, Fuhlbrigge A, Weiss KB. Resource costs for asthma-related care among pediatric patients in managed care. Ann Allergy Asthma Immunol 2003;91:251-7. 18. Grupp-Phelan J, Lozano P, Fishman P. Health care utilization and cost in children with asthma and selected comorbidities. J Asthma 2001;38:363-73. 19. Lenney W. The burden of pediatric asthma. Pediatr Pulmonol Suppl 1997;15:13-6. 20. Lozano P, Fishman P, VonKorff M, Hecht J. Health care utilization and cost among children with asthma who were enrolled in a health maintenance organization. Pediatrics 1997;99:757-64. 21. Lozano P, Sullivan SD, Smith DH, Weiss KB. The economic burden of asthma in US children: estimates from the National Medical Expenditure Survey. J Allergy Clin Immunol 1999;104:957-63. 22. Wang LY, Zhong Y, Wheeler L. Direct and indirect costs of asthma in school-age children. Prev Chronic Dis 2005;2:A11. 23. Weinmann S, Kamtsiuris P, Henke K-D, Wickman M, Jenner A, Wahn U. The costs of atopy and asthma in children: assessment of direct costs and their determinants in a birth cohort. Pediatr Allergy Immunol 2003;14:18-26. 24. Smith DH, Malone DC, Lawson KA, Okamoto LJ, Battista C, Saunders WB. A national estimate of the economic costs of asthma. Am J Respir Crit Care Med 1997;156:787-93. 25. Weiss KB, Gergen PJ, Hodgson TA. An economic evaluation of asthma in the United States. N Engl J Med 1992;326:862-6. 26. Weiss KB, Sullivan SD. Understanding the costs of asthma: the next step. CMAJ 1996;154:841-3. 27. Weiss KB, Sullivan SD. The health economics of asthma and rhinitis. I. Assessing the economic impact. J Allergy Clin Immunol 2001;107:3-8. 28. Table 20 Economic Cost of Asthma, United States, 2007. Available at: http://www. lungusa.org/assets/documents/ASTHMA-JAN-2009.pdf. Accessed January 18, 2010. 29. Tolpin HG, Bentkover JD. Economic cost of illness: decision-making applications and practical considerations. Adv Health Econ Health Serv Res 1983;4:165-98. 30. Akobundu E, Ju J, Blatt L, Mullins CD. Cost-of-illness studies: a review of current methods. Pharmacoeconomics 2006;24:869-90.
BARNETT AND NURMAGAMBETOV 151
31. Honeycutt AA, Segel JE, Hoerger TJ, Finkelstein EA. Comparing cost-of-illness estimates from alternative approaches: an application to diabetes. Health Serv Res 2009;44:303-20. 32. Finkelstein EA, Fiebelkorn IC, Wang GJ. National medical spending attributable to overweight and obesity: how much, and who’s paying? Health Aff (Millwood) 2003;22:W219-26. 33. Howard DH, Molinari NA, Thorpe KE. National estimates of medical costs incurred by nonelderly cancer patients. Cancer 2004;100:883-91. 34. Kamble S, Bharmal M. Incremental direct expenditure of treating asthma in the United States. J Asthma 2009;46:73-80. 35. Trogdon JG, Finkelstein EA, Hoerger TJ. Use of econometric models to estimate expenditure shares. Health Serv Res 2008;43:1442-52. 36. Trogdon JG, Finkelstein EA, Nwaise IA, Tangka FK, Orenstein D. The economic burden of chronic cardiovascular disease for major insurers. Health Promot Pract 2007;8:234-42. 37. MEPS HC-105: 2006 full year consolidated data file, 2008. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h105/h105doc.pdf. Accessed June 12, 2009. 38. MEPS HC-070 2002 full year consolidated data file, 2004. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h70/h70doc.pdf. Accessed June 12, 2009. 39. MEPS HC-079 2003 full year consolidated data file, 2005. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h79/h79doc.pdf. Accessed June 12, 2009. 40. MEPS HC-089 2004 full year consolidated data file, 2006. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h89/h89doc.pdf. Accessed June 12, 2009. 41. MEPS HC-097 2005 full year consolidated data file, 2007. Available at: http://www. meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h97/h97doc.pdf. Accessed June 12, 2009 42. MEPS HC-113 2007 full year consolidated data file, 2009. Available at: http://www. meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h113/h113doc.pdf. Accessed March 29, 2010 43. StataCorp. Stata Statistical Software: Release 11. College Station (TX): StataCorp; 2007. 44. Machlin S, Yu, W, Zodet M. Computing standard errors for MEPS estimates, 2005. Available at: http://www.meps.ahrq.gov/mepsweb/survey_comp/standard_errors. jsp. Accessed May 1, 2009. 45. Table 1A. consumer price index for all urban consumers (CPI-U): U.S. city average, by expenditure category and commodity and service group, 2002–2009. Available at: http://www.bls.gov/cpi/cpi_dr.htm#2007. Accessed June 9, 2009. 46. MEPS HC-102G 2006 office based medical provider visits, 2008. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h102g/h102gdoc.pdf. Accessed August 7, 2009. 47. Wakefield DB, Cloutier MM. Modifications to HEDIS and CSTE algorithms improve case recognition of pediatric asthma. Pediatr Pulmonol 2006;41:962-71. 48. Centers for Disease Control and Prevention. Early release of selected estimates based on data from the 2008 National Health Interview Survey, 2009. Available at: http:// www.cdc.gov/nchs/data/nhis/earlyrelease/earlyrelease200906.pdf. Accessed July 6, 2009. 49. D’Hoore W, Bouckaert A, Tilquin C. Practical considerations on the use of the Charlson comorbidity index with administrative data bases. J Clin Epidemiol 1996;49:1429-33. 50. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373-83. 51. Schneeweiss S, Maclure M. Use of comorbidity scores for control of confounding in studies using administrative databases. Int J Epidemiol 2000;29:891-8. 52. Manning WG, Mullahy J. Estimating log models: to transform or not to transform? J Health Econ 2001;20:461-94. 53. Buntin MB, Zaslavsky AM. Too much ado about two-part models and transformation?: comparing methods of modeling Medicare expenditures. J Health Econ 2004;23:525-42. 54. Duan N, Manning WG, Morris CN, Newhouse JP. A comparison of alternative models for the demand for medical care. J Bus Econ Stat 1983;1:115-26. 55. Duan N, Manning WG, Morris CN, Newhouse JP. Choosing between the sampleselection model and the multi-part model. J Bus Econ Stat 1984;2:283-9. 56. Vistnes JP. Gender differences in days lost from work due to illness. Ind Labor Relat Rev 1997;50:304-23. 57. Jonsson B. Measuring the economic burden of asthma. In: Weiss KB, Buist AS, Sullivan SD, editors. Asthma’s impact on society: the social and economic burdenNew York: Markel Dekker; 2000. p. 251-67. 58. Weisbrod B. Economics of public health. Philadelphia: University of Pennsylvania Press; 1961.
152 BARNETT AND NURMAGAMBETOV
59. Median weekly earning of full-time wage and salary workers by detailed occupation and sex—continued, 2002–2006. Available at: ftp://ftp.bls.gov/pub/special. requests/lf/. Accessed April 18, 2009. 60. Efron B, Tibshirani R. An Introduction to the Bootstrap. New York: Chapman & Hall; 1993. 61. Kolenikov S. Resampling variance estimation for complex survey data. Stata J 2010;10:165-99. 62. Heron M, Hoyert DL, Murphy SL, Xu J, Kochanek KD, Tejada-Vera B. Deaths: final data for 2006. Natl Vital Stat Rep 2009;57:1-134. 63. Hoyert DL, Heron MP, Murphy SL, Kung HC. Deaths: final data for 2003. Natl Vital Stat Rep 2006;54:1-120. 64. Kochanek KD, Murphy SL, Anderson RN, Scott C. Deaths: final data for 2002. Natl Vital Stat Rep 2004;53:1-115. 65. Kung HC, Hoyert DL, Xu J, Murphy SL. Deaths: final data for 2005. Natl Vital Stat Rep 2008;56:1-120. 66. Minino AM, Heron MP, Murphy SL, Kochanek KD. Deaths: final data for 2004. Natl Vital Stat Rep 2007;55:1-119. 67. Xu J, Kochanek MA, Murphy SL, Tejada-Vera B. Deaths: final data for 2007. Natl Vital Stat Rep 2009;58:1-135. 68. Grosse SD, Krueger KV, Mvundura M. Economic productivity by age and sex: 2007 estimates for the United States. Med Care 2009;47(suppl 1):S94-103. 69. MEPS annual methodology report Rockville (MD): Westat; 2009, Mar. Deliverable No.:142. Contract No.:290-02-0005. Sponsored by the Agency for Healthcare Research and Quality. 70. Cherry DK, Hing E, Woodwell DA, Rechtsteiner EA. National Ambulatory Medical Care Survey: 2006 summary. Natl Health Stat Rep 2008;3):1-39. 71. DeFrances CJ, Lucas CA, Buie VC, Golosinskiy A. 2006 National Hospital Discharge Survey. Natl Health Stat Rep 2008;(5):1-20. 72. Schappert SA, Rechtsteiner EA. Ambulatory medical care utilization estimates for 2006. Natl Health Stat Rep 2008;(8):1-29.
J ALLERGY CLIN IMMUNOL JANUARY 2011
73. Moonie SA, Sterling DA, Figgs L, Castro M. Asthma status and severity affects missed school days. J Sch Health 2006;76:18-24. 74. Silverstein MD, Mair JE, Katusic SK, Wollan PC, O’Connell ED, Yunginger JW. School attendance and school and school performance: a population-based study of children with asthma. J Pediatr 2001;139:278-83. 75. Weiss KB, Sullivan SD, Lyttle CS. Trends in the cost of illness for asthma in the United States, 1985-1994. J Allergy Clin Immunol 2000;106:493-9. 76. Aligne CA, Auinger P, Byrd RS, Weitzman M. Risk factors for pediatric asthma. Contributions of poverty, race, and urban residence. Am J Respir Crit Care Med 2000;162:873-7. 77. Christiansen SC, Martin SB, Schleicher NC, Koziol JA, Mathews KP, Zuraw BL. Current prevalence of asthma-related symptoms in San Diego’s predominantly Hispanic inner-city children. J Asthma 1996;33:17-26. 78. Persky VW, Slezak J, Contreras A, Becker L, Hernandez E, Ramakrishnan V, et al. Relationships of race and socioeconomic status with prevalence, severity, and symptoms of asthma in Chicago school children. Ann Allergy Asthma Immunol 1998;81:266-71. 79. Brugge D, Woodin M, Schuch TJ, Salas FL, Bennett A, Osgood ND. Communitylevel data suggest that asthma prevalence varies between U.S. and foreign-born black subpopulations. J Asthma 2008;45:785-9. 80. Holguin F, Mannino DM, Anto J, Mott J, Ford ES, Teague WG, et al. Country of birth as a risk factor for asthma among Mexican Americans. Am J Respir Crit Care Med 2005;171:103-8. 81. Brim SN, Rudd RA, Funk RH, Callahan DB. Asthma prevalence among US children in underrepresented minority populations: American Indian/ Alaska Native, Chinese, Filipino, and Asian Indian. Pediatrics 2008;122: e217-22. 82. Home-based multi-trigger multicomponent environmental interventions, 2009. Available at: http://www.thecommunityguide.org/asthma/multicomponent.html. Accessed August 7, 2009.
Asthma is a chronic, sometimes debilitating disorder of the airways characterized by recurrent symptoms of coughing (particularly at night), wheezing, chest tightness, and difficulty
From the Air Pollution and Respiratory Health Branch, Centers for Disease Control and Prevention. Supported by the Air Pollution and Respiratory Health Branch, Centers for Disease Control and Prevention and supported in part by an appointment to the Research Participation Program at the Centers for Disease Control and Prevention, administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the US Department of Energy and the Centers for Disease Control and Prevention. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention. Disclosure of potential conflict of interest: S. B. L. Barnett has received research support from the OakRidge Institute for Science and Education. T. A. Nurmagambetov has declared no conflict of interest. Received for publication November 10, 2009; revised October 16, 2010; accepted for publication October 20, 2010. Reprint requests: Sarah Beth L. Barnett, MA, Air Pollution and Respiratory Health Branch, Centers for Disease Control and Prevention, 4770 Buford Hwy NE, Atlanta, GA 30341. E-mail: [email protected]. 0091-6749/$36.00 Ó 2010 American Academy of Allergy, Asthma & Immunology doi:10.1016/j.jaci.2010.10.020
Atlanta, Ga
Abbreviations used GLM: Generalized linear model ICD-9-CM: International Classification of Diseases, Ninth Revision, Clinical Modification MEPS: Medical Expenditure Panel Survey
breathing. Although asthma affects persons of all ages, there is a higher prevalence of asthma among children younger than 18 years (9.3%) than among adults (7.3%).1 Asthma affects patient’s lives and those of their families, including their quality of life, productivity at work and school, and health care use, and can even result in death. The economic burden of the disease is a crucial measure of the societal effect of the disease for policy makers in setting priorities for public health programs. There have been numerous studies on the costs of asthma internationally2-12 and in the United States on the costs of adult13-16 and pediatric17-23 asthma. However, there have been few studies and reports24-28 that have estimated direct costs and productivity losses due to asthma for persons of all ages from a societal perspective. A frequently cited study, conducted by Weiss et al25 in 1992, estimated the cost of asthma to society, including direct costs and productivity losses due to morbidity and mortality, as $6.2 billion in 1990 (1990 dollars). This study used an analytic method sometimes referred to as the bottom-up method.29 Although the bottom-up method for estimating direct costs is straightforward, this method likely underestimates direct costs when only primary diagnostic events are summed. On the other hand, if secondary or further diagnostic codes are included in addition to primary diagnosis, the estimation might result in double counting of the costs of diseases.30 Regression-based methods have emerged as a standard for estimating the incremental costs attributable to a disease or risk factor.17,18,21,22,31-36 In the most recent study on the direct costs of asthma, Kamble and Bharmal34 used generalized linear regression models to estimate the incremental direct costs of asthma for adults and children separately using data from the 2004 Medical Expenditure Panel Survey (MEPS). The study finds that the 2004 incremental direct costs of asthma per person were $2,077.50 for adults and $1,004.60 for children, amounting to an estimated $37.17 billion (2007 dollars) in total direct annual medical expenditures associated with asthma. The purpose of this study is to provide a current estimate of the incremental direct medical costs and productivity losses due to morbidity and mortality from asthma at both the individual and national levels for the years 2002-2007. This study contributes to the literature on asthma because it is, to our knowledge, the first study in more than a decade to estimate total costs (incremental direct costs and productivity losses) due to asthma for persons of all ages. Furthermore, we use nationally representative personlevel data, which are the most comprehensive single source of medical expenditures. We use regression-based techniques that take into account the distribution of medical costs and productivity days lost to estimate the incremental direct costs of asthma and 145
146 BARNETT AND NURMAGAMBETOV
productivity days lost because of asthma for the years 2002-2007. We estimate the value of mortality by using vital statistics reports and published estimates of the value of productivity losses.
METHODS Data Data for calendar years 2002-2007 came from the MEPS, a large-scale nationally representative set of surveys of families and individuals and their medical providers and employers. The surveys collect detailed data on health care use, expenditures, sources of payment, and health insurance coverage. The MEPS has been conducted annually since 1996; it is cosponsored by the Agency for Healthcare Research and Quality and the National Center for Health Statistics. The survey uses a complex design and provides population weights to create nationally representative estimates of the US noninstitutionalized population.37 The publicly available Household Component of the survey, used in the current study, uses an overlapping panel design in which participants are interviewed in person for 5 rounds over 2 full calendar years. Households and individuals provide information on their demographic and socioeconomic characteristics, employment, days disrupted by injury or illness, health care and medication use, medical conditions, and health status. The Medical Provider Component of the survey samples participants’ medical providers and pharmacies and is used primarily as an imputation source to supplement household-reported expenditure information.37 The full sample during the years 2002-2007 ranged from 30,964 to 39,163, and the response rate ranged from 56.9% to 65.9%.37-42
Sample and identification of persons with asthma For each year, data from the Household Component of the survey, including the medical conditions file and event files (office-based medical provider visits, hospital outpatient visits and special clinic visits, emergency department visits, hospital inpatient stays, and prescribed medicines), were merged with the full-year consolidated file by using the unique identification variables, thus creating a merged file of person-level data for each of the years 20022007. The data files from the 6 years were also pooled to produce estimates with greater reliability, with a total sample size of 206,851 persons. The complex survey design of MEPS was accounted for by use of person-level weights and survey commands within Stata 11 software43 for all analyses and descriptive statistics, as recommended by the Agency for Healthcare Research and Quality.44 All expenditures and wages were adjusted to 2009 dollars by use of the Medical Care Consumer Price Index and the Consumer Price Index.45 Individuals were identified as persons with asthma if an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code of 493 was associated with an office-based medical provider visit, a hospital outpatient visit, an emergency department visit, a hospital inpatient stay, or a prescription medication. Self-reported medical events and prescription medications that were said to be related to asthma were recorded as verbatim text and then converted by professional coders to an ICD-9-CM code of 493 within the data.46 This case definition is similar to one version of the modified Council of State and Territorial Epidemiologists definition for probable asthma that performs well in identifying subjects with asthma in claims data.47 The definition for asthma used by this study is utilization based, and hence it does not include persons who reported having asthma as a condition but did not have any asthma-related events or prescription medications during the calendar year. Persons with asthma consisted of 4.2% of the pooled sample, representing 11.6 to 13.2 million persons during the years 2002-2007. Although not directly comparable, this percentage is close to the average annual prevalence percentage of self-reported asthma episodes, which ranged from 3.9% to 4.3% during the years 2002-2007.48
Dependent variables The exclusive major categories of expenditures, which were dependent variables for the analyses of direct cost, are office-based medical provider
J ALLERGY CLIN IMMUNOL JANUARY 2011
expenditures, hospital outpatient expenditures, emergency department expenditures, hospital inpatient expenditures, and prescribed medicine expenditures. The sum of these 5 major components of expenditures for the year, called total expenditures, was a dependent variable created for additional analyses of direct costs. The MEPS definition of expenditures is the sum of direct payments for care during the year, including out-of-pocket payments and payments by private insurance, Medicaid, Medicare, and other sources. Expenditures were chosen as the measurements of cost instead of charges because charges are often discounted and include uncollected liability, bad debt, and charitable care.37 The number of school days lost and the number of work days lost per calendar year* were the dependent variables in estimating productivity losses attributable to asthma. During each survey round, respondents reported the number of days during which they missed a half day or more of work or school because of illness, injury, or mental or emotional problems.37 Because partial and full days lost were indistinguishable in the data, all days lost were treated as full days in this study, which is consistent with previous studies using MEPS.22,24 The incremental school days lost because of asthma were estimated for students aged 3 to 19 years, and the incremental work days lost because of asthma were estimated for persons who had worked during at least 1 survey round in a year.
Independent variables The independent variable of interest is asthma, a binary variable, with 1 indicating that the person had an asthma-related medical visit or prescription medication during the calendar year and 0 indicating otherwise. Independent binary variables included in the analyses were for age groups, marital status, minority races, educational levels, sex, income less than 200% of the poverty line, uninsured status (not having private or public insurance), and calendar year (only for the pooled sample). Estimates of the incremental effect of asthma might be biased if asthma is correlated with other conditions, and therefore comorbidities were adjusted for by including the D’Hoore adaptation49 of the Charlson comorbidity index50 as a continuous independent variable. The D’Hoore version allows for the index’s use with ICD-9-CM codes, and it includes 17 conditions ; it has had good predictive power in an evaluation of published comorbidity indices for administrative databases.51
Modeling the incremental direct costs of asthma To select an appropriate econometric model, we followed the criteria outlined by Manning and Mullahy52 and Buntin and Zaslavsky.53 On the basis of their criteria, generalized linear models (GLMs) with log-link functions, which are differentially weighted nonlinear least-squares estimators, were selected; they are efficient with the correct choice of a variance function. To choose a specification among GLMs, we used the modified Park test52 for each of the expenditure models on both the full and positive portion of the distribution of the dependent variables. The results of the modified Park tests indicated that a gamma distribution with a log link was the best-fitting GLM model for estimating incremental expenditures. Although the choice of the GLM addressed the positive skewedness of the dependent variables, many of the dependent variables had a high
*Some of the disability days reported by participants in panels 6 and 7 of the survey pertain to both 2002 and 2003. Therefore we multiplied the number of days lost by the percentage of days in each person’s reference period pertaining to the year 2002. From 2003 onward, the disability days variables reflect only the data pertinent to the calendar year. The conditions include myocardial infarct, congestive heart failure, peripheral vascular disease, dementia, cerebrovascular disease, chronic pulmonary disease, connective tissue disease, ulcer disease, mild liver disease, hemiplegia, moderate-to-severe renal disease, diabetes, any tumor, leukemia, lymphoma, moderate or severe liver disease, and metastatic solid tumor. Although usually included in the index, ICD-9-CM codes related to asthma were not included in the computation of the index for this study because the condition is accounted for with a separate binary variable.
BARNETT AND NURMAGAMBETOV 147
J ALLERGY CLIN IMMUNOL VOLUME 127, NUMBER 1
concentration of observations with zero medical expenditure. To improve the precision of the estimators, we used the Two-part model.54,55 The first part of the Two-part model is a logistic that predicts the probability of positive medical expenditures, and the second part of the model is a GLM with a gamma distribution and a log link to estimate medical expenditures; such expenditures are estimated only on observations with positive expenditures. For each subject in the sample with asthma, we predicted 2 results from the Two-part model, one based on the subject having asthma and one simulating the removal of asthma with all other characteristics of the subject remaining the same. The first predicted result is the predicted probability of positive expenditures multiplied by the predicted value of expenditures for persons with asthma. The second predicted result is the predicted probability of positive expenditures multiplied by the predicted value of expenditures simulating the removal of asthma. We report the average of the difference in those predicted results, the incremental cost of asthma, for each category of expenditure. The sampling weights provided by MEPS are used with the predicted incremental direct costs of asthma to produce nationally representative estimates of the total direct cost of asthma in the United States for the years 2002-2007.
Modeling the incremental productivity costs of asthma We used a count data model to analyze the number of days lost from work for working adults and the number of days lost from school for students in one calendar year during 2002-2007 and over the entire time period through use of the pooled sample. The negative binomial model was chosen over the Poisson model because the Poisson model assumes that events are distributed independently over time, an assumption likely inappropriate for healthrelated absences56; moreover, the violation of that assumption might lead to overdispersion. The coefficient estimates from the productivity loss models were used to predict 2 results for each subject with asthma, the number of days lost with asthma and the number of days lost simulating the removal of asthma, to estimate the number of days lost attributable to asthma. The valuation of days lost follows the human capital approach.57,58 Work days were valued at the average daily wage of workers with asthma in the sample. Regarding school days lost, we assume that a parent or caregiver would forgo their usual activities to take care of a child who misses a day of school. The parent with the lower wage or the parent outside the labor force is assumed to stay home with the child. We use the parent’s wage or a wage for household work59 to value the loss to society as a result of a child missing a day of school. For the models estimating incremental direct costs and productivity costs due to asthma, we calculated percentile CIs using bootstrapping with 1000 iterations.60 Our CIs might be too wide because bootstrapping uses random resampling of the data; however, our point estimates of incremental costs and days lost use the survey design variables and should not be biased.61
Mortality cost of asthma The number of deaths caused by asthma was taken from the National Vital Statistics System’s final data on underlying causes of death by illnesses and age group; these data are based on all death certificates from the 50 states and the District of Columbia.62-67 In an attempt to express the loss of life in monetary terms, we relied on published estimates of economic productivity over the lifetime adjusted for survival and discounting.68
RESULTS Of the 206,851 observations in the pooled sample, 8,719 persons had asthma, and 198,132 persons did not have asthma (Table I). The persons in the asthma subsample tended to be younger and unmarried. They had lower education attainment, were more likely to be in poverty, and were more likely to have comorbidities.
TABLE I. Pooled sample summary statistics by asthma status Asthma
Unweighted no. Age (y) Age 0-4 (reference) Age 5-14 Age 15-34 (reference, work days lost model) Age 35-64 _65 Age > Marital status as not married (reference) Married White race (reference) Black race American Indian/Alaska Native race Asian race Native Hawaiian/Pacific Islander race Multiple races West region (reference) Northeast region Midwest region South region Children without a degree (reference) No degree earned High school diploma or GED Bachelor’s degree Graduate degree Other degree Male sex (reference) Female sex Income >200% of the poverty line (reference) Income <200% of the poverty line Insured (reference) Uninsured Charlson index for comorbidities Year 2002 (reference) Year 2003 Year 2004 Year 2005 Year 2006 Year 2007
No asthma
8,719 198,132 Mean Mean 34.558 (0.434) 36.515 (0.170) 0.081 0.066 0.221 0.135 0.205 0.276
P value
.000 .004 .000 .000
0.356 0.129 0.673
0.393 0.122 0.583
.000 .269 .000
0.327 0.781 0.156 0.009
0.417 0.806 0.123 0.008
.000 .000 .000 .582
0.024 0.004
0.043 0.003
.000 .918
0.027 0.214 0.215 0.222 0.343 0.341
0.017 0.230 0.182 0.220 0.359 0.234
.000 .030 .000 .894 .090 .000
0.134 0.314 0.104 0.047 0.059 0.430 0.570 0.636
0.148 0.374 0.125 0.062 0.057 0.492 0.508 0.694
.013 .000 .003 .001 .640 .000 .000 .000
0.364 0.306 0.947 0.873 0.053 0.127 0.397 (0.016) 0.252 (0.005) 0.160 0.163 0.158 0.165 0.157 0.166 0.166 0.168 0.181 0.169 0.178 0.170
.000 .000 .000 .000 .540 .148 .042 .728 .016 .146
Notes: SDs are reported in parentheses for continuous variables (all others are binary). Missing values for education attainment and marital status, accounting for less than 1% of the data, are recorded as zeros. Reference groups for categorical variables are specified in parentheses.
Incremental direct costs of asthma The results of the Two-part GLMs of the incremental costs attributable to asthma for each major expenditure category are shown by year in Table II. We chose to discuss the results from the pooled sample. The total incremental cost of asthma for a subject was estimated at $3,259 per year during the years 2002-2007. The predicted incremental cost of hospital outpatient visits was $151; for emergency department visits, it was $110, and for inpatient visits, it was $446. The incremental cost of office-based visits for persons with asthma was estimated at $581 per year. Prescription medication expenditures are estimated to cost an additional $1,680 a year for a person with asthma.
148 BARNETT AND NURMAGAMBETOV
J ALLERGY CLIN IMMUNOL JANUARY 2011
TABLE II. Estimates from 2-step GLM models of the per-person incremental costs of asthma by expenditure category (2009$) Total expenditure
2002 2003 2004 2005 2006 2007 Pooled sample
4,193* 3,413* 3,623* 2,745* 2,582* 3,856* 3,259*
(0 to 5,276) (0 to 4,142) (0 to 4,880) (2,012 to 3,748) (2,012 to 3,258) (2,821 to 5,153) (2,912 to 3,676)
Prescription medication
Office-based visits
1,583* 1,867* 1,779* 1,567* 1,490* 1,861* 1,680*
646* 636* 585* 483* 515* 632* 581*
(1,314 (1,563 (1,517 (1,335 (1,295 (1,571 (1,557
to to to to to to to
1,916) 2,207) 2,113) 1,929) 1,803) 2,230) 1,827)
(485 (480 (394 (303 (381 (399 (505
to to to to to to to
814) 810) 803) 689) 696) 931) 661)
Emergency department visits
143 113 74à 155 66 124 110
(90 (67 (35 (97 (31 (64 (90
to to to to to to to
197) 166) 121) 228) 104) 195) 133)
Outpatient visits
201 220 63à 153 167 143 151
(73 to 351) (66 to 369) (235 to 178) (24 to 313) (48 to 307) (39 to 262) (98 to 211)
Inpatient visits
985 329 68à 218 349 1,137 446à
(488 to 1,573) (21 to 758) (2383 to 545) (2411 to 398) (213 to 752) (406 to 1,961) (236 to 655)
Notes: The incremental cost of asthma is estimated from the 2-step model (step 1, logit; step 2, GLM with gamma distribution and log link), with both steps including age groups, married status, minority race, region, level of education, female sex, poverty, uninsured status, and Charlson comorbidity index as covariates. Asthma was positive and significant in all logit models. Ninety-five percent bootstrap CIs are presented in parentheses. *Asthma was positive and significant at 1% in GLMs Asthma was negative and significant at 1% in GLMs. àAsthma was negative and significant at 5% in GLMs.
TABLE III. Incremental days lost 2002-2007 from negative binomial model and value of days lost Days lost
2002 2003 2004 2005 2006 2007 Pooled sample
Work School Work School Work School Work School Work School Work School Work School
Incremental days lost
2.20 0.91 2.47 2.58 1.45 0.60 3.15 0.73 2.18 0.60 4.53 1.73 2.62 0.92
Weighted N with asthma
Mean day wage
Total value (billions of 2009$)
5,756,902 3,593,892 5,105,860 3,957,528 5,102,763 3,775,023 5,599,393 4,078,877 5,885,926 4,608,083 5,710,309 4,200,346 5,500,000 4,000,000
$106 $100 $117 $105 $124 $96 $112 $96 $115 $107 $117 $101 $115 $101
$1.342 $0.328 $1.472 $1.069 $0.921 $0.216 $1.981 $0.285 $1.472 $0.296 $3.024 $0.735 $1.657 $0.371
(0.92-3.67) (-0.22-2.77) (0.75-4.58) (-0.02-7.72) (-0.04-3.50) (-0.11-1.63) (1.10-5.72) (-0.07-1.86) (0.75-4.05) (-0.33-2.12) (2.36-7.43) (-0.69-7.98) (1.88-3.48) (0.28-1.81)
Average value of work days and school days lost per year over the pooled years 2002-2007 5 $2.028; total value for all years 5 $13.140. Notes: Predicted days lost for asthma and non-asthma are estimated from negative binomial regressions. Covariates for work days lost include age groups, married status, minority races, region, level of education, female, poverty, uninsured, and Charlson comorbidity index. Covariates for school days lost are the same as above less married and level of education. Ninety-five percent bootstrap confidence intervals are presented in parentheses.
Productivity losses attributable to asthma The estimates of the incremental days lost attributable to asthma, as estimated by negative binomial regressions, are presented in Table III. According to the results from the pooled sample from 2002-2007, workers with asthma experienced an additional 2.62 days lost from work per year, and students with asthma lost an additional 0.92 days from school per year. Asthma accounted for a 66% increase in work days lost and a 98% increase in school days lost. The value of additional days lost attributable to asthma per year was approximately $301 for each worker and $93 for each student. Thus, on the national scale, society loses approximately 14.41 million work days and 3.68 million school days per year because of asthma, with a combined estimated value of $2.03 billion per year. Over the study time period, the number of reported deaths caused by asthma ranged from 3445 in 2007 to 4260 in 2002 (Table IV). The present value of lives lost because of asthma during the 6-year period was an estimated $14.25 billion, an average of approximately $2.37 billion per year.
Total costs attributable to asthma The total costs of asthma to society during the years 2002-2007 are presented in Table V. The predicted incremental direct and indirect costs for a person with asthma were multiplied by the nationally weighted number of persons with asthma per year. The total cost of asthma to society in 2007 was estimated at $56 billion. Productivity losses from work days and school days lost because of morbidity and productivity losses from mortality represent 8% to 12% of annual total costs from 2002-2007. DISCUSSION Limitations Higher income whites and Asians had the lowest response rates to MEPS.69 Although MEPS oversamples minorities, including Asians, we cannot be assured that our estimates are accurate for nonrespondents. The effect of possible underrepresentation of higher-income whites and Asians on the incremental cost of asthma is unclear because there is no conclusive evidence on
BARNETT AND NURMAGAMBETOV 149
J ALLERGY CLIN IMMUNOL VOLUME 127, NUMBER 1
TABLE IV. Mortality costs of asthma Number of deaths with asthma as underlying cause
Under 1 year 1-4 years 5-14 years 15-24 years 25-34 years 35-44 years 45-54 years 55-64 years 65-74 years 75-84 years 85 years and over Total deaths by year Present value loss per year (billions)
2002
2003
2004
2005
2006
2007
4 43 123 169 235 472 608 536 583 812 675 4,260 $2.664
7 37 110 158 227 411 632 562 532 752 671 4,099 $2.544
7 29 105 159 197 376 562 520 504 673 684 3,816 $2.333
4 37 97 131 207 369 595 520 475 709 740 3,884 $2.322
6 26 99 135 194 373 566 492 443 626 653 3,613 $2.237
4 41 107 133 201 320 538 461 412 569 659 3,445 $2.146
Total deaths by age
Present value of one life (2009$)
32 213 641 885 1,261 2,321 3,501 3,091 2,949 4,141 4,082
$1,221,771 $1,221,771 $1,420,888 $1,660,503 $1,627,306 $1,324,446 $918,243 $505,923 $242,316 $125,369 $105,048
Total present value of deaths related to asthma from 2002-2007 5 $14.247 billion; Average present value of deaths related to asthma per year over the 6-years 5 $2.374 billion. Notes: numbers of deaths with asthma as the underlying cause are reported in National Vital Statistics Reports. Estimates of the present value of one life are adapted from Grosse et al (2009).68
TABLE V. Cost of asthma to society by year (billions of 2009$)
Nationally weighted N with asthma (millions) Total direct incremental cost Total cost of work and school days lost Productivity loss due to death Total cost of asthma to society
2002
2003
2004
2005
2006
2007
11.6 48.64 1.67 2.66 52.97
11.5 39.25 2.54 2.54 44.33
11.4 41.30 1.14 2.33 44.77
12.1 33.21 2.27 2.32 37.80
13.2 34.08 1.77 2.24 38.09
13.0 50.13 3.76 2.15 56.03
Notes: Estimates of the total incremental cost of asthma per person from the 2-step regressions are multiplied by the weighted number of individuals in the sample with asthma. Estimates of the incremental cost of work and school days lost per person are multiplied by the weighted number of workers and students, respectively. Estimates of the value of productivity loss due to death are the same as reported in Table IV.
TABLE VI. Alternative estimation of the direct cost of asthma in 2006 (2009$) Weighted MEPS estimates Category
Prescription medication Office-based visits Emergency department visits Outpatient visits Inpatient visits
NCHS sources estimates
Mean price
N events
Cost in billions
N events
Cost in billions
$93 $123 $638 $450 $4,767
110,000,000 17,000,000 1,000,000 645,420 325,647 Total direct cost
$10.260 $2.095 $0.638 $0.291 $1.552 $14.836
10,590,000 1,592,880 1,088,468 444,000 Total direct cost (less prescription medication)
$1.305 $1.016 $0.490 $2.116 $4.927
Notes: The mean prices reported are the mean expenditures for medical events related to asthma in 2006 from MEPS, which is multiplied by both the number of estimated events from MEPS and the number of estimated events from NCHS sources. There is not an estimate of the prescription medication usage from NCHS sources.
the independent effect of income on asthma-related health expenditures. Persons who have not used health care goods or services related to asthma in the last year and who reported having asthma are not coded as having asthma for the purposes of this study. This population accounted for 1.1% of the pooled sample, and their mean direct medical expenditures did not significantly differ from those of persons who did not have any asthma-related events or reported having asthma. We do not assert that this group of persons does not have asthma; we merely do not believe the data suggest that they meaningfully contribute to the cost of asthma during the calendar year reported.
The present study does not attempt to estimate nonmedical direct costs, presenteeism, and the intangible costs of asthma, all of which can have a substantial economic burden and would enhance the total estimates of the cost of asthma to society. The estimates for the cost of asthma are sensitive to the technique and model used,à and thus we attempted to use the most appropriate analyses on the basis of econometric tests and to control for covariates.
àFor example, when the Duan estimator54 is used in the 2-part regression technique, the direct cost of asthma in the current study for the year 2006 is $55 billion.
150 BARNETT AND NURMAGAMBETOV
Putting the results into perspective: A brief comparison of methods and estimates The current study estimates the total direct cost of asthma at $38.1 billion in 2006, with an underlying asthmatic population totaling 13.2 million in that year. In an attempt to provide another perspective on the cost of asthma, we present an alternative estimate of the direct costs of asthma for the year 2006, the most recent year possible with these data, by using the bottom-up method (Table VI). The mean expenditures for services related to asthma from the 2006 MEPS are used as the price of services. The numbers of nationally weighted events by category were estimated by using both asthma-related events from the 2006 MEPS and events with asthma as the primary diagnosis from the 2006 Nationally Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey.70-72 The alternative direct cost of asthma for 2006, using solely MEPS data, is $14.84 billion, which is less than half (39%) of the regression-based estimate of the incremental direct cost of asthma in 2006. The difference in the regression-based and bottom-up estimates can be explained in part by studies reporting that asthmatic patients incurred higher annual medical costs than persons without asthma but that asthma-specific medical care accounted for only a fraction of those higher annual medical costs.13,18,20 Persons with asthma might incur additional costs for other illnesses or events, even when asthma is not recorded as a diagnosis in the medical records. The pathology of asthma and its relatedness to other conditions is not completely understood. We postulate that the regression-based estimates of the incremental direct cost of asthma account for costs that are associated with asthma but are not necessarily coded as asthma diagnoses.
Conclusion We have estimated that between 2002 and 2007, asthma resulted in an additional 2.62 days lost from work each year and 0.92 days from school each year per person. The estimates of incremental school days lost per student per year are reasonably consistent with previous studies estimating excess school days lost because of asthma by using MEPS, as well as other data.22,73,74 The value of productivity loss is estimated at $2.03 billion because of morbidity and $2.37 billion because of mortality per year. Although productivity losses represent a significant loss to society, they are relatively small in comparison with the incremental direct costs of asthma per year, likely because of the measurement of days lost in the MEPS survey and the conservative dollar values assigned to days lost. The incremental direct cost of asthma alone over the years 2002-2007 is estimated at $3,259 per person per year. The predicted total medical expenditure simulating the removal of asthma was $4,200; consequently, having asthma implied a 78% higher total medical cost per person per year (Fig 1). Prescription medications account for more than 50% ($1,680) of the incremental direct cost of asthma. This finding is consistent with previous studies that found prescription medications to comprise the largest percentage of total medical expenditures attributable to asthma in the adult population.14,34 Hospital costs no longer are the largest share of direct costs of asthma, as they were in the 1990s.25,75
J ALLERGY CLIN IMMUNOL JANUARY 2011
FIG 1. Individual predicted annual medical expenditures by category, 20022007, estimated from 2-step GLMs on the pooled sample. Red, Inpatient; orange, outpatient; yellow, emergency department; green, office visits; blue, prescription.
Many studies have found that asthma disproportionately affects disadvantaged subpopulations, including persons in poverty, minorities, and children living in urban areas.76-78 Recent studies have also found that US–born persons of minority race and ethnic groups were more likely to have current asthma than their foreignborn counterparts.79-81 These studies suggest that environmental factors might play a role both in the development and burden of asthma. Determining the contribution of various environmental factors in the development and burden of asthma is a necessary next step toward reducing the effect of those environmental factors on asthma and the costs associated with them. The Task Force on Community Preventative Services recently recommended home-based, multitrigger, multicomponent interventions with an environmental focus for children and adolescents with asthma.82 Environmental remediation items and trigger-reducing behaviors might be a strategy to reduce health care use and subsequently costs. This study contributes to the current cost-of-illness and asthma literature because we provide both national and individual cost estimates using several years of recent nationally representative data and regression-based methods. Our estimates suggest that the costs of asthma are substantial. Understanding the costs of asthma assists policy makers, program administrators, and individuals in evaluating options to reduce the burden of the disease and improve quality of life. It is our hope that this article will promote further discussion and research aimed at improving the lives of persons with asthma. We thank Shubhayu Saha, PhD, for his input during the development of the study; Dana Flanders, MD, MPH, DSc, for his comments; and Elizabeth Herman, MD, MPH, Maureen Wilce, MS, and Paul Garbe, DVM, MPH, for their comments and support of this project. We also thank the reviewers for their thoughtful suggestions.
Key Messages d
The estimated incremental direct cost of asthma was $3,259 (2009 dollars) per person per year during the years 2002-2007.
d
In 2007, the estimated total cost of asthma (incremental direct cost and productivity costs) was $56 billion (2009 dollars).
J ALLERGY CLIN IMMUNOL VOLUME 127, NUMBER 1
REFERENCES 1. Table 4-1 current asthma prevalence percents by age, United States: National Health Interview Survey, 2006. Available at: http://www.cdc.gov/asthma/nhis/06/ table4-1.htm. Accessed July 6, 2009. 2. Antonicelli L, Bucca C, Neri M, De Benedetto F, Sabbatani P, Bonifazi F, et al. Asthma severity and medical resource utilisation. Eur Respir J 2004;5:723-9. 3. Chew FT, Goh DY, Lee BW. The economic cost of asthma in Singapore. Aust N Z J Med 1999;29:228-33. 4. Herjavecz I, Nagy GB, Gyurkovits K, Magyar P, Dobos K, Nagy L, et al. Cost, morbidity, and control of asthma in Hungary: the Hunair Study. J Asthma 2003; 40:673-81. 5. Krahn MD, Berka C, Langlois P, Detsky AS. Direct and indirect costs of asthma in Canada, 1990. CMAJ 1996;154:821-31. 6. Lodha R, Puranik M, Kattal N, Kabra SK. Social and economic impact of childhood asthma. Indian Pediatr 2003;40:874-9. 7. Rutten-van Molken MP, Postma MJ, Joore MA, Van Genugten ML, Leidl R, Jager JC. Current and future medical costs of asthma and chronic obstructive pulmonary disease in The Netherlands. Respir Med 1999;93:779-87. 8. Schramm B, Ehlken B, Smala A, Quednau K, Berger K, Nowak D. Cost of illness of atopic asthma and seasonal allergic rhinitis in Germany: 1-yr retrospective study. Eur Respir J 2003;21:116-22. 9. Schwenkglenks M, Lowy A, Anderhub H, Szucs TD. Costs of asthma in a cohort of Swiss adults: associations with exacerbation status and severity. Value Health 2003;6:75-83. 10. Stevens CA, Turner D, Kuehni CE, Couriel JM, Silverman M. The economic impact of preschool asthma and wheeze. Eur Respir J 2003;21:1000-6. 11. Szucs TD, Anderhub H, Rutishauser M. The economic burden of asthma: direct and indirect costs in Switzerland. Eur Respir J 1999;13:281-6. 12. Van Ganse E, Laforest L, Pietri G, Boissel JP, Gormand F, Ben-Joseph R, et al. Persistent asthma: disease control, resource utilisation and direct costs. Eur Respir J 2002;20:260-7. 13. Birnbaum HG, Berger WE, Greenberg PE, Holland M, Auerbach R, Atkins KM, et al. Direct and indirect costs of asthma to an employer. J Allergy Clin Immunol 2002;109:264-70. 14. Cisternas MG, Blanc PD, Yen IH, Katz PP, Earnest G, Eisner MD, et al. A comprehensive study of the direct and indirect costs of adult asthma. J Allergy Clin Immunol 2003;111:1212-8. 15. Stanford R, McLaughlin T, Okamoto LJ. The cost of asthma in the emergency department and hospital. Am J Respir Crit Care Med 1999;160:211-5. 16. Yelin E, Henke J, Katz PP, Eisner MD, Blanc PD. Work dynamics of adults with asthma. Am J Ind Med 1999;35:472-80. 17. Gendo K, Sullivan SD, Lozano P, Finkelstein JA, Fuhlbrigge A, Weiss KB. Resource costs for asthma-related care among pediatric patients in managed care. Ann Allergy Asthma Immunol 2003;91:251-7. 18. Grupp-Phelan J, Lozano P, Fishman P. Health care utilization and cost in children with asthma and selected comorbidities. J Asthma 2001;38:363-73. 19. Lenney W. The burden of pediatric asthma. Pediatr Pulmonol Suppl 1997;15:13-6. 20. Lozano P, Fishman P, VonKorff M, Hecht J. Health care utilization and cost among children with asthma who were enrolled in a health maintenance organization. Pediatrics 1997;99:757-64. 21. Lozano P, Sullivan SD, Smith DH, Weiss KB. The economic burden of asthma in US children: estimates from the National Medical Expenditure Survey. J Allergy Clin Immunol 1999;104:957-63. 22. Wang LY, Zhong Y, Wheeler L. Direct and indirect costs of asthma in school-age children. Prev Chronic Dis 2005;2:A11. 23. Weinmann S, Kamtsiuris P, Henke K-D, Wickman M, Jenner A, Wahn U. The costs of atopy and asthma in children: assessment of direct costs and their determinants in a birth cohort. Pediatr Allergy Immunol 2003;14:18-26. 24. Smith DH, Malone DC, Lawson KA, Okamoto LJ, Battista C, Saunders WB. A national estimate of the economic costs of asthma. Am J Respir Crit Care Med 1997;156:787-93. 25. Weiss KB, Gergen PJ, Hodgson TA. An economic evaluation of asthma in the United States. N Engl J Med 1992;326:862-6. 26. Weiss KB, Sullivan SD. Understanding the costs of asthma: the next step. CMAJ 1996;154:841-3. 27. Weiss KB, Sullivan SD. The health economics of asthma and rhinitis. I. Assessing the economic impact. J Allergy Clin Immunol 2001;107:3-8. 28. Table 20 Economic Cost of Asthma, United States, 2007. Available at: http://www. lungusa.org/assets/documents/ASTHMA-JAN-2009.pdf. Accessed January 18, 2010. 29. Tolpin HG, Bentkover JD. Economic cost of illness: decision-making applications and practical considerations. Adv Health Econ Health Serv Res 1983;4:165-98. 30. Akobundu E, Ju J, Blatt L, Mullins CD. Cost-of-illness studies: a review of current methods. Pharmacoeconomics 2006;24:869-90.
BARNETT AND NURMAGAMBETOV 151
31. Honeycutt AA, Segel JE, Hoerger TJ, Finkelstein EA. Comparing cost-of-illness estimates from alternative approaches: an application to diabetes. Health Serv Res 2009;44:303-20. 32. Finkelstein EA, Fiebelkorn IC, Wang GJ. National medical spending attributable to overweight and obesity: how much, and who’s paying? Health Aff (Millwood) 2003;22:W219-26. 33. Howard DH, Molinari NA, Thorpe KE. National estimates of medical costs incurred by nonelderly cancer patients. Cancer 2004;100:883-91. 34. Kamble S, Bharmal M. Incremental direct expenditure of treating asthma in the United States. J Asthma 2009;46:73-80. 35. Trogdon JG, Finkelstein EA, Hoerger TJ. Use of econometric models to estimate expenditure shares. Health Serv Res 2008;43:1442-52. 36. Trogdon JG, Finkelstein EA, Nwaise IA, Tangka FK, Orenstein D. The economic burden of chronic cardiovascular disease for major insurers. Health Promot Pract 2007;8:234-42. 37. MEPS HC-105: 2006 full year consolidated data file, 2008. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h105/h105doc.pdf. Accessed June 12, 2009. 38. MEPS HC-070 2002 full year consolidated data file, 2004. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h70/h70doc.pdf. Accessed June 12, 2009. 39. MEPS HC-079 2003 full year consolidated data file, 2005. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h79/h79doc.pdf. Accessed June 12, 2009. 40. MEPS HC-089 2004 full year consolidated data file, 2006. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h89/h89doc.pdf. Accessed June 12, 2009. 41. MEPS HC-097 2005 full year consolidated data file, 2007. Available at: http://www. meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h97/h97doc.pdf. Accessed June 12, 2009 42. MEPS HC-113 2007 full year consolidated data file, 2009. Available at: http://www. meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h113/h113doc.pdf. Accessed March 29, 2010 43. StataCorp. Stata Statistical Software: Release 11. College Station (TX): StataCorp; 2007. 44. Machlin S, Yu, W, Zodet M. Computing standard errors for MEPS estimates, 2005. Available at: http://www.meps.ahrq.gov/mepsweb/survey_comp/standard_errors. jsp. Accessed May 1, 2009. 45. Table 1A. consumer price index for all urban consumers (CPI-U): U.S. city average, by expenditure category and commodity and service group, 2002–2009. Available at: http://www.bls.gov/cpi/cpi_dr.htm#2007. Accessed June 9, 2009. 46. MEPS HC-102G 2006 office based medical provider visits, 2008. Available at: http:// www.meps.ahrq.gov/mepsweb/data_stats/download_data/pufs/h102g/h102gdoc.pdf. Accessed August 7, 2009. 47. Wakefield DB, Cloutier MM. Modifications to HEDIS and CSTE algorithms improve case recognition of pediatric asthma. Pediatr Pulmonol 2006;41:962-71. 48. Centers for Disease Control and Prevention. Early release of selected estimates based on data from the 2008 National Health Interview Survey, 2009. Available at: http:// www.cdc.gov/nchs/data/nhis/earlyrelease/earlyrelease200906.pdf. Accessed July 6, 2009. 49. D’Hoore W, Bouckaert A, Tilquin C. Practical considerations on the use of the Charlson comorbidity index with administrative data bases. J Clin Epidemiol 1996;49:1429-33. 50. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373-83. 51. Schneeweiss S, Maclure M. Use of comorbidity scores for control of confounding in studies using administrative databases. Int J Epidemiol 2000;29:891-8. 52. Manning WG, Mullahy J. Estimating log models: to transform or not to transform? J Health Econ 2001;20:461-94. 53. Buntin MB, Zaslavsky AM. Too much ado about two-part models and transformation?: comparing methods of modeling Medicare expenditures. J Health Econ 2004;23:525-42. 54. Duan N, Manning WG, Morris CN, Newhouse JP. A comparison of alternative models for the demand for medical care. J Bus Econ Stat 1983;1:115-26. 55. Duan N, Manning WG, Morris CN, Newhouse JP. Choosing between the sampleselection model and the multi-part model. J Bus Econ Stat 1984;2:283-9. 56. Vistnes JP. Gender differences in days lost from work due to illness. Ind Labor Relat Rev 1997;50:304-23. 57. Jonsson B. Measuring the economic burden of asthma. In: Weiss KB, Buist AS, Sullivan SD, editors. Asthma’s impact on society: the social and economic burdenNew York: Markel Dekker; 2000. p. 251-67. 58. Weisbrod B. Economics of public health. Philadelphia: University of Pennsylvania Press; 1961.
152 BARNETT AND NURMAGAMBETOV
59. Median weekly earning of full-time wage and salary workers by detailed occupation and sex—continued, 2002–2006. Available at: ftp://ftp.bls.gov/pub/special. requests/lf/. Accessed April 18, 2009. 60. Efron B, Tibshirani R. An Introduction to the Bootstrap. New York: Chapman & Hall; 1993. 61. Kolenikov S. Resampling variance estimation for complex survey data. Stata J 2010;10:165-99. 62. Heron M, Hoyert DL, Murphy SL, Xu J, Kochanek KD, Tejada-Vera B. Deaths: final data for 2006. Natl Vital Stat Rep 2009;57:1-134. 63. Hoyert DL, Heron MP, Murphy SL, Kung HC. Deaths: final data for 2003. Natl Vital Stat Rep 2006;54:1-120. 64. Kochanek KD, Murphy SL, Anderson RN, Scott C. Deaths: final data for 2002. Natl Vital Stat Rep 2004;53:1-115. 65. Kung HC, Hoyert DL, Xu J, Murphy SL. Deaths: final data for 2005. Natl Vital Stat Rep 2008;56:1-120. 66. Minino AM, Heron MP, Murphy SL, Kochanek KD. Deaths: final data for 2004. Natl Vital Stat Rep 2007;55:1-119. 67. Xu J, Kochanek MA, Murphy SL, Tejada-Vera B. Deaths: final data for 2007. Natl Vital Stat Rep 2009;58:1-135. 68. Grosse SD, Krueger KV, Mvundura M. Economic productivity by age and sex: 2007 estimates for the United States. Med Care 2009;47(suppl 1):S94-103. 69. MEPS annual methodology report Rockville (MD): Westat; 2009, Mar. Deliverable No.:142. Contract No.:290-02-0005. Sponsored by the Agency for Healthcare Research and Quality. 70. Cherry DK, Hing E, Woodwell DA, Rechtsteiner EA. National Ambulatory Medical Care Survey: 2006 summary. Natl Health Stat Rep 2008;3):1-39. 71. DeFrances CJ, Lucas CA, Buie VC, Golosinskiy A. 2006 National Hospital Discharge Survey. Natl Health Stat Rep 2008;(5):1-20. 72. Schappert SA, Rechtsteiner EA. Ambulatory medical care utilization estimates for 2006. Natl Health Stat Rep 2008;(8):1-29.
J ALLERGY CLIN IMMUNOL JANUARY 2011
73. Moonie SA, Sterling DA, Figgs L, Castro M. Asthma status and severity affects missed school days. J Sch Health 2006;76:18-24. 74. Silverstein MD, Mair JE, Katusic SK, Wollan PC, O’Connell ED, Yunginger JW. School attendance and school and school performance: a population-based study of children with asthma. J Pediatr 2001;139:278-83. 75. Weiss KB, Sullivan SD, Lyttle CS. Trends in the cost of illness for asthma in the United States, 1985-1994. J Allergy Clin Immunol 2000;106:493-9. 76. Aligne CA, Auinger P, Byrd RS, Weitzman M. Risk factors for pediatric asthma. Contributions of poverty, race, and urban residence. Am J Respir Crit Care Med 2000;162:873-7. 77. Christiansen SC, Martin SB, Schleicher NC, Koziol JA, Mathews KP, Zuraw BL. Current prevalence of asthma-related symptoms in San Diego’s predominantly Hispanic inner-city children. J Asthma 1996;33:17-26. 78. Persky VW, Slezak J, Contreras A, Becker L, Hernandez E, Ramakrishnan V, et al. Relationships of race and socioeconomic status with prevalence, severity, and symptoms of asthma in Chicago school children. Ann Allergy Asthma Immunol 1998;81:266-71. 79. Brugge D, Woodin M, Schuch TJ, Salas FL, Bennett A, Osgood ND. Communitylevel data suggest that asthma prevalence varies between U.S. and foreign-born black subpopulations. J Asthma 2008;45:785-9. 80. Holguin F, Mannino DM, Anto J, Mott J, Ford ES, Teague WG, et al. Country of birth as a risk factor for asthma among Mexican Americans. Am J Respir Crit Care Med 2005;171:103-8. 81. Brim SN, Rudd RA, Funk RH, Callahan DB. Asthma prevalence among US children in underrepresented minority populations: American Indian/ Alaska Native, Chinese, Filipino, and Asian Indian. Pediatrics 2008;122: e217-22. 82. Home-based multi-trigger multicomponent environmental interventions, 2009. Available at: http://www.thecommunityguide.org/asthma/multicomponent.html. Accessed August 7, 2009.