Chest ACCP Member Benefits
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mullins, C. D.
Right arrow Articles by Stoller, J. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mullins, C. D.
Right arrow Articles by Stoller, J. K.
(Chest. 2003;124:826-831.)
© 2003 American College of Chest Physicians

Major Components of the Direct Medical Costs of {alpha}1-Antitrypsin Deficiency*

C. Daniel Mullins, PhD; Junling Wang, MS and James K. Stoller, MD, MS, FCCP

* From the University of Maryland School of Pharmacy (Dr. Mullins and Ms. Wang), Center on Drugs and Public Policy, Baltimore, MD; and the Department of Pulmonary and Critical Care Medicine (Dr. Stoller), The Cleveland Clinic Foundation, Cleveland, OH.

Correspondence to: C. Daniel Mullins, PhD, University of Maryland, Center on Drugs and Public Policy, Second Floor, 515 W Lombard St, Baltimore, MD 21201; e-mail: dmullins{at}rx.umaryland.edu


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Study objectives: To examine the sources of the direct medical costs of {alpha}1-antitrypsin (AAT) deficiency based on survey data from affected individuals.

Background: Prior research has reported the total cost of AAT deficiency but has not examined the specific components of the direct medical costs of affected individuals.

Methods: In order to detail the sources of the direct medical costs, we sent surveys to 688 respondents of a previous survey. We estimated the costs in three ways, which differed in the method of managing missing values. With method 1, the group mean value of cost per unit of utilization, multiplied by the occurrences of utilization, was used to replace the missing value. Two sensitivity analyses (ie, methods 2 and 3) were conducted to test the robustness of our estimate. With method 2, values of zero were entered for all missing values. With method 3, the missing values were replaced by the group mean value. The Wilcoxon test was used to test the cost differences between patients of different phenotypes. All cost data were expressed in 1998 US dollars.

Results and conclusions: Two hundred ninety-two individuals responded to the survey. The annual total health-care costs were high (mean range, $36,471 to $46,114; median range, $12,485 to $37,100 [according to the method for managing missing data]) for AAT deficiency. The total costs for individuals with the PI*ZZ phenotype exceeded those for individuals with a non-PI*ZZ phenotype. The use of IV augmentation therapy accounted for more than half of all direct medical costs for the respondents. Besides the costs for therapy with {alpha}1-proteinase inhibitor (Prolastin; Bayer; West Haven, CT), other major cost sources were prescription drugs other than {alpha}1-proteinase inhibitor, hospitalization, health insurance, and physician visits.

Key Words: {alpha}1-antitrypsin deficiency • cost of illness • economics • phenotype • registries


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
{alpha}1-antitrypsin (AAT) deficiency is an autosomal codominant condition that predisposes an individual to emphysema and liver disease, and is reasonably common (ie, 1 in 1,575 to 1 in 5,000 individuals) but widely underrecognized by clinicians.1 2 The currently available therapy for individuals with AAT deficiency-related COPD consists of the usual treatments for emphysema (eg, bronchodilator therapy, pulmonary rehabilitation, and oxygen therapy where indicated) and of specific treatment of AAT deficiency. The latter treatment consists of the infusion of pooled human plasma antiprotease (ie, {alpha}1-proteinase inhibitor), which is called augmentation therapy (currently available in the United States as the drug Prolastin; Bayer; West Haven, CT).

In the context of the recent attention to the cost of treating common illnesses, we have focused previously and in the current report on the costs of treatment reported by individuals with AAT deficiency. The current study extends our prior research, which reported mean annual health-care expenditures by AAT-deficient individuals of $30,948, for patients with the PI*ZZ phenotype, and $20,673, for patients with a non-PI*ZZ phenotype,3 by assessing the specific sources of the total cost. Also, in contrast to prior decision analyses,4 5 these analyses are, to our knowledge, the first to estimate the direct medical expenses of AAT deficiency based on survey data from affected individuals.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
In the current study, a questionnaire (available from Dr. Mullins on request) addressing the previous 12-month period was sent to the 688 respondents to our earlier report.3 The questionnaire addressed the following areas: demographic characteristics; utilization of health services due to AAT deficiency and other health problems; and treatment costs for AAT deficiency. Specifically, information was sought regarding costs due to the following: {alpha}1-proteinase inhibitor; health insurance premiums; other prescription drugs; physician visits; emergency department visits; hospitalizations; and long-term care costs, where long-term care is defined as "health, mental health, social, and residential services provided to a temporarily or chronically disabled person over an extended period of time with a goal of enabling the person to function as independently as possible."6 The continuum of long-term care contains extended inpatient care, acute inpatient care, ambulatory care services, home care, outreach programs, wellness programs, and housing.6 As one of these categories, home care "represents a variety of nursing, therapy, and support services provided to people who are homebound and have some degree of illness or functional disability, but who are able to satisfy their needs by bringing services into the home setting."6 In completing the questionnaires, subjects were asked to review their actual medical bills. For each cost item, except for insurance premiums, subjects were asked to indicate whether the costs were paid by themselves, their insurance company, or by others (eg, family members). For insurance premiums, the three sources were patients, employers, and others. All costs were expressed in 1998 US dollars.

Three separate methods for imputing missing values were undertaken. In general, missing values were imputed by considering whether the patient reported using a specific service/item, whether the patient reported any costs for that service/item, and whether the patient reported costs by specific sources of expenditure (ie, self, insurance, or other). With method 1, first, the missing values for long-term care costs and prescription drug costs were entered as zero. For other cost items, if a patient did not report an expense with at least one of the three sources of payment (ie, self, insurance, or other) and did not report any utilization for that service/item, the amount for that cost item was considered to be zero. If the patient reported an expense from at least one of the three payment sources for the item, the amount was the sum across the three payment sources (ie, self + insurance + other), while substituting a value of zero only for the missing value components for that cost item. Finally, if a patient reported the utilization of a resource but did not report any costs associated with that item, then the missing value was imputed by using the group mean value per unit of utilization multiplied by the actual quantity used. These methods are summarized in Table 1 .


View this table:
[in this window]
[in a new window]

  		Table 1. Description of the Methods for Imputing Missing Values

 
In order to test the robustness of our estimate, two sensitivity analyses were performed, which will be referred to as methods 2 and 3. With method 2, which was a lower-bound analysis, values of zero were used for all missing values. With method 3, the group mean value was used for all missing values when the patient only reported using the service but did not report the costs. A value of zero was entered otherwise. Method 3 differs from method 1 in that it ignores the reported number of times a respondent used a particular health service. Among the three methods that were used, method 1 incorporated the most information that was available from the survey. It used both information about the cost of a cost item/health service and information about the quantity of utilization of the health service. Thus, method 1 likely provided the most accurate and unbiased estimate among the three methods used.

To exemplify the differences among these three methods, we will consider how the missing values for the cost of physician visits were handled with each method. In method 1, the mean cost value per physician visit first was calculated for the sample of respondents who reported the cost of physician visits. Next, for patients who reported the number of visits but not the associated costs, the number of physician visits for that patient was multiplied by the mean cost value of all physician visits. For example, the number of physician visits (Y) was multiplied by the average cost (X dollars) such that the missing cost value (Z) was represented by the formula XxY = Z. If the individual did not report either a cost for a physician visit or a number of visits, the value of zero was entered for the missing value of physician visit costs. Missing values for emergency department visits and hospitalization cost items were imputed in the same way as that for physician visits. For health insurance costs, if a patient had health insurance coverage at the time of the survey, the missing value was entered as the group mean value. Otherwise, zero was entered.

Using method 2, values of zero were entered for all missing values. In method 3, the mean value of physician visits was calculated first. Then, whenever the subject reported having physician visits but did not report the actual cost, the calculated mean cost was used as the total physician visit cost. As an example, if the mean physician visit cost was calculated as Z dollars, the missing physician visit cost for patients who reported physician visits was replaced by Z dollars, regardless of their reported number of visits.

Using method 3, values for emergency department visits, hospitalization, and health insurance costs were handled identically to physician visit costs. Specifically, if a patient had health insurance coverage at the time of the survey, used the emergency department, or was hospitalized, the missing value was entered as the group mean value. Otherwise, zero was entered.

According to our imputation methods, cost analyses were stratified by separate cost items (eg, {alpha}1-proteinase inhibitor or health insurance premiums) and by different patient groups (eg, patients with PI*ZZ phenotype vs those with other phenotypes). Descriptive statistics were calculated to describe the demographic characteristics of the survey respondents. A nonparametric test (ie, Wilcoxon rank sum) was used to compare the costs between groups because the cost data were not normally distributed. One-sided tests were used because the costs for individuals with the PI*ZZ phenotype were expected to exceed those of individuals with other phenotypes. p Values < 0.05 were deemed statistically significant.

The study was reviewed and was determined to be exempt by the Institutional Review Board of the University of Maryland (exemption No. DM-119801).


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Responses were received from 292 of the 688 individuals to whom questionnaires had been sent (42.4%). The response rates for PI*ZZ individuals was 42.7% vs 41.8% for non-PI*ZZ subjects. Of the 292 respondents, 23 were excluded from analysis (20 because the questionnaires lacked cost information, and 3 because the total cost estimates were outliers, exceeding the group mean values by 10 times the SD). Thus, the results presented are based on 269 respondents (39.1% of those persons who had been sent questionnaires).

Table 2 presents the demographic features and clinical characteristics of our sample group. The mean age of subjects was 52.0 years. More men than women participated, and, in keeping with the rarity of AAT deficiency in non-white groups, most respondents were white (98.1%). Most respondents were former smokers (72.9%). Phenotype information was reported by 264 of the 269 respondents (98.1%), of whom 80.7% reported having the PI*ZZ phenotype.


View this table:
[in this window]
[in a new window]

  		Table 2. Demographic Characteristics of Respondents*

 
Table 3 presents the health services utilization of the respondents. The mean (± SD) number of physician visits for all health problems was 8.5 ± 9.6, with most subjects reporting one to five visits per year both for AAT deficiency only and for all health problems. As expected, respondents used fewer emergency department and hospitalization services than physician visits. Indeed, most respondents did not visit the emergency department for AAT deficiency, and few had been hospitalized for any reason.


View this table:
[in this window]
[in a new window]

  		Table 3. Use of Health Services by Respondents

 
Table 4 presents estimates of the direct medical costs of AAT deficiency. Estimates of the mean (± SD) total costs ranged from a maximum value of $46,114 ± $49,664 (using method 1 for imputing missing values) to a minimum value of $36,471 ± $47,929 (using method 2). As a measure of the potential impact of the imputing method used, 46.3% of all cost fields contained missing data. The inability to determine when missing values represent zeros vs unknown values prompted our high and low sensitivity analyses.


View this table:
[in this window]
[in a new window]

  		Table 4. Direct Medical Costs for Individuals With AAT Deficiency*

 
For all three methods, the order of the five largest cost sources was identical, with {alpha}1-proteinase inhibitor accounting for the largest share (ie, 55.4 to 60.9%), followed by prescription drugs (14.0 to 17.7%), hospitalization (9.5 to 10.3%), health insurance (6.9 to 7.3%), and physician visits (5.0 to 5.4%). Whether the costs for emergency department visits exceeded those for long-term care differed by the imputation method used. As expected because of {alpha}1-proteinase inhibitor use, the costs for individuals with the PI*ZZ phenotype exceeded those for individuals with non-PI*ZZ phenotypes by $8,028 (using method 2) to $8,790 (using method 1). These differences between costs for PI*ZZ and for non-PI*ZZ individuals failed to achieve statistical significance. On average, ex-smokers and current smokers reported higher costs than those reported by nonsmokers by $2,889 (using method 2) to $3,887 (using method 3). The differences between these two groups of patients did not achieve statistical significance.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The results of this survey of direct health-care costs associated with AAT deficiency show the following: (1) annual total health-care costs are high (mean range, $36,471 to $46,114; median range, $12,485 to $37,100), with estimates varying (by up to 20.9 to 66.3%) according to the method used for managing missing data; (2) total costs for PI*ZZ individuals showed a trend to exceed those for non-PI*ZZ individuals; (3) the use of IV augmentation therapy accounted for more than half of all health-care costs for the respondents, regardless of the method used for imputing missing values; and (4) besides {alpha}1-proteinase inhibitor, which accounted for the largest proportion of health-care costs, other major cost sources were prescription drugs other than {alpha}1-proteinase inhibitor, hospitalization, health insurance, and physician visits.

By assessing the specific health-care cost sources for individuals with AAT deficiency, the current study extends the results of our earlier report.3 Indeed, the estimate of the total health-care costs for AAT-deficient individuals in the current report is higher than the earlier reported value (PI*ZZ patients: current study, $48,462; our previous survey, $30,948; non-PI*ZZ patients: current study, $39,672; our previous study, $20,673), likely because the current survey focused more closely on the sources of cost and was, it seems, likely to identify costs that may have not been reported earlier. Specifically, in contrast to the earlier survey, the current survey sought information on itemized cost components.

That PI*ZZ patients incur higher costs than non-PI*ZZ patients is expected. Among all different phenotypes, the PI*ZZ phenotype is associated with the lowest level of AAT activity (10 to 15%).7 PI*ZZ patients are more likely to be prescribed {alpha}1-proteinaseinhibitor, which is an expensive therapy and accounts for the largest share of direct medical costs of treatment, as reported in our study.

Our finding higher costs in ex-smokers and current smokers than in nonsmokers is consistent with previous research. For example, McElvaney and colleagues8 reported that in the National Heart, Lung, and Blood Institute Registry, most AAT-deficient individuals with relatively normal lung function never had smoked. Among individuals with more severe airflow obstruction, most were ex-smokers and current smokers.8 Because cigarette smoking clearly is associated with accelerated airflow obstruction, smoking would be expected to be associated with higher costs for AAT-deficient individuals.

Because {alpha}1-proteinase inhibitor constitutes the largest share of the direct medical costs, we compared our estimate of this cost item with those from other studies. Table 5 compares results across studies, with all cost estimates adjusted to 1998 US dollars using consumer price indexes.9 Notably, the estimated cost in the current study ($28,075) was lower than our previous estimate of $32,606 to $67,430 (depending on infusion frequency and the recipient’s weight), as well as the estimates from all other available studies. In the context that overall health-care costs are higher here than our previously estimated amount, we suspect that {alpha}1-proteinase inhibitor costs are lower in the current study because a greater proportion of subjects currently were receiving {alpha}1-proteinase inhibitor less frequently (ie, with up to 54.2% of currently reported patients receiving {alpha}1-proteinase inhibitor biweekly or less frequently and with periods of less frequent use related to unavailability of the drug).


View this table:
[in this window]
[in a new window]

  		Table 5. Comparison of Prolastin Costs in Available Studies

 
The high cost of having AAT deficiency is underscored by comparison with cost estimates of having other pulmonary conditions. For example, compared with estimates from two studies (Table 6 ) of the annual health-care costs of individuals with COPD, AAT deficiency-associated costs are much higher. Specifically, based on the 1997 National Medical Expenditure Survey, Strassels et al10 reported mean direct medical expenditures of $6,469 in 1987, which adjusts to $9,282 in 1998 US dollars, the reference used in the current study. Stratifying the health-care expenses of COPD patients by disease severity, Hilleman et al11 reported median treatment costs of $1,681 for stage I patients, $5,037 for stage II patients, and $10,812 for stage III patients. Not only were the total costs of health-care higher for AAT-deficient individuals than for patients with COPD, but the main cost sources differed also. The leading cost driver for individuals with AAT deficiency was {alpha}1-proteinase inhibitor, whereas for AAT-replete COPD patients, hospitalization accounted for the largest fraction of costs (68% and 40 to 63%, respectively) in the studies by Strassels et al10 and Hilleman et al.11


View this table:
[in this window]
[in a new window]

  		Table 6. Annual Health-Care Costs Associated With Several Conditions

 
In extending the comparison beyond COPD, Table 6 also compares estimates of health-care costs of AAT deficiency with those of a variety of other chronic conditions. As shown, AAT deficiency-associated costs are 3-fold to > 10-fold higher.

Several shortcomings in the current study warrant comment. First, because our cost estimates were based on responses from only 39% of the potential respondents, we cannot discount the possibility of selection bias. Another shortcoming of this study is that our estimates are based on participants’ self-reports and so are limited by the inaccuracies of respondents’ knowledge and/or recall. Indeed, our results will overestimate the true cost of having AAT deficiency to the extent that subjects not receiving {alpha}1-proteinase inhibitor are underrepresented. Still, the fact that the cost estimates in the current study, although still high, are lower than other estimates of augmentation therapy-associated costs weighs against our having significantly overestimated costs. Also, the lack of spirometry values precludes an analysis of the relationship between FEV1 and cost. The expectation would be that individuals with better preserved airflow would require fewer physician visits and hospitalizations, and, therefore, would incur lower costs than those with worse airflow obstruction.

In conclusion, our current study showed even higher costs of AAT deficiency compared with our previous study. In the context that {alpha}1-proteinase inhibitor therapy is costly, it is not surprising that the cost of {alpha}1-proteinase inhibitor constitutes the largest share of the direct medical costs.


   Footnotes
 
Abbreviation: AAT = {alpha}1-antitrypsin

This research was funded by the Alpha-1 Foundation.

Received for publication August 28, 2002. Accepted for publication January 30, 2003.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Sociedad Española de Neumología y Cirugía Toracica. World Health Organization’s meeting on Alpha-1-Antitrypsin Deficiency. Available at: http://www.separ.es/areas/interes5.htm. Accessed August 11, 2003
  2. Stoller, JK, Smith, P, Yang, P, et al Physical and social impact of {alpha}1-antitrypsin deficiency. Cleve Clin J Med 1994;61,461-467[ISI][Medline]
  3. Mullins, CD, Huang, X, Merchant, S, et al The direct medical costs of {alpha}1-antitrypsin deficiency. Chest 2001;119,745-752[Abstract/Free Full Text]
  4. Hay, JW, Robin, ED Cost-effectiveness of alpha-1 antitrypsin replacement therapy in treatment of congenital chronic obstructive pulmonary disease. Am J Public Health 1991;81,427-433[Abstract/Free Full Text]
  5. Alkins, SA, O’Malley, P Should health-care systems pay for replacement therapy in patients with {alpha}1-antitrypsin deficiency? A critical review and cost-effectiveness analysis. Chest 2000;117,875-880[Abstract/Free Full Text]
  6. Evashwick, CJ The continuum of long-term care. Williams, SJ Torrens, PR eds. Introduction to health services 6th ed. 2002,234-279 Delmar Publishers. Albany, NY:
  7. Gelehrter, TD, Collins, FS, Ginsburg, D Principles of medical genetics 2nd ed. 1998,124-126 Williams & Wilkins. Baltimore, MD:
  8. McElvaney, NG, Stoller, JK, Buist, AS, et al Baseline characteristics of enrollees in the National Heart, Lung and Blood Institute Registry (NHLBI) of alpha 1-antitrypsin deficiency: Alpha 1-Antitrypsin Deficiency Registry Study Group. Chest 1997;111,394-403[Abstract/Free Full Text]
  9. US Department of Labor Bureau of Labor Statistics. Consumer price indexes: inflation calculator. Available at: http://www.bls.gov/cpi/#data. Accessed August 11, 2003
  10. Strassels, SA, Smith, DH, Sullivan, SD, et al The cost of treating COPD in the United States. Chest 2001;119,344-352[Abstract/Free Full Text]
  11. Hilleman, DE, Dewan, N, Malesker, M, et al Pharmacoeconomic evaluation of COPD. Chest 2000;118,1278-1285[Abstract/Free Full Text]
  12. Smith, DH, Malone, DC, Lawson, KA, et al A national estimate of the economic costs of asthma. Am J Respir Crit Care Med 1997;156,787-793[Abstract/Free Full Text]
  13. Ernst, RL, Hay, JW, Fenn, C, et al Cognitive function and the costs of Alzheimer disease: an exploratory study. Arch Neurol 1997;54,687-693[Abstract]
  14. Lanes, SF, Lanza, LL, Radensky, PW, et al Resource utilization and cost of care for rheumatoid arthritis and osteoarthritis in a managed care setting: the importance of drug and surgery costs. Arthritis Rheum 1997;40,1475-1481[ISI][Medline]
  15. Brown, JB, Nichols, GA, Glauber, HS, et al Type 2 diabetes: incremental medical costs during the first 8 years after diagnosis. Diabetes Care 1999;22,1116-1124[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 ThoraxHome page
J K Stoller and L S Aboussouan
{alpha}1-Antitrypsin deficiency {middle dot} 5: Intravenous augmentation therapy: current understanding
Thorax, August 1, 2004; 59(8): 708 - 712.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mullins, C. D.
Right arrow Articles by Stoller, J. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mullins, C. D.
Right arrow Articles by Stoller, J. K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS