This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (62) Citing Articles via Google Scholar Google Scholar Articles by Druss, B. G. Articles by Pincus, H. A. Search for Related Content PubMed PubMed Citation Articles by Druss, B. G. Articles by Pincus, H. A. Related Collections Chronic Care Health Spending

DataWatch

Comparing The National Economic Burden Of Five Chronic Conditions

Abstract

 
Using a nationally representative sample of 23,230 U.S. residents, we examine patterns of economic burden across five chronic conditions: mood disorders, diabetes, heart disease, asthma, and hypertension. Almost half of U.S. health care costs in 1996 were borne by persons with one or more of these five conditions; of that spending amount, only about one-quarter was spent on treating the conditions themselves and the remainder on coexistent illnesses. Each condition demonstrated substantial economic burden but also unique characteristics and patterns of service use driving those costs. The findings highlight the differing challenges involved in understanding needs and improving care across particular chronic conditions.

This study uses a nationally representative sample of the U.S. population to compare demographic characteristics and costs and patterns of service use across five prevalent, costly, and disabling chronic conditions: mood disorders (that is, depressive and manic-depressive disorders), diabetes, heart disease, hypertension, and asthma.³ We hypothesized that while all would entail substantial economic burden, each would also have unique features resulting in distinct clinical and health policy challenges.

Study Data And Methods

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 
The 1996 Medical Expenditure Panel Survey (MEPS) was conducted by the Agency for Healthcare Research and Quality (AHRQ) to provide nationally representative estimates of health care use, spending, sources of payment, and insurance coverage for the U.S. civilian noninstitutionalized population.⁴ Overall, 23,200 eligible persons responded to the core MEPS, which represented a final response rate of 70.2 percent.

Analyses calculated the proportion of persons with a given condition who received any treatment for it and, among those who received any treatment, the costs for treatment for the condition. Costs included (1) mean per capita health costs for services directly resulting from each condition (both across the total sample and stratified by insurance status), (2) mean per capita health costs for persons with the condition, and (3) costs associated with work loss for persons with the condition. We estimated these three components as the sum of the individual costs weighted to represent the entire U.S. population. Statistical tests for comparisons across conditions were made using Z-tests comparing a given disorder to the condition closest to it on a given parameter. To provide nationally representative estimates and to account for the multistage sampling design of MEPS, all analyses used the SUDAAN statistical package, with appropriate weighting and nesting variables.⁵

Study Results

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 
Prevalence and characteristics. Participants with one or more of the five conditions represented an estimated 25 percent of the U.S. population. Hypertension was by far the most prevalent condition, with an estimated national prevalence of 10.2 percent (p <.001 compared with mood disorders).⁶ This was followed by mood disorders (the vast majority of which were depressive disorders), asthma, diabetes, and ischemic heart disease (Exhibit 1). Comorbidity among the conditions was most common in patients with diabetes and ischemic heart disease, of whom 55.5 percent and 60.8 percent, respectively, had at least one of the other four conditions (p <.001 for diabetes compared with hypertension). Those with asthma were most likely to be younger (p <.001 compared with mood disorders) and nonwhite (p < .001 compared with hypertension) than those with the other four conditions, and almost a quarter were living below the federal poverty level (p =.02 compared with diabetes).

When we examined the costs of all health services used by persons with a given condition, we found a different pattern. At the national level, persons with hypertension accrued almost twice the per capita costs of those with any of the other conditions—largely because of the condition’s high prevalence. Only 13.4 percent of spending among persons with hypertension nationally was for direct treatment for the disorder; the remainder was for treatment for coexistent conditions. In contrast, 18.6 percent of the total spending among persons with mood disorders and with diabetes, 20.7 percent among persons with asthma, and 55.8 percent among persons with heart disease was for direct treatment of the index condition.

Persons with heart disease paid the least out of pocket for treatment of their condition (p <.001 compared with diabetes). Those with the other conditions all paid approximately 40 percent of total health costs out of pocket (Exhibit 2).

Work loss and total costs. Ischemic heart disease was almost twice as likely as any of the other conditions to be the cause of one or more missed workdays (p <.001 compared with asthma). However, because of its relatively low prevalence, persons with this disease incurred relatively low work-loss costs nationally (Exhibit 3).

Use of services and costs by type of coverage. Aggregated across all types of insurance, more than four-fifths of persons with diabetes, heart disease, asthma, and hypertension received at least some care (outpatient, inpatient, or medication) for the condition (Exhibit 4). In contrast, only two-thirds of persons with mood disorders received any treatment for their condition during the year (p <.001 compared with asthma).

Discussion And Policy Implications

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 
The five conditions we examined all demonstrated substantial costs, but each also varied along key characteristics. Each of these five parameters plays an important role in determining total costs in patients with chronic illnesses; each, in turn, may have distinct implications for improving their medical care.

Comorbidity and prevalence. We found a large difference between costs directly attributed to the five conditions and the total costs borne by persons with these conditions. This gap is primarily a function of the number and costs of coexistent illnesses. Since a comorbid disorder may be either independent or caused by an index condition, the "true" costs of any chronic condition likely fall somewhere between those attributed to a particular condition and the total costs incurred by persons with that condition.

Diabetes exemplifies how comorbid conditions such as infections, peripheral vascular disease, and microvascular damage account for much of the clinical and economic burden of disease. Therefore, these sequelae represent an important target for secondary prevention among diabetics, including screening and early intervention for complications.⁷ Because most health care settings are poorly equipped to provide these coordinated services, disease management team models have been developed.⁸ These teams have rapidly proliferated in managed care settings; however, many bear little resemblance to the best-practice models described in the literature and hence may be limited in their effectiveness and reach.⁹

For hypertension, the primary cost driver was its prevalence—more than twice that of any of the other four conditions. High disease prevalence is one of the most important factors determining appropriateness of population-based screening efforts, since it increases diagnostic yield and also the public health benefits of early detection.¹⁰ However, public funding for primary prevention is limited; Medicare does not cover screening for hypertension, and adding new preventive services requires an act of Congress. The Medicare Wellness Act of 2001, under review in Congress, would add coverage for hypertension screening (as well as for cholesterol screening and smoking-cessation counseling) and also would streamline the process for including new preventive services in the future.

Treatment costs. Ischemic heart disease posed the highest direct treatment costs, largely because of the disproportionate use of inpatient care. Many of the treatments available to treat ischemic heart disease, from angiography to cardiac surgery, are among the most costly medical treatments today. In part because these treatments are both expensive and highly visible to insurers (who shoulder most of the expenditures), they have become priority areas for clinical guideline development. If there is even a small increase in the prevalence of ischemic heart disease as the U.S. population ages, there will be a major increase in spending for this condition. The high cost of these interventions, and the potential for even higher costs to society in the coming years, make rigorous cost-effectiveness/cost-benefit analysis of paramount importance.¹¹

Sociodemographics. The importance of sociodemographic variables in chronic conditions is best illustrated by asthma. Low-income and minority patients, as well as children, may be at increased risk not only for developing asthma but also for reduced access to appropriate care.¹² Quality improvement programs for chronic illness will be most useful when targeted toward subpopulations with elevated prevalence of disease or barriers to treatment. For asthma, educational programs for poor and minority patients have been shown not only to improve clinical outcomes but also to redistribute services away from emergency room and inpatient visits and toward more appropriate and less costly outpatient care.¹³

Coverage. The impact of lack of insurance on access to and intensity of treatment was greatest for mood disorders. Furthermore, because the presence of general health insurance is only a weak proxy for mental health coverage, the findings likely underestimate the actual gap in access to services between patients with and without full mental health benefits. While much of the mental health policy debate in recent years has focused on achieving parity of mental health benefits among insured persons, this study’s findings serve as a reminder that many persons with mental disorders have no insurance at all.¹⁴ Increasing rates of coverage, important for all persons with chronic conditions, may be a particular priority for those with mental disorders.

Managed care. There was not a significant difference in treatment under managed and unmanaged plans for any of the five conditions. This finding likely reflects a progressive blurring of the lines between traditional health maintenance organizations (HMOs) and fee-for-service plans, occurring as traditional indemnity plans have adopted managed care techniques and as managed care plans have adopted a host of blended payment methods.¹⁵

Study limitations. Several limitations should be discussed. First, the health conditions were determined by self-report. Reassuringly, internal MEPS studies have found an approximately 70 percent correlation between patient- and provider-reported diagnoses and also have found that the correlation is highest for inpatient hospitalizations, which comprise the bulk of expenditures.¹⁶ Second, measures of indirect costs were more limited than were measures of health costs. These costs were calculated only among workers (and so did not account for persons who were unemployed), and work loss could only be calculated as a total per person, rather than being attributed to specific conditions. Finally, the study design necessitated a focus on a finite number of conditions and thus an inevitable exclusion of other important conditions.

In place of aggregate cost-of-disease figures, policymakers would be better served by data that provide insight into the factors driving those costs. Several parameters, including comorbidity and prevalence, available treatments, socioeconomic factors, and impact of insurance, appear to be examples of such domains in chronic conditions. Better understanding of these factors can provide a first step both in understanding and ameliorating the burden of chronic disease in the United States.

Editor's Notes

 
Benjamin Druss is an assistant professor in the Departments of Psychiatry and Public Health at the Yale School of Medicine in New Haven, Connecticut. Steven Marcus is a research assistant professor at the School of Social Work, University of Pennsylvania, in Philadelphia. Mark Olfson is an associate professor in the Department of Psychiatry, Columbia University, in New York City. Terri Tanielian is a research administrator for RAND Health in Washington, D.C. Lynn Elinson is an assistant professor in the Department of Psychiatry, University of Pittsburgh, and Harold Alan Pincus is executive vice-chairman of that department.

This work was funded in part by in part by a grant from the Robert Wood Johnson Foundation; National Institute of Mental Health (NIMH) Grant no. K08-MH01556-01A; and by the National Alliance for Research on Schizophrenia and Depression (NARSAD).

NOTES

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 

1. M.L. Berk and A.C. Monheit, "The Concentration of Health Care Expenditures, Revisited," Health Affairs (Mar/Apr 2001): 9–18; and C. Hoffman et al., "Persons with Chronic Conditions: Their Prevalence and Costs," Journal of the American Medical Association 276, no. 18 (1996): 1473–1479.[Abstract]
2. See, for example, the Robert Wood Johnson Foundation–funded Partnership for Solutions, <www.chronicnet.org>, and the Agency for Healthcare Research and Quality’s evidence-based practice centers. Evidence-Based Practice Centers: Overview, Pub. no. 00-P013 (Rockville, Md.: AHRQ, March 2000).
3. V. Benson and M.A. Marano, "Current Estimates from the National Health Interview Survey, 1995," Vital and Health Statistics 10, no. 199 (1998): 1–428; P. Fishman et al., "Chronic Care Costs in Managed Care,"Health Affairs (May/June 1997): 239–247; and C. Murray and A. Lopez, The Global Burden of Disease (Geneva: World Heath Organization, 1996), 247–324.
4. J. Cohen et al., "The Medical Expenditure Panel Survey: A National Health Information Resource," Inquiry 33, no. 4 (1996): 373–389.[Medline]
5. J. Cohen, Sample Design of the 1996 MEPS Household Component, MEPS Methodology Report no. 2, Pub. no. 97-0027 (Rockville, Md.: AHRQ, 1997).
6. All p values are for a Z-test comparing the condition with the next most expensive or prevalent condition. This provides a more conservative method than comparing a particular condition to the mean for all other conditions.
7. C. Clark et al., "Promoting Early Diagnosis and Treatment of Type-2 Diabetes: The National Diabetes Education Program," Journal of the American Medical Association 284, no. 3 (2000): 363–365.[Free Full Text]
8. E. Wagner, "The Role of Patient Care Teams in Chronic Disease Management," British Medical Journal 320, no. 7234 (2000): 569–572.[Free Full Text]
9. E. Wagner et al., "A Survey of Leading Chronic Disease Management Programs: Are They Consistent with the Literature?" Managed Care Quarterly 7, no. 3 (1999): 56–66.[Medline]
10. U.S. Preventive Services Task Force, Guide to Clinical Preventive Services, 2d ed. (Baltimore: Williams and Wilkins, 1996).
11. W.S. Weintraub et al., "Economics, Health-Related Quality of Life, and Cost-Effectiveness Methods for the TACTICS (Treat Angina with Aggrastat [tirofiban]; and Determine Cost of Therapy with Invasive or Conservative Strategy)–TIMI 18 Trial," American Journal of Cardiology 83, no. 3 (1999): 317–322.[Medline]
12. See, for example, J. Miller, "The Effects of Race/Ethnicity and Income on Early Childhood Asthma Prevalence and Health Care Use," American Journal of Public Health 90, no. 3 (2000): 428–430.[Abstract/Free Full Text]
13. M. George et al., "A Comprehensive Educational Program Improves Clinical Outcome Measures in Inner-City Patients with Asthma," Archives of Internal Medicine 159, no. 15 (1999): 1710–1716.[Abstract/Free Full Text]
14. B. Druss, R. Hoff, and R. Rosenheck, "Underuse of Antidepressants in Major Depression: Prevalence and Correlates in a National Sample of Young Adults," Journal of Clinical Psychiatry 61, no. 3 (2000): 234–237.
15. L. Baker, "Association of Managed Care Market Share and Health Expenditures for Fee-for-Service Medicare Patients," Journal of the American Medical Association 281, no. 5 (1999): 432–437[Abstract/Free Full Text]; and J. Robinson, "Blended Payment Methods in Physician Organizations under Managed Care," Journal of the American Medical Association 282, no. 13 (1999): 1258–1263.[Abstract/Free Full Text]
16. N. Krauss and B. Kass, "Comparison of Household and Medical Provider Reports of Medical Conditions" (Paper presented at the Joint Statistical Meetings, Indianapolis, Indiana, August 2000).

                      What's this?

This article has been cited by other articles:

				K. R. Merikangas, M. Ames, L. Cui, P. E. Stang, T. B. Ustun, M. Von Korff, and R. C. Kessler The Impact of Comorbidity of Mental and Physical Conditions on Role Disability in the US Adult Household Population Arch Gen Psychiatry, October 1, 2007; 64(10): 1180 - 1188. [Abstract] [Full Text] [PDF]

				P. S. Wang, G. E. Simon, J. Avorn, F. Azocar, E. J. Ludman, J. McCulloch, M. Z. Petukhova, and R. C. Kessler Telephone Screening, Outreach, and Care Management for Depressed Workers and Impact on Clinical and Work Productivity Outcomes: A Randomized Controlled Trial JAMA, September 26, 2007; 298(12): 1401 - 1411. [Abstract] [Full Text] [PDF]

				S. M. Shortell, T. G. Rundall, and J. Hsu Improving Patient Care by Linking Evidence-Based Medicine and Evidence-Based Management JAMA, August 8, 2007; 298(6): 673 - 676. [Full Text] [PDF]

				L. E. Ford and S. H. Gilbert The importance of maturational studies in airway smooth muscle Am J Physiol Lung Cell Mol Physiol, December 1, 2005; 289(6): L898 - L901. [Full Text] [PDF]

				J. H. Wasson Why Isn't It Better? Ann. Fam. Med, July 1, 2004; 2(4): 292 - 293. [Full Text] [PDF]

				R. Jagsi, T. F. DeLaney, K. Donelan, and N. J. Tarbell Real-Time Rationing of Scarce Resources: The Northeast Proton Therapy Center Experience J. Clin. Oncol., June 1, 2004; 22(11): 2246 - 2250. [Full Text] [PDF]

				E. A. Jacobs, D. S. Shepard, J. A. Suaya, and E.-L. Stone Overcoming Language Barriers in Health Care: Costs and Benefits of Interpreter Services Am J Public Health, May 1, 2004; 94(5): 866 - 869. [Abstract] [Full Text] [PDF]