This Article Extract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when Comments 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 Web of Science (25) Citing Articles via Google Scholar Google Scholar Articles by Zuvekas, S. H. Search for Related Content PubMed PubMed Citation Articles by Zuvekas, S. H. Related Collections Health Reform Mental Health/Substance Abuse

Health Tracking

TRENDS

Trends In Mental Health Services Use And Spending, 1987–1996

Trends in the overall financing of health care services have affected patterns of MH/SA service use and spending as well. Between 1987 and 1996 the number of uninsured persons increased, while the number covered by private insurance declined, although the number of uninsured has stabilized recently.² Expansions of Medicaid during the late 1980s and early 1990s has likely affected the use of MH/SA services, because Medicaid tends to have the most generous MH/SA coverage.

At the same time, the technology for treating mental and addictive disorders has improved with the introduction of new pharmacological treatments for depression, anxiety, and other disorders. Better available treatments and increased media exposure have led to increased demand for MH/SA services among the general population.³

The net of effect of these changes on patterns of use, spending, and access is difficult to predict, because these changes often move in opposite directions. To date, information about changing patterns of MH/SA services has generally been limited to aggregate expenditures and short-term case studies of selected groups.⁴ This study examines trends in inpatient and outpatient treatment, psychotropic drug use, access, and spending between 1987 and 1996. Particular attention is paid to trends by age (especially children and the elderly), sex, racial/ethnic subgroups, family income, and health insurance status.

Study Methods

 
This study uses the rich information on health care use and spending, and personal and household characteristics from the 1987 National Medical Expenditure Survey (NMES) and its successor, the 1996 Medical Expenditure Panel Survey (MEPS).⁵ These surveys elicited information from households for each health care event they reported in the 1987 and 1996 calendar years, respectively, including the main reasons for the visit, associated conditions, and expenditures and payment sources. Follow-back surveys of health care providers for a subsample of events (100 percent for hospital-based care, 50 percent for office-based physician care) yielded more complete information on spending and payment sources.⁶

Scope of sample and services NMES (n = 34,459) and MEPS (n = 21,571) are nationally representative surveys of the civilian, noninstitutionalized population. Hence, a significant fraction of total MH/SA use and spending are excluded, including all stays at long-term psychiatric facilities, residential treatment facilities, nursing homes, and other assisted-living facilities. Acute care MH/SA services for institutionalized populations, including those in jails and prisons, are also excluded. MH/SA services for homeless populations, which tend to have high rates of severe mental disorders and addictive disorders, are similarly excluded. Use of and spending for services received at community mental health centers and acute care services from short-term general and private psychiatric hospitals are included, as are services received at physician, psychologist, and other specialty mental health providers’ offices and clinics.

MH/SA treatment is defined to include services for the treatment of a disorder covered by International Classification of Diseases, Ninth Revision (ICD-9) codes 291, 292, and 295–314 (for example, major depression, bipolar disorder, anxiety disorders, schizophrenia, and substance use disorders). Alzheimer’s disease and other dementia, mental retardation, and developmental disabilities are excluded, as are disorders that result from MH/SA disorders, such as cirrhosis of the liver.

Outpatient treatment. Ambulatory MH/SA visits include all office-based, outpatient hospital, or emergency room visits where either (1) the main reason for the visit was psychotherapy or mental health treatment; (2) the provider was a psychologist, social worker, or counselor; or (3) an MH/SA condition was associated with the visit.⁷ Physician specialty was not ascertained in the 1996 MEPS, which precludes analyses of trends in the specialty and general medical sectors since psychiatrists cannot be distinguished from other physicians.

Inpatient treatment. MH/SA-related hospital stays were identified based on household-reported conditions associated with inpatient stays. The small number of respondents in each year with inpatient stays (which excludes stays in long-term care facilities) limits analyses to trends in aggregates.

Psychotropic drug use. Identification of psychotropic drug use and spending is complicated. Many drugs generally described as psychotropic (antidepressants, antianxiety, antipsychotics, central nervous system stimulants such as Ritalin, and antixiolytics/sedative hypnotics) have non-MH/SA uses, both on and off label. The strategy here is to include only those psychotropic drugs for which the household also reported an associated MH/SA condition. Although this may lead to some underestimation, it is consistent with the method for identifying ambulatory and inpatient use based on conditions and reasons for use.⁸

Expenditures. Expenditures are the sum of out-of-pocket and third-party payments for ambulatory, psychotropic drug, and inpatient MH/SA treatment. The MEPS expenditures are as reported in the public-use files. The 1987 NMES spending variables on the publicly released files generally reflect provider charges rather than payments. To create comparable spending data to the MEPS, adjustment factors based on information about discounts (difference between charges and payments) from the provider follow-back surveys were applied to the NMES spending estimates.⁹ These adjusted estimates were then put in constant 1996 dollars using the Consumer Price Index (CPI) for all items, a widely used price deflator for general inflation.

Analyses. The analyses compare distributions of use, expenditures, and payers for ambulatory, psychotropic drugs, and inpatient care in calendar years 1987 and 1996, both in aggregate and for various subgroups reflected in the surveyed populations.¹⁰ All differences between years reported in the text are significant at the .05 level unless otherwise noted. Standard errors were corrected for the complex survey designs.

Study Results

Top Study Methods Study Results Discussion NOTES

 
Aggregate use and spending. MH/SA use and spending rose sharply from 1987 to 1996. The total number of ambulatory visits increased 29.2 percent (p = .06), while the population increased only 12.3 percent over the same period (Exhibit 1). Total spending increased nearly 46 percent in constant-dollar terms. Per capita spending rose about 30 percent; increases in psychotropic drug use and spending account for about two-thirds of the increase in per capita total spending. In contrast, inpatient MH/SA use remained flat.

Average spending for ambulatory MH/SA users dropped slightly, but not statistically significantly, to $606 in 1996. Average costs per visit did not change significantly, either.¹¹ Managed care, through either negotiated discounts with providers, limits on the types of services, or changes in the mix of providers, may have limited increases in prices seen for medical services, in general.

Psychotropic drug use. Psychotropic drug use increased dramatically (63 percent) between 1987 and 1996, and average spending for psychotropic drug users more than doubled. The use of newer, more expensive psychotropic medications most likely explains the increases in cost per user. Interestingly, half of those with psychotropic drug use reported no other MH/SA use in 1987, and 39 percent did likewise in 1996, which suggests that many persons receive primarily psychotropic drug treatment.

Inpatient use. There was no change in the percentage with inpatient MH/SA stays between 1987 and 1996. However, the average length of an MH/SA stay declined. Spending per hospital day increased, but not significantly, averaging $761 in constant dollars in 1996.

Distribution of payment sources. The distribution of payment sources for all MH/SA services combined changed relatively little between 1987 and 1996, with the most significant increase in the amount paid for by Medicare (Exhibit 3). The distribution of payment sources for ambulatory MH/SA treatment was also relatively unchanged over the period. About one-third of spending for ambulatory MH/SA care was paid out of pocket in both years—twice the percentage paid out of pocket for all outpatient services. This reflects, in part, poor MH/SA coverage in most health insurance plans, but also decisions to either go out of the network or pay for MH/SA services entirely out of pocket.¹² Private insurance accounted for one-third of expenditures and Medicaid, about one-fifth, in 1996. Medicare accounted for very few outpatient services, reflecting the lower number and intensity of MH/SA users among the elderly. A sharp decline in the average cost per visit offset a relatively greater number of visits for the publicly insured under age sixty-five, contributing to the stability in the distribution across payers.

Comparisons with total health spending and use. Total spending for MH/SA increased by 37 percent in real terms between 1987 and 1996; it increased only 28 percent for all health care services in the study population. As a result, the share of total health spending accounted for by MH/SA services rose slightly from 4.1 percent in 1987 to 4.4 percent in 1996. This proportionately greater increase was driven by the near quadrupling of MH/SA-related prescription medications over this period. Overall, spending on all prescription medications increased 132 percent.

In contrast, outpatient spending for MH/SA treatment rose only 22 percent, compared with 95 percent for all outpatient services. While MH/SA-related visits increased by more than twice the amount of all outpatient visits (29 percent versus 14 percent), the average expense per MH/SA visit remained flat but increased in real terms by one-quarter for all visits. Inpatient spending did not increase significantly for either MH/SA or all inpatient services. However, average length-of-stay declined 34 percent for MH/SA treatment but remained unchanged for other types of care.

Age. Use of MH/SA services increased about 1.1–1.5 percentage points for all age groups between 1987 and 1996 (Exhibit 4). By 1996, 6.2 percent of children ages six to seventeen used ambulatory MH/SA services, compared with 7.3–7.4 percent of nonelderly adults and 5.2 percent of the elderly. The percentage paid out of pocket declined by half for children and by two-thirds for the elderly. Nonelderly adults experienced little change in out-of-pocket payments and were still paying more than one-third of total costs out of pocket for ambulatory treatment.

The elderly present an entirely different story. The elderly were much more likely than the nonelderly were to use psychotropic medications and much less likely to use ambulatory MH/SA treatment in both periods. However, there was no change in the percentage of the elderly using psychotropic drugs between 1987 and 1996 (Exhibit 5). In addition, average spending among elderly psychotropic drug users was much lower ($228) in 1996 than was true for other age groups, especially adults ages 18–44 ($411). The percentage paid out of pocket by the elderly for their psychotropic drugs remained unchanged in 1996, reflecting the elderly’s poor coverage for drugs. High out-of-pocket costs may have contributed to the lower growth in psychotropic drug use and spending among the elderly.

Sex. Females were more likely than males were to have ambulatory MH/SA treatment and receive psychotropic medications in both 1987 and 1996 (Exhibit 4). The gap in psychotropic medication use was even greater than for ambulatory treatment (Exhibit 5). There was little difference in 1996 between males and females in the intensity of ambulatory MH/SA treatment or psychotropic medications among users and no difference in the average percentage paid out of pocket for either. However, the out-of-pocket percentage dropped for female users of ambulatory treatment between 1987 and 1996, partially the result of increased Medicaid coverage among females receiving MH/SA services.

Race/ethnicity. Consistent with expectations, whites were much more likely than other racial/ethnic groups were to use MH/SA services in both 1987 and 1996 (Exhibit 4). The proportion of ambulatory MH/SA treatment users grew in approximately the same amounts for whites, blacks, and Hispanics in 1987–1996.Whiteswere also much more likely than other groups were to use psychotropic medications in 1996 and also had the fastest rate of growth in use between 1987 and 1996 (Exhibit 5).

The percentage paid out of pocket by whites dropped from 1987 to 1996 but tended to be greater than by other groups, especially blacks. By 1996 out-of-pocket payments for ambulatory MH/SA treatment were more than three times higher for whites than they were for blacks (Exhibit 4); payments for psychotropic medications were 1.7 times higher (Exhibit 5). The lower out-of-pocket percentages among blacks reflect, in part, the higher proportion of blacks on Medicaid.

Family income. The relationship between family income and MH/SA treatment use presents a complicated picture. In 1987 there were few significant differences between income groups in the proportion using ambulatory MH/SA services (Exhibit 4). However, between 1987 and 1996 lower income groups had much greater increases in the use of ambulatory services. For example, 5.5 percent of persons with family incomes below the poverty level used ambulatory services in 1987, rising to 8.0 percent in 1996. Among persons with the highest family incomes (greater than 400 percent of poverty), use went from only 5.6 percent in 1987 to 6.2 percent (not statistically different) in 1996. The greater increases among persons with lower incomes may reflect, in part, increased Medicaid coverage among these groups.

Increases in psychotropic drug use between 1987 and 1996 were uniform across income groups, in both the proportion with use and their average spending (Exhibit 5). In 1996, 6.4–6.5 percent of the poor and near-poor used psychotropic drugs, compared with 5.2–5.3 percent among higher-income groups.

While lower-income groups tended to pay a lower percentage of ambulatory MH/SA services out of pocket in 1996 as in 1987, the percentage paid out of pocket still remained high for these groups. Persons with family incomes below the poverty level paid 19.2 percent out of pocket in 1996 for ambulatory MH/SA care, compared with 12.0 percent for ambulatory care services in general. Lower-income groups paid the same large percentages out of pocket for psychotropic medications that higher-income groups did in 1996.

Insurance. Trends in ambulatory MH/SA differed greatly between persons under age sixty-five by health insurance status. Among those with any private insurance during the year (including Medigap), the number of persons with ambulatory use increased slightly, while the average number of visits dropped slightly (Exhibit 4). The decrease in visits was driven mainly by the reduced number of persons with twenty or more visits. The increasing percentage of privately insured persons using outpatient services was offset by this decline in intensity of use, so that ambulatory visits in aggregate did not change significantly between 1987 and 1996.

The percentage of persons under age sixty-five with public insurance using ambulatory services grew from 7.8 percent in 1987 to 10.1 percent in 1996. The increase in the number of visits per user was not statistically significant. However, those with public insurance accounted for a much larger percentage of all MH/SA outpatient visits in 1996 (35 percent) than in 1987 (16 percent). In contrast, the percentage of the uninsured using ambulatory services did not change between 1987 and 1996 and was significantly lower than for people with either public or private insurance in 1996.

The publicly and privately insured experienced similar growth in the use of psychotropic medications (Exhibit 5). However, those with public insurance were significantly more likely to use psychotropic drugs in both periods. The uninsured were much less likely than either insurance group to use psychotropic medications.

The higher rates of both ambulatory MH/SA service and psychotropic drug use may be explained by the greater generosity of coverage for MH/SA services in public plans, especially Medicaid, compared with private plans. However, a number of MH/SA users with Medicaid or Medicare qualified for that coverage because of their mental disorders.

The percentage paid out of pocket changed little for ambulatory MH/SA services but dropped for psychotropic medications between 1987 and 1996 for all types of insurance coverage for persons under age sixty-five. Persons with public-only insurance paid much less in 1996 out of pocket (16.6 percent) than did those with any private insurance (40.1 percent) (Exhibit 4). However, some of the out-of-pocket expenses for the privately insured may have resulted from their choosing out-of-network providers or paying out of pocket entirely to avoid reporting MH/SA treatment to their insurers and others. The differences in percentage paid out of pocket between public only and any private insurance for psychotropic medications were much smaller and not significant in 1996 (Exhibit 5). The uninsured faced out-of-pocket costs of nearly 60 percent for ambulatory MH/SA services and 80 percent for psychotropic medications in 1996. Other sources of payment for the uninsured include the Veterans Affairs (VA) system and other federal, state, and local funding.

Discussion

Top Study Methods Study Results Discussion NOTES

 
Limitations. There are two important limitations with the data used in this analysis. First, the scope of NMES and MEPS is limited to the civilian, noninstitutionalized population and therefore excludes a number of policy-relevant populations. Second, health care and MH/SA service use is based on household-reported data. Household surveys may lead not only to underreporting of all health care events, because of imperfect recall, but also underreporting of MH/SA services because of the associated stigma. A further complication arises if stigma has decreased over time.

Tracking polls have shown greater awareness and acceptance in the general population for persons with mental disorders, suggesting that stigma surrounding these disorders has decreased.¹⁴ While reduced stigma has likely increased demand for MH/SA treatment, a secondary effect may have been to increase people’s willingness to report MH/SA treatment in household surveys over time. To test this, comparisons were made between household-reported conditions associated with health care events and those reported by medical providers for those same events in both the 1987 NMES and 1996 MEPS. In general, there was very good agreement between households and providers on mental health conditions in both 1987 and 1996 and little evidence of systematic differences between the two years. This suggests that when households report health care events, they are fairly good at identifying MH/SA-related events. However, it is possible that households (and providers) may omit reporting some such events altogether and also that this likelihood decreased between 1987 and 1996.

Comparison to other estimates. The spending figures derived here are much lower than those reported by Tami Mark and colleagues, which are based on a National Health Accounts (NHA) model.¹⁵ The lower estimates are primarily attributable to differences in scope of included populations and spending between NMES and MEPS and the NHA based estimates. These differences lead to the exclusion not only of spending in institutions but the acute care MH/SA spending of those residing in those institutions.¹⁶ A careful analysis by researchers at the Agency for Healthcare Research and Quality (AHRQ) and the Health Care Financing Administration (HCFA) Office of the Actuary (which produces the NHA estimates) concludes that differences in scope explain almost all of the differences between the $554 billion estimate from MEPS and the $1.04 trillion NHA estimate of total U.S. health care spending in 1996, and that there is a remarkable degree of concordance between the two sets of estimates.¹⁷ However, household underreporting, as previously discussed, may still account for some differences between the estimates, especially for MH/SA services. A major strength of NHA and NHA-model aggregate spending estimates is that they provide a more complete picture of total spending, which is particularly important for MH/SA spending, while NMES and MEPS allow more comprehensive microlevel analyses.

A full, detailed comparison of the estimates presented here and the estimates by Mark and colleagues is beyond the scope of this paper, but it is important to note that the two sets of estimates imply similar rates of growth in total MH/SA spending between 1987 and 1996. The estimates presented here suggest lower increases in ambulatory MH/SA treatment than the NHA-based estimates do but much greater increases in psychotropic drug spending. However, when all drugs within psychotropic therapeutic classes are included in the NMES and MEPS analyses, regardless of what they were prescribed for, almost identical growth rates are obtained.¹⁸

Finally, Mark and colleagues find some-what slower increases in MH/SA spending than for all U.S. health care spending, while the results here suggest just the opposite for the community population. Again, the difference likely results from differences in scope.

Implications. A central finding is that access to ambulatory MH/SA services and psychotropic medications in the study population greatly increased between 1987 and 1996. However, despite these apparent gains in access, the recent surgeon general’s report on mental health makes clear that there is still much unmet need.¹⁹ Moreover, our analyses suggest wide variations in access and use among population subgroups, especially racial/ethnic groups.

Although it is clear that there have been substantial increases in the number of people using ambulatory MH/SA and psychotropic drug treatment, we can say little about the appropriateness of this increased use. The explosive growth in prescriptions for Ritalin and Prozac, for example, has raised concerns about the overdiagnosis of ADHD and major depression, respectively, and overuse of these medications for some patients. The dramatic rise in psychotropic drug spending is in direct contrast to ambulatory and inpatient MH/SA spending—which on a per capita basis changed little between 1987 and 1996—and likely led to increased pressure to contain costs. At the same time, the surgeon general’s report suggests that ADHD and depression are often underdiagnosed and undertreated. Thus, while there have been gains in the use of treatment, it is not clear that they have been distributed optimally, on the basis of medical or other societal judgments.

Finally, the community-dwelling population, including the well-insured, continued to face high out-of-pocket expenses for ambulatory MH/SA and psychotropic drug treatment in 1996. The financial impact for those receiving intensive MH/SA can be extremely large.²⁰ The well-known results from the RAND Health Insurance Experiment suggest that use of mental health services is quite sensitive to the level of cost sharing, so that high out-of-pocket costs are likely to deter many from seeking treatment for mental and addictive disorders.²¹

Editor's Notes

 
Sam Zuvekas is a senior economist at the Center for Cost and Financing Studies, Agency for Healthcare Research and Quality (AHRQ), in Rockville, Maryland.

Steve Hill, Marc Freiman, Joel Cohen, Alan Monheit, and two anonymous referees provided many helpful comments and suggestions. The views expressed in this paper are those of the author, and no official endorsement by the Agency for Healthcare Research and Quality or the Department of Health and Human Services is intended or should be inferred.

NOTES

Top Study Methods Study Results Discussion NOTES

 

1. U.S. Department of Health and Human Services, Mental Health: A Report of the Surgeon General (Rockville, Md.: DHHS, 1999).
2. J. Rhoades and M. Chu, Health Insurance Status of the Civilian, Noninstitutionalized Population, 1999, MEPS Research Findings (Rockville, Md.: Agency for Healthcare Research and Quality, forthcoming).
3. S.M. Foote and L. Etheredge, "Increasing Use of New Prescription Drugs: A Case Study," Health Affairs (July/Aug): 165–170; and H.A. Pincus et al., "Prescribing Trends in Psychotropic Medications: Primary Care, Psychiatry, and Other Medical Specialties," Journal of the American Medical Association 279, no. 7 (1998): 526–531.
4. T.L. Mark et al., "Spending on Mental Health and Substance Abuse Treatment, 1987–1997," Health Affairs (July/Aug 2000): 108–120.
5. W.S. Edwards and M. Berlin, Questionnaires and Data Collection Methods for the Household Survey and the Survey of American Indians and Alaska Natives, DHHS Pub. no. (PHS)89-3450, NMES Methods 2 (Rockville, Md.: National Center for Health Services Research and Health Care Technology Assessment, 1989); and J.W. Cohen et al., "The Medical Expenditure Panel Survey: A National Health Information Resource," Inquiry 33, no. 4 (1996): 373–389.
6. Provider-reported diagnoses from follow-back surveys were not used to identify MH/SA treatment. They were not available for many events due to the selected sampling, various forms of nonresponse, and matching errors.
7. Professional coders classified household-reported conditions into standard ICD-9 codes. In addition to ICD-9 codes 291, 292, and 295–314, other codes indicating MH/SA treatment are included, specifically, V11, V15.4, V40, V61, V62, V66.3, V67.3, V70.1, V70.2, V710, V79, and 799.2 ("nervousness").
8. Previous NMES analyses included some or all additional uses of psychotropic drugs. However, the pattern of household reports of associated conditions follows a logical pattern. Lithiumand Clozaril are almost always reported as MH/SA-related, Prozac and Zoloft are mostly reported as MH/SA-related, while a much smaller proportion of sedatives/hypnotics, which have a large number of other uses, are reported as MH/SA-related. See, for example, M.P. Freiman, P. Cunningham, and L. Cornelius, Use and Expenditures for the Treatment of Mental Health Problems, AHCPR Pub. no. 94-0085, NMES Research Findings 22 (Rockville, Md.: Agency for Health Care Policy and Research, 1994); and M. Olfson and H.A. Pincus, "Measuring Outpatient Mental Health Care in the United States," Health Affairs (Winter 1994): 172–180.
9. S.H. Zuvekas and J.W. Cohen, "A Guide to Comparing Expenditures in the NMES and MEPS" (Rockville, Md.: Center for Cost and Financing Studies, AHRQ, 2000).
10. The population was divided into the following age groups: 0–5, 18–44, 45–64, and 65 and older. Results for children ages 0–5 are omitted because of low MH/SA prevalence and treatment use. Family income was measured in relationship to the poverty line, with poor (family income under the poverty level), near-poor (100–125 percent of the poverty level), low income (125–200 percent of poverty), middle income (200–400 percent of poverty), and high income (above 400 percent of poverty) categories. Health insurance status for those under age sixty-five is represented by three mutually exclusive categories: those privately insured at any point during the year, including military coverage for civilian dependents and those with both Medicare and private coverage (Medigap); those with public insurance coverage but without private insurance during the year; and those uninsured all year. It is important to note that some uninsured may receive services through the Veterans Affairs (VA) health care system (the VA is not considered insurance in NMES and MEPS), the Indian Health Service, community mental health centers, and other public sources, as well as through private philanthropy.
11. Changes in the mix of ambulatory MH/SA services or providers may lead to changes in prices. However, the household-reported data lack sufficient detail to control for this mix.
12. J.W. Cohen et al., Health Care Expenses in the United States, 1996, MEPS Research Findings (Rockville, Md.: AHRQ, forthcoming).
13. Ibid.
14. J. Phelan et al., "Public Conceptions of Mental Illness in 1950 and 1996: What Is Mental Illness and Is It to Be Feared?" Journal of Health and Social Behavior 41, no. 2 (2000): 188–207.
15. Mark et al., "Spending on Mental Health and Substance Abuse Treatment."
16. Other differences include the exclusion of insurance and administration costs, most public health activities, most philanthropy, and psychotropic drugs prescribed for non-MH/SA purposes.
17. T.M. Selden et al., "Reconciling Medical Expenditure Estimates from the Medical Expenditure Panel Survey and the National Health Accounts, 1996" (Unpublished mimeo, July 2000).
18. A much larger percentage of drugs within psychotropic therapeutic classes were identified as being prescribed for non-MH/SA reasons in 1987 than in 1996, where newer psychotropic medications predominate.
19. DHHS, Mental Health.
20. S.H. Zuvekas, J.S. Banthin, and T.M. Selden, "Mental Health Parity: What Are the Gaps in Coverage?" Journal of Mental Health Policy and Economics 1, no. 3 (1998): 135–146.
21. E.B. Keeler, W.G. Manning, and K.B. Wells, "The Demand for Episodes of Mental Health Services," Journal of Health Economics 7, no. 4 (1988): 369–392.

                      What's this?

This article has been cited by other articles:

				S. H. Zuvekas and C. D. Meyerhoefer State Variations In The Out-Of-Pocket Spending Burden For Outpatient Mental Health Treatment Health Aff., May 1, 2009; 28(3): 713 - 722. [Abstract] [Full Text] [PDF]

				R. G. Frank and S. Glied Changes in mental health financing since 1971: implications for policymakers and patients. Health Aff., May 1, 2006; 25(3): 601 - 613. [Abstract] [Full Text] [PDF]

				B. G. Druss Rising mental health costs: what are we getting for our money? Health Aff., May 1, 2006; 25(3): 614 - 622. [Abstract] [Full Text] [PDF]

				S. H. Zuvekas Prescription Drugs And The Changing Patterns Of Treatment For Mental Disorders, 1996-2001 Health Aff., January 1, 2005; 24(1): 195 - 205. [Abstract] [Full Text] [PDF]

				D. J. Safer, J. M. Zito, and S. dosReis Concomitant Psychotropic Medication for Youths Am J Psychiatry, March 1, 2003; 160(3): 438 - 449. [Abstract] [Full Text] [PDF]

				J. M. Zito, D. J. Safer, S. dosReis, J. F. Gardner, L. Magder, K. Soeken, M. Boles, F. Lynch, and M. A. Riddle Psychotropic Practice Patterns for Youth: A 10-Year Perspective Arch Pediatr Adolesc Med, January 1, 2003; 157(1): 17 - 25. [Abstract] [Full Text] [PDF]

				R. G. Frank, H. H. Goldman, and T. G. McGuire Will Parity in Coverage Result in Better Mental Health Care? N. Engl. J. Med., December 6, 2001; 345(23): 1701 - 1704. [Full Text] [PDF]