Administration and Policy in Mental Health, Vol. 32, No. 4, March 2005 ( 2005) DOI: 10.

1007/s10488-004-1670-3

RACE, MANAGED CARE, AND THE QUALITY OF SUBSTANCE ABUSE TREATMENT

Marilyn C. Daley

ABSTRACT: The adoption of managed behavioral health care by state Medicaid agencies has the potential to increase the quality of treatment for racial minorities by promoting access to substance abuse treatment and creating more appropriate utilization patterns. This paper examines three indicators of quality for white, Black, and Hispanic Medicaid clients who received substance abuse treatment in Massachusetts between 1992 and 1996. It evaluates whether a managed behavioral health care carve-out in FY1993 had a positive or negative effect on access, continuity of care, and 30-day re-admissions. Prior to managed care, access and continuity were worse for minorities than for whites. For all clients under managed care, access and continuity improved between 1992 and 1996. Access improved more for Hispanic clients relative to other racial groups. Continuity improved more for Black clients relative to other racial groups. Although seven-day and 30-day re admissions also increased following managed care, the rate of increase was not signicantly greater for minorities. Although managed care had a benecial impact on the quality of treatment for minority clients, the percent of minority Medicaid-eligible clients who accessed treatment and the percent who achieved continuity of care remained lower than for whites in every year of the study. Managed care reduced, but did not overcome, racial disparities in behavioral health care. KEY WORDS: managed care; race; substance abuse treatment.

The adoption of managed behavioral health care (MBHC) plans by state Medicaid agencies had the potential to improve mental health outcomes for racial minority groups by promoting better access to treatment and creating more appropriate utilization patterns. However, little
Marilyn C. Daley, Ph.D., is a Senior Research Associate at the Schneider Institute for Health Policy in the Heller School for Social Policy and Management at Brandeis University, Waltham, MA. The author would like to acknowledge the important contributions of Donald Shepard, Ph.D., Professor at the Heller School for Social Policy and Management, who was the Principal Investigator on the project, and Yvonne Anthony, Ph.D., who reviewed earlier versions of the paper. Funding for this paper came from the National Institute on Alcohol Abuse and Alcoholism, grant #R01-AA-10880, Managed Care in Michigan and Massachusetts, to Professor Donald Shepard, Ph.D. Address for correspondence: Marilyn C. Daley, Ph.D., Senior Research Associate, Schneider Institute for Health Policy, Heller School for Social Policy and Management, Brandeis University, MS 035, P.O. Box 9110, 415 South Street, Waltham, MA 02454-9110. E-mail: daley@brandeis.edu.

457

2005 Springer Science+Business Media, Inc.

458

Administration and Policy in Mental Health

research has separately evaluated the impact of MBHC on the quality of substance abuse treatment for Blacks and Hispanics. The quality of treatment is a critical issue, since substance abuse increases the risk for accidents, overdose, injuries, diabetes, crime, homicide, low birth weight, hypertension, Hepatitis C, and HIV/AIDS, problems that contribute to a reduced life expectancy for Blacks and Hispanics (Collins et al., 2002; Fiscella & Franks, 1997; Lillie-Blanton & LaVeist, 1996; National Center for Health Statistics, 2003; U.S. Census Bureau, 2000). This report describes how the rst four years (FY1992FY1996) of an MBHC carveout affected three aspects of the quality of substance abuse treatment servicesaccess, continuity, and rapid re admissionsfor Black, white, and Hispanic Medicaid recipients in Massachusetts. Since the Presidents New Freedom Commission on Mental Health (SAMHSA, 2003) has identied the elimination of disparities in mental health care as one of its six main goals, the ndings of this report should be particularly relevant for policy and practice.
Access and Use of Behavioral Health Services for Blacks and Hispanics

Historically, Blacks and Hispanics have reported much poorer access to mental health and substance abuse treatment and use fewer services than their non-minority counterparts. For example, ndings from a large community epidemiological study (Vega, Kolody, Aguilar-Gaxiolar, & Catalano, 1999) suggest that only 11% of Hispanics with mood disorders and 9% with anxiety disorders had received treatment. Although lack of health insurance, poverty, and unemployment certainly contribute to poorer access, it is clearly not the whole picture. Although the 2000 Census reported that over a third of Hispanics did not have health insurance, compared to 20% of African-Americans and 10% of whites (U.S. Census Bureau, 2000), a recent study by the Commonwealth Fund found that even among insured populations, minorities were less likely to receive mental health treatment (Collins et al., 2002). Minorities are less likely to have a regular health care provider, and are more likely to receive treatment for psychiatric disorders through the emergency room or from primary care physicians, rather than seek specialty mental health treatment. They are more likely to be misdiagnosed, less likely to be treated in outpatient settings, less likely to receive psychiatric medications, and more likely to have expensive inpatient hospitalizations. These ndings are unfortunate, since they suggest that the quality of care minorities receive is delayed, fragmented, and inadequate (Collins et al., 2002; Centers for Disease Control, 2004; Institute of Medicine, 2001; Regier et al., 1993; Vega et al., 1999).

Marilyn C. Daley

459

Poor communication skills and cultural bias may also be part of the reason why access to treatment is poor and misdiagnosis is high. Since cultural factors may cause minorities to express symptoms somewhat differently, depression and anxiety disorders tend to be under-diagnosed for African-Americans, while schizophrenia is over-diagnosed, a practice that may result in lengthy and unnecessary hospitalizations in state facilities (Regier et al., 1993). Despite this, less than 2% of psychiatrists and 4% of social workers are AfricanAmerican (Alegria et al., 2002; SAMHSA, 1998; Weineck, Jacobs, Stome, Ortega, & Burstin, 2004).

Minorities are less likely to have a regular health care provider, and are more likely to receive treatment for psychiatric disorders through the emergency room.
Since the 2000 Census reported that 40% of Hispanic Americans did not speak English very well, language is another factor that may present a signicant impediment to access (Lopez, 2002). Language problems are exacerbated in the area of mental health treatment where the ability to communicate is so essential. Since less than 1% of practicing psychologists are Hispanic, Hispanics who do seek treatment are unlikely to nd a Spanish-speaking therapist, resulting in miscommunication, misdiagnosis, and inappropriate treatment (Malgady, Roger, & Constantino, 1987; Mukherjee, Shukla, Woodle, Rosen, & Olarte, 1983).
Outcomes of Behavioral Health Care for AfricanAmericans and Hispanics

Once minority groups access the right kind of treatment, their outcomes appear to be similar to those of non-minorities. For example, two large federally funded, nationally representative studies found no differences in overall treatment outcomes for whites, AfricanAmericans, and Hispanics when using multivariate analysis. After controlling for age, education, severity of drug use at intake, and other demographic factors, the Services Research Outcome Study (SAMHSA, 1998) found no differences by race in Addiction Severity Index scores for the three ethnic groups as self-reported one year post-treatment. Outcomes for Blacks and Hispanics varied, however, by the type of treatment they received and the type of outcome assessed. For example, in terms of treatment facility, whites did better in inpatient and methadone programs, Blacks did better in residential, and Hispanics did better in outpatient. The National Treatment Improvement Evaluation (NTIES) also failed to nd differences in overall treatment outcomes by minority group status (SAMHSA, 1996).

460

Administration and Policy in Mental Health

Howard, LaViest, and McCaughrin (1996) looked at outpatient treatment programs and did not nd a difference in retention rates by race once socio-environmental and program organizational factors were taken into account. Another investigation used Addiction Severity Index self-reports and urinalysis to compare six-month treatment outcomes for Blacks, whites, and Hispanics, and found almost identical rates of relapse for minorities and non-minorities (Mulvaney et al., 1999). On the other hand, Moos, Moos, and Finney (2001) reported that African-Americans were more likely to experience deterioration in a subset of clients whose ASI scores got worse (increased) in the year following discharge from inpatient treatment. However, all of these clients received treatment in Veterans Administration hospitals, which served a population of older, disabled, low-income veterans, and ndings may not generalize to the larger population of minorities who received substance abuse treatment in public and private facilities. In recent decades, three policy changes improved the quality of substance abuse treatment for racial minority groups: the emergence of Medicaid and Medicare in the 1960s, the development of new treatment programs that are linguistically appropriate and sensitive to subtle differences in minority culture, and the growing use of public sector managed behavioral health care programs to address rising behavioral health care spending (Leigh, Lillie-Blanton, Martinez, & Collins, 1999). MBHCs have the potential to improve access, promote continuity, and enhance outcomes for minority groups. They could create more appropriate utilization patterns by ensuring that minorities get more preventive care, make better use of outpatient counseling and day treatment, are matched with minority therapists, and have fewer lengthy inpatient hospitalizations. They could certainly ensure better coordination between primary care physicians (PCPs) and substance abuse treatment providers, since most minorities access treatment through their PCP. They could ensure that minorities receive appropriate diagnoses and receive pharmacotherapy if warranted. On the other hand, the major impetus for the spread of Medicaid managed behavioral health care has been cost containment, as opposed to quality improvement. There are several cost containment strategies that could work to the detriment of minorities who wish to enter substance abuse treatment. For example, MBHCs, in an effort to contain costs, could deny treatment to minority clients if they are perceived as high resource utilizers. Since minorities make less use of low-intensity and preventive services, they may only reach specialty substance abuse treatment when their drug abuse problems have deteriorated to a crisis state. Further, many managed care programs will attempt to contain costs by selectively contracting with a network of providers who agree to provide services at a reduced rate. While

Marilyn C. Daley

461

this strategy may decrease costs and improve quality for the typical client, it might mean that small providers in minority communities might not be able to compete, and minority clients will have to sever relationships with the community health providers they currently use.

[Minorities] are more likely to be misdiagnosed. . .less likely to receive psychiatric medications, and more likely to have expensive inpatient hospitalizations.
Despite the importance of this controversy, there have been very few evaluations that have examined how managed care has affected access, utilization, or outcomes separately for racial minorities. Leigh et al. (1999) compared multiple measures of access and satisfaction for low-income whites, Hispanics, and AfricanAmericans in managed care and fee-for-service plans in Texas, Florida, and Tennessee. Although access improved for the typical white client under managed care, access was the same or worse for Blacks and Hispanics. In another recent study (Cunningham & Trude, 2001; Hargreaves, Cunningham, & Hughes, 2001), ve commonly used indicators of access, including having a regular provider, number of emergency room visits, and visiting a specialist, were compared for a nationally representative sample of 4,811 AfricanAmericans, 3,379 Hispanics, and over 30,000 white clients in managed care and fee-for-service plans. For both minorities and nonminorities, access was better in managed care relative to fee-for-service plans. However, even under managed care plans, access for whites was better than access for minorities. Therefore, the gap between whites and minorities persisted even in the aftermath of managed care. In another evaluation that examined racial differences in access and satisfaction for Medicaid enrollees in North Carolina, Greenberg, Brandon, Schoeps, Tingle, and Shull (2003) compared a county that had implemented a mandatory managed care plan with an adjacent county that had not. While several measures of access improved for the racial minorities who were enrolled in managed care, including having a regular source of health care and having made a visit to a doctor, these measures also improved for the racial minorities in the control county who were not under managed care. The authors speculate that managed care had no independent impact on access or satisfaction (Greenberg et al., 2003). On July 1, 1992, Massachusetts Department of Medical Assistance (DMA) became the rst state Medicaid agency to enter into a contract with a risk-sharing managed behavioral health care carve-out, MHMA. An independent evaluation of the rst year of MHMAs performance

462

Administration and Policy in Mental Health

found that spending declined by 76%, access improved signicantly, and 30-day readmissions declined by a non-signicant amount (Callahan et al., 1995). While these ndings were encouraging, quality measures were not examined separately by ethnic group. Hence, the current study adds to the literature by comparing changes in access, rapid readmissions and continuity of care for whites, Blacks, and Hispanics prior to (FY1992) and following (FY1993FY1996) the implementation of the Massachusetts carve-out.

METHODS
Data Sources

All Medicaid claims for substance abuse treatment received by MHMA enrollees less than 65 years of age from FY1992 through FY1996 (July 1, 1991 through June 30, 1996) were obtained from what is now the Department of Transitional Assistance for Needy Families, Division of Medical Assistance (DMA). Substance abuse claims were dened as any invoice of type 1, 3, 5, or 9 (hospitalizations, clinic visits, physicians services, or miscellaneous services, respectively) and met at least one of the following criteria: (1) a primary diagnosis of drug or alcohol abuse or dependence; (2) a primary diagnosis of mental health disorder and a second, third, fourth or fth diagnosis of drug or alcohol abuse or dependence; (3) identication by DMA as a specialty substance abuse service based on invoice type, procedure code, provider type, or provider specialty code. Claims for pharmaceuticals, transportation, and dental visits were not included. Using the unique identier used by Medicaid (Recipient Historical Number or RHN), each client was matched with his enrollment record to obtain demographic information and ensure that the client was eligible for Medicaid and participating in MHMA at the time of the claim.

Even under managed care plans, access for whites was better than access for minorities.
Based on an algorithm used in the prior Brandeis MHMA evaluation (Shepard et al., 2001), all claims were categorized into service types based on invoice type, primary diagnosis, procedure code, provider type, and provider specialty, and collapsed into four modalities:(1) acute inpatient treatment for alcohol or other drug (AOD) withdrawal (ASAM Level 4), (2) acute residential services (ASAM Level 3), (3) day treatment/regular outpatient (ASAM Levels 1 and 2), and (4) methadone counseling. Claims for methadone dosing were excluded because preliminary analyses indicated that virtually all of these clients had also received counseling.

Marilyn C. Daley

463

Outcome Measures

Access. The penetration rate is the total number of unique Black, white, or Hispanic clients admitted to any 24-hour service in each scal year per 1,000 managed care enrollees within that particular ethnic category. The penetration rate thus controls for changes in the enrolled population. The average number of Black, white, and Hispanic managed care enrollees was determined by taking the total person-days of eligibility for each of those ethnic groups during each scal year, and dividing by 365 (or 366 in leap years). This allowed us to measure access relative to the changing number of enrolled Medicaid recipients in each ethnic category. Continuity. Continuity is a desirable process measure that has been associated with favorable outcomes in numerous evaluations and cost-effectiveness analyses. Many clinicians feel that the clients chances for stable recovery are enhanced if s/he is moved along a continuum of care that decreases intensity in preparation for living independently. This continuum would begin with detoxication and proceed to acute residential, halfway houses, day treatment, and regular outpatient. Plans should encourage the providers in their network to cooperate with each other to maximize the number of clients who are able to complete referrals to the next level of care. Rapid Re admissions. Re admissions to acute treatment within a relatively short period of time (seven or 30 days) are considered an undesirable outcome because they suggest the client has returned to drug use, perhaps due to inadequate or inappropriate treatment in the original facility. Spending. Spending is the total amount paid by Medicaid for each episode of care, which should not be confused with the amount billed or the true cost of the resources consumed. Since the cost analysis was conducted from Medicaids perspective, we felt the amount paid was the most relevant measure. All Medicaid expenditures were adjusted to the midpoint of FY 1996 dollars using the Consumer Price IndexAll Urban Consumers, Boston area, Medical component (Bureau of Labor Statistics, 2002).
Data Analysis

Claims for the ve years were merged, and a chronological le was created for each client, sorting by admission date and modality. Because multiple claims are frequently submitted within a single inpatient stay, claims were collapsed into episodes. For inpatient and residential

464

Administration and Policy in Mental Health

treatment, an episode was dened as any sequential claims of the same modality with zero or one day between the discharge date of the rst claim and the admission date of the second claim. A lapse of two days or longer was considered a re admission and not a continuation of the same episode. For outpatient and methadone claims, a new episode was initiated whenever there was an interval of 45 or more days between successive visits. Within each episode, the total days or sessions attended and the Medicaid payments were summed, resulting in a le of 55,304 distinct episodes with admission and discharge dates. The number of days between episodes was then computed. For inpatient or residential treatment, an admission to a lower level of care (outpatient, methadone, or lower level of residential) within 14 days of discharge from a rst inpatient or residential episode was assigned a 1 for continuity of care, a favorable outcome. If the client was admitted to the same or a higher level of care within two to 30 days, it was assigned a 1 for a rapid re admission, an unfavorable outcome. Since the calculation looked forward up to six episodes to identify re admissions within 30 days, it was possible for a client to have both continuity and rapid re admission for the same episode, if he was discharged to a lower level of care and subsequently re admitted to a higher level of care within 30 days of the rst discharge. If a client received outpatient or methadone services concurrently with a residential or inpatient treatment episode, continuity was achieved only if the outpatient or methadone visits extended beyond the discharge date for the residential or inpatient episode. Preliminary frequency distributions and summary statistics were conducted on all variables by year of service to assess the integrity and face-validity of this episode-based data set. All episodes were then classied by year of admission, and whether the episode was initiated prior to or following the introduction of managed care on July 1, 1993 (POSTMC = 1). Clients whose rst admission occurred after May 31, 1996 were excluded because they would not have had a full month to monitor services with the available data.

Although some measures of quality improved for minority clients, penetration rates remained consistently lower for Hispanics and Blacks than they were for whites in every year of the study.
All client characteristics available from the claims or enrollment les were selected as control variables, including age, race, gender, eligibility status (disabled versus non-disabled), and primary diagnosis according to ICD-9 code (cocaine, heroin, alcohol, other drug, or mental health). A mental health diagnosis indicated comorbidity, since all clients were

Marilyn C. Daley

465

substance abusers. Each of ve age groups, race, and drug of choice were expressed as categorical variables to permit the calculation of odds ratios. Although income and education were not available, all clients should have been below 185% of the federal poverty level as a pre-requisite for Medicaid eligibility. An important programmatic variable, the length of stay of the current treatment episode was considered an intermediate outcome as it reected not only the complexity of the admission, but also MHMA approval policies. To estimate the impact of these programmatic changes on quality, separate logistic regressions were estimated including and excluding program variables. All regressions were estimated with the logistic regression procedure (PROC LOGISTIC) of the SAS program, Version 8.
Statistical Analysis

Logistic regression was conducted to assess the likelihood of achieving continuity of care and being re-admitted to acute treatment services, after adjusting for changes in demographic characteristics and historical trends that may have independently affected these two outcomes over time. Odds ratios from the logistic regressions could be multiplied by the mean rate of 30-day re admission or continuity to assess the magnitude of each explanatory variables contribution, after other changes were held constant. Interaction effects were used to test whether managed care had a greater or lesser impact on Blacks and Hispanics relative to whites.

RESULTS
Demographics

There were major differences in the type of treatment received, the primary substance of abuse, and other lifestyle characteristics when specic minority groups were examined (see Table 1). Although only a small number of variables were available from the Medicaid enrollment les, we found radical differences between the ethnic categories in age, gender, primary diagnosis, and eligibility category. The Hispanic clients were more likely to be females who were receiving TANF, less likely to be SSI-disabled, and 45 years younger than the other ethnic groups. In terms of primary diagnosis, Blacks and whites received more treatment for alcohol disorders, Hispanics reported more heroin problems, and AfricanAmericans were twice as likely to be diagnosed with a cocaine problem. These ndings have been reported in other investigations that used the BSAS MIS system to examine ethnic groups (Argeriou & Daley, 1997; McCarty, Caspi, Panas, Krakow, & Mulligan, 2000).

466

Administration and Policy in Mental Health

TABLE 1 Demographic Characteristics for Whites, Blacks, and Hispanics (Percent of Episodes) White Female Age (years) Age Under 21 2130 3140 4150 Over 50 Drug of Choice Heroin Cocaine Alcohol Mental health Other Disabled No. of episodes 49.6% 34.9 9.3 21.6 39.6 20.6 9.0 23.2 6.8 44.0 13.0 13.0 65.8 43,014 AfricanAmerican Hispanic 56.7% 33.8 6.8 25.0 46.1 17.7 4.4 15.6 26.6 37.6 9.1 11.2 60.3 8,481 66.5% 28.8 20.2 32.8 36.9 8.1 2.1 42.8 13.7 20.0 15.2 8.2 24.6 2,902 Other Total 53.9% 51.7% 28.3 33.1 19.2 28.0 15.4 4.3 14.3 9.5 33.6 34.1 8.5 39.5 907 9.9 22.6 4.3 19.4 7.8 22.9 10.3 41.6 12.8 12.4 62.4 55,304

Types of Services Utilized

When we compared the types of 24-hour care utilized by Blacks, Hispanics, and whites, there were some signicant differences between the groups (see Table 2). Please note that the category other race has too few cases to be reliable. While a higher percentage of Blacks were admitted more frequently to substance abuse residential and freestanding detoxication facilities, whites were slightly more likely to enter substance abuse hospitals and mental health residential programs. Hispanics, who were younger, used more services targeted to clients under 21 years of age.
Access

The rate of access (Figure 1) for Blacks and Hispanics was consistently lower than for white clients, and this persisted both before and after managed care. However, following the introduction of MHMA in 1993, access increased dramatically for all three ethnic groups. These improvements were particularly pronounced for Hispanics, who were

TABLE 2 Percentage of Treatment Episodes in Various Services, by Race (with N) White (43,014) AfricanAmerican (8,481) Hispanic (2,902) Other (907) Total (55,304)

Type and Indicator of Treatment

Marilyn C. Daley

Percentage of Episodes Substance abuse hospital Freestanding detoxication Level 3 detoxication Acute residential Acute residential under 21 Mental health residential Mental health residential <21 Total 16.1 42.4 12.4 15.0 1.0 7.7 5.3 100% 5.49 28.0% $1,188 6.09 21.7% $1,258 5.95 25.4% $1,241 14.3 51.8 6.9 17.7 0.3 4.5 4.7 100% 12.1 47.6 11.3 12.0 1.8 3.8 11.5 100%

10.0 32.1 8.1 13.0 2.8 12.2 21.8 100% 7.08 27.8% $1,670

15.5 44.0 11.4 15.2 1.0 7.1 5.8 100% 5.74 26.9% $1,220

Service Intensity Average length of stay (days) LOS <3 days (%) Average per episode cost

467

468

Administration and Policy in Mental Health

67% more likely to enter alcohol or drug treatment services after managed care (in FY1996) than before (in FY1992). Despite this, the rate of access for Hispanics remained only half that of Blacks and one-third that of whites, even in FY 1996. Although access to treatment was consistently higher for Blacks than for Hispanics, the penetration rate for Blacks improved less relative to Hispanics in the years following managed care.
Continuity of Care

Overall, clients were 73% more likely to complete referrals for continuing treatment after managed care was introduced (see Figure 2). Furthermore, the percent of Blacks who were able to complete referrals for continuing care became signicantly higher after managed care was introduced, relative to the other two groups. When compared to other racial groups, Black clients were 50% more likely to achieve continuity in the period following managed care. Although continuity improved dramatically after managed care for all three ethnic groups, Blacks and Hispanics remained less likely to receive continuing services in every year of the study (see Table 3).

FIGURE 1 Changes in Access to Behavioral Health Care for Racial Minorities Before and After Managed Care
25
Unique clients per thousand enrollees who received AOD services
20.6 19.8 21.0

20
16.4

15

14.4

13.0

13.2

10

10.8

11.7 6.1

11.9 7.5 7.8 7.3

4.5

White Black Hispanic

0 1992 1993 1994 Fiscal year 1995 1996

Marilyn C. Daley

469

FIGURE 2 Changes in Continuity of Care for Racial Minorities Before and After Managed Care
Percent admitted to a lower level of care within 14 days of discharge from acute treatment 30%

26.0%
25%

27.8% 24.6% 21.1% 20.6% 24.1% 21.9% 22.2% 25.2% 23.3%

23.1% 21.6% 17.4% 16.7%

20%

15%

11.3%
10%

White Black Hispanic

5%

0% 1992 1993 1994 1995 1996

Rapid Re-admissions

30-day re-admission rates (Figure 3) increased signicantly (+34%) following managed care, even after controlling for demographic changes in the enrolled population. In a reversal of other trends, the rates of TABLE 3 Logistic Regression: Impact of Managed Care on 30-day Re admissions and Continuity of Care for Blacks and Hispanics Re admission Within 30 Days Explanatory Variable Post-managed Black Black post-MC Hispanic Hispanic post-MC Under 21 Male Heroin Disabled Odds Ratio 1.24 0.71 0.86 0.73 1.13 0.54 1.27 1.03 1.80 Lower CL 1.15 0.59 0.70 0.50 0.77 0.49 1.22 0.98 1.70 Upper CL 1.33 0.86 1.06 1.07 1.67 0.60 1.33 1.08 1.90 Odds Ratio 1.73 0.55 1.50 0.85 0.93 0.23 0.91 1.06 0.74 Continuity Lower CL 1.60 0.45 1.22 0.63 0.67 0.21 0.87 1.01 0.71 Upper CL 1.86 0.67 1.84 1.16 1.28 0.26 0.95 1.11 0.78

470

Administration and Policy in Mental Health

FIGURE 3 Changes in 30-day Re admissions to Acute Treatment for Racial Minorities Before and After Managed Care
Percent readmitted to acute treatment services within 30 days of discharge
25%

22.9% 21.3%

22.0%

23.2% 18.7%

20%

18.3%
15%

14.6% 12.3% 11.5% 11.5% 10.3% 12.2%

15.1% 15.0%

17.2%

10%

White Black Hispanic

5%

0% 1992 1993 1994 1995 1996

Fiscal Year

re-admissiona poor outcomewere higher for whites than they were for either Blacks or Hispanics.
Spending

Bivariate analysis indicated no signicant differences in spending or length of stay in treatment between whites, Blacks, and Hispanics (see Table 2). A regression analysis (not shown) suggested that reductions in per episode spending as a result of managed care did not differ signicantly by race.

DISCUSSION
Prior to managed care, access and continuity reected the same pattern in Massachusetts as in the national literaturethey were worse for minorities than for whites. The introduction of a MBHC carve-out in 1993 was largely benecial for racial minority groups in Massachusetts. In two of the three quality indicators that we chose to examineaccess to substance abuse treatment and continuity of careminorities enjoyed a greater level of improvement as a result of managed behavioral health care than did whites. For the third quality indicator, rapid readmissions,

Marilyn C. Daley

471

outcomes for African American, Hispanic, and white Medicaid clients did not differ. Hispanics enjoyed the largest increase in access to behavioral health treatment over the period of observation, and the percent of Blacks who received continuing treatment services improved signicantly relative to any other ethnic group. Another encouraging observation was that rapid re-admissions remained signicantly lower for Blacks throughout the study. Most importantly, not one of our quality indicators deteriorated for minority groups relative to whites following managed care. This suggests that the carve-out did not neglect services to minorities in an effort to contain costs as some professionals had feared. On the contrary, managed care may have employed some strategies to make it easier for minority groups to receive services. The Department of Medical Assistance granted MHMA more leverage than a lot of MBHCs, so that providers would generally follow their directives. MHMA had the power to close facilities if they were not operating in a cost-effective manner. At the outset, they contracted with a network of established community-based providers, many of whom had experience treating minorities. In addition, they tried to encourage cooperation between them through performance measurements and other strategies. In contrast, other MBHCs have been criticized for selectively contracting with the lowest bidder, which is especially damaging for mental health clinics serving the minority community who may be struggling nancially and lack the resources to compete. If clinics based in ethnic communities were forced to close, the minority clients who use them might not be able to access treatment in more distant neighborhoods, and might lose a therapeutic relationship with their clinician that may have taken years to develop. By comparison, MHMA allowed all existing outpatient providers to renew their contracts without a competitive bidding process. This permitted clients to stay with their current clinicians, thus preserving an established relationship between clients and staff (Counihan, Nelson, & Patullo 1996). MHMA also emphasized discharge planning, case management, and performance measurement (Counihan et al., 1996). In 1994, after renewing MHMAs contract for two years, an aggressive quality management program was implemented. A director of quality management was hired, and a quality improvement plan was drawn up. As one of their rst tasks, DMA and MHMA established eight interdepartmental teams to develop and implement eight key quality improvement activities amenable to measurement and improvement. Three of the eight identied that performance targets included increasing access to non English-speaking clients, increasing access to outpatient services, and improving continuity of care (Witherbee, 1994).

472

Administration and Policy in Mental Health

Although some measures of quality improved for minority clients, penetration rates remained consistently lower for Hispanics and Blacks than they were for whites in every year of the study. Historically, treatment programs have had problems recruiting and retaining minority clients. Persons of color may avoid large bureaucracies and instead try to solve problems on their own. A history of negative experiences with the mainstream culture combined with a legacy of self-reliance makes these groups especially resistant to the idea of treatment by the society they have grown to mistrust (Hogue, 2002). While poverty and socioeconomic status can explain some of the variation, there are still disparities even when controlling for these factors (Schulz et al., 2000). For example, all of the clients in our study had Medicaid, but minorities were still less likely to utilize behavioral health services. Since many types of therapy rely heavily on verbal communication, language barriers are a signicant problem. Aside from the risk of misdiagnosis, non-English speaking Hispanics will be less likely to understand the physicians instructions, less likely to receive psychotropic medications (Sclar, Robin, Staer, & Galin, 1999; Malgady et al., 1987), and perhaps most importantly, less likely to benet from support groups or talk therapy. Communication problems can adversely affect many aspects of mental health and substance abuse counseling, where the ability to communicate is necessary to diagnose, plan, and benet from treatment. Regardless of whether the therapy is provided in a family, individual, or group context, it is clearly necessary for feelings, emotions, and problems to be communicated and discussed clearly and openly. In addition, since social support is felt to be a key element in the effectiveness of group therapy, therapeutic communities and 12-step groups, those who are perceived as outside of the mainstream culture could be excluded from these networks. If minorities cannot feel a sense of fellowship with other addicts who attend mutual support groups, it may be difcult for them to recover (Arroyo, Westerberg, & Tonigan, 1998; Katsukas Weisner, Lee, & Humphreys, 1999). The literature on minority health discusses many other factors that may affect the quality of treatment for Blacks and Hispanics. Minorities may have misconceptions about the causes of diseases (Hubell, Chavez, Mishra, & Vega, 1996), they might lack of knowledge about health care options (Gerstein et al., 2002; Vega et al., 1999), know to get free care, have problems with transportation (Vega et al., 1999), need child care, be intimidated by doctors (Hubbell et al., 1996), be stigmatized (Hubbell et al., 1996), use home remedies (Marks et al., 1987), or rely more on their family (Kline, 1996). Results from Massachusetts suggest that the states public sector managed behavioral health care program reduced gaps on access and continuity between minorities and whites, but further work is needed to eliminate these disparities.

Marilyn C. Daley

473

Study Limitations

There are several limitations to the study. It is not possible to establish whether access and continuity rates improved for minorities due directly to managed care or to other changes in demographics, policy, or practices that operated during this period. For example, there was increasing recognition of the importance of cultural forces in recovery from drug abuse, which resulted in several state and federal initiatives to design programs that would be more minority-targeted, culturally-focused, and linguistically appropriate. The Surgeon Generals Report on mental health and minorities was issued in 2001 (SAMHSA/CMHS 2001a, b), followed by Healthy People 2010 and The Presidents New Freedom Initiative (SAMHSA, 2003). These culminated in the creation of the DSM-IV task force on cross-cultural issues, which designed a glossary of culture-bound syndromes to assist physicians in diagnosis and treatment (Marks et al., 1987; Mathews, Glidden, Murray, Forster, & Hargreaves, 2002; Musgrave, Allen, & Allen, 2002; Opler, 2002).

Not one of our quality indicators deteriorated for minority groups relative to whites following managed care.
Locally, the Massachusetts Department of Public Health, Bureau of Substance Abuse Services (BSAS) made efforts to increase access for minority clients by expanding the number of programs that were located in minority communities, hiring staff who were culturally and linguistically matched to the clients in these programs, and implementing training programs for existing staff members in cultural competence. Several targeted residential treatment programs were opened such as Casa Esperanza, a therapeutic community for pregnant Hispanic women and their children located in Jamaica Plain, an area of Boston with a large number of Hispanic immigrants (McCarty et al., 2000). Between FY1992 and FY1996, BSAS MIS data showed that admissions for Hispanics increased by 31% across Massachusetts, but this was only part of a larger 501% increase in Hispanic admissions that occurred between FY1984 and FY1996. At the same time, there was a 147% increase in access for Blacks (Lundgren et al., 2001; McCarty et al., 2000). There was also a rapid escalation in the purity and availability of heroin in the greater Boston area in the 1990s, which may have caused a few Hispanics to develop habits severe enough to precipitate treatment entry for the rst time (Clark, Krakow, Panas, & Brolin, 1997; McCarty, LaPrade, & Botticelii, 1996). Most of the Hispanics who live in the

474

Administration and Policy in Mental Health

Boston area are Puerto Ricans, one of the Hispanic groups that are most prone to abuse heroin in national and local surveys of drug treatment programs (Clark et al., 1997; Collins et al., 2002; Weineck et al., 2004). Partially in response to the spread of AIDS among intravenous drug users (IVDU), the Bureau of Substance Abuse Services expanded methadone programs and other services to IVDU. For example, a mobile methadone unit made it possible to provide treatment to many minority IVDU in some communities that objected to the permanent sitting of a methadone clinic (McCarty et al., 1997). Finally, as a result of mandatory drug penalties and stricter enforcement, there may also have been an increase in the number of admissions for minority clients who were mandated by the courts during this time (Pescosolido, Garder, & Lubell, 1998). Another limitation is that Medicaid claims and eligibility les were used to construct our quality indicators, so we were unable to measure treatment received in halfway houses, therapeutic communities, or other settings not reimbursed through Medicaid. However, since this exclusion affects the pre- and post-managed care periods equally, it should not introduce bias into the study. In addition, the data are nearly eight years old, and since that time, MHMA has been replaced by another vendor. Further research is in progress to see whether the favorable results for minorities reported here have continued under the second managed care vendor. Measures of minority groups are very crude, since they reect racial background, not ethnicity, and Blacks and Hispanics may vary widely in their country of origin, culture, and lifestyle (Weineck et al., 2004). Despite these limitations, however, there is no evidence from this study that MHMA had a detrimental impact on the quality of services received by racial minorities in Massachusetts, and the carve-out may in fact have improved the ability of minorities to obtain needed treatment services.

REFERENCES
Alegria, M., Canino, G., Rios, R., Vera, M., Calderon, J., Rusch, D., & Ortega, A.N. (2002). Inequalities in use of specialty mental health services among Latinos, AfricanAmericans, and non-Latino Whites. Psychiatric Services, 53(12), 15471555. Argeriou, M., & Daley, M. (1997). An examination of racial and ethnic differences within a sample of Hispanic, white (non-Hispanic), and AfricanAmerican Medicaid-eligible pregnant substance abusers: The MOTHERS project. Journal of Substance Abuse Treatment, 14(5), 489498. Arroyo, J.A., Westerberg, V.S., & Tonigan, J.S. (1998). Comparison of treatment utilization and outcome for Hispanics and non-Hispanic whites. Journal of Studies on Alcohol, 59(3). Bureau of Labor Statistics. (2002). Consumer price index for all urban consumers, Boston area, Medical component. Washington, DC: US Department of Labor. Callahan, J.J., Shepard, D.S., Beinecke, R.H., Hodgkin, D., Larsen, M.J., & Cavanaugh, D. (1995). Mental health/substance abuse treatment in managed care: The Massachusetts Medicaid experience. Health Affairs, 14(3), 173184.

Marilyn C. Daley

475

Centers for Disease Control National Center for Health Statistics (2003). Health, United States, Chartbook. National Center for Health Statistics. Hyattsville MD. Available at: www.nchs.cdc.gov. Centers for Disease Control Ofce for Minority Health Research. (2004). Available at: www.omhrc.gov. Clark, T.M., Krakow, M., Panas, L., & Brolin, M. (1997). Drug use trends in greater Boston and Massachusetts: 1991 to 1996. Boston, MA: Health & Addictions Research Inc. Collins, K.S., Hughes, D.L., Doty, M.M., Ives, B.L., Edwards, J.N., & Tenney, K. (2002). Diverse communities, common concerns: Assessing health care quality for minority Americans: Findings from the Commonwealth Fund 2001 Health Care Quality Survey. New York: The Commonwealth Fund. Counihan, C.W., Nelson, D., & Patullo, E. (1996). A Medicaid mental health carve-out program: The Massachusetts experience. Managed Care Quarterly, 4(3), 8592. Cunningham, P.J., & Trude, S. (2001). Does managed care enable low income persons to identify a usual source of care? Implications for access to care. Medical Care, 39(7), 716726. Fiscella, K., & Franks, P. (1997). Does psychological distress contribute to racial and socioeconomic disparities in mortality? Social Science and Medicine, 45(12), 18051809. Gerstein, D.R., Greenblatt, J., Johnson, R.A., Larison, C., Bacellar, H., & Choy, W. (2002). Prevalence of substance abuse among racial and ethnic subgroups in the United States. Rockville, MD: U.S. DHHS, SAMHSA, Ofce of Applied Studies and the National Opinion Research Center (NORC). Greenberg, G., Brandon, W.P., Schoeps, N., Tingle, L.R., & Shull, L.D. (2003). Medicaid managed care and racial differences in satisfaction and access. Journal of Health Care for the Poor and Underserved, 14(3), 351371. Hargreaves, J.L., Cunningham, P.J., & Hughes, R.G. (2001). Racial and ethnic differences in access to medical care in managed care plans. Health Services Research, 36(5), 853868. Hogue, C.J.R. (2002). Toward a systematic approach to understandingand eliminatingAfrican American womens health disparities. Womens Health Issues, 12(5), 222237. Howard, D.L., LaVeist, T.A., & McCaughrin, W.C. (1996). The effect of social environment on treatment outcomes in outpatient substance abuse treatment: Does race really matter? Substance Use & Misuse, 31(5), 617638. Hubbell, F.A., Chavez, L.R., Mishra, S.I., & Valdez, R.B. (1996). Differing beliefs about breast cancer among Latinas and Anglo women. Western Journal of Medicine, 164, 405409. Institute of Medicine. (2002). Unequal treatment: Confronting racial and ethnic disparities in healthcare. Washington, DC: National Academy of Sciences. Katsukas, L.A., Weisner, C., Lee, M., & Humphreys, K. (1999). Alcoholics Anonymous afliation at treatment intake among white and Black Americans. Journal of Studies on Alcohol, 60(6), 810815. Kline, A. (1996). Pathways into drug user treatment: The inuence of gender and racial/ethnic identity. Substance Use & Misuse, 31(3), 323342. Leigh, W.A., Lillie-Blanton, M., Martinez, R.M., & Collins, K.S. (1999). Managed care in three states: Experiences of low-income AfricanAmericans and Hispanics. Inquiry, 36, 318331. Lillie-Blanton, M., & LaVeist, T. (1996). Race/ethnicity, the social environment, and health. Social Science and Medicine, 43(1), 8391. Lopez, S.R. (2002). A research agenda to improve the accessibility and quality of mental health care for Latinos. Psychiatric Services, 53, 15691573. Lundgren, L.M., Amodeo, M., Ferguson, F., & Davis, K. (2001). Racial and ethnic differences in drug treatment entry of injection drug users in Massachusetts. Journal of Substance Abuse Treatment, 21, 145153. Malgady, R.G., Rogler, L.H., & Constantino, G. (1987). Ethnocultural and linguistic bias in mental health evaluation of Hispanics. American Psychologist, 42, 228234. Marks, G., Solis, J., Richardson, J.L., Collins, L.M., Birba, L., & Hisserich, J.C. (1987). Health behavior of Hispanic women: Does cultural assimilation make a difference? American Journal of Public Health, 77(10), 13151319. Mathews, C.A., Glidden, D., Murray, S., Forster, P., & Hargreaves, W.A. (2002). The effect on treatment outcomes of assigning patients to ethnically focused inpatient psychiatric units. Psychiatric Services, 53(7), 830835. McCarty, D., LaPrade, J., & Botticelli, M. (1996). Substance abuse treatment and HIV services: Massachusetts policies and programs. Journal of Substance Abuse Treatment, 13(5), 429438. McCarty, D., Caspi, Y., Panas, L., Krakow, M., & Mulligan, D. (2000). Detoxication centers: Who is in the revolving door? Journal of Behavioral Health Services and Research, 27(3), 245256. Moos, R.H., Moos, B.S., & Finney, J.W. (2001). Predictors of deterioration among patients with substance-use disorders. Journal of Clinical Psychology, 57(12), 14031419.

476

Administration and Policy in Mental Health

Mulvaney, F.D., Brown, L.S., Alterman, A.I., Sage, R.E., Cnaan, A., Cacciola, J., & Rutherford, M. (1999). Methadone maintenance outcomes for Hispanic and AfricanAmerican men and women. Drug and Alcohol Dependence, 5, 1118. Mukherjee, S., Shukla, S., Woodle, J., Rosen, A.M., & Olarte, S. (1983). Misdiagnosis of schizophrenia in bipolar patients: A multiethnic comparison. American Journal of Psychiatry, 140, 15711574. Musgrave, C.F., Allen, C.E., & Allen, G.J. (2002). Spirituality and women of color. American Journal of Public Health, 92(4), 557560. Opler, L.A. (2002). The culture of emotions: A cultural competence and diversity training program, review. Psychiatric Services, 53(12), 15371538. Pescosolido, B.A., Garnder, C.B., & Lubell, K.M. (1998). How people get into mental health services: Stories of choice, coercion and muddling through from rst-timers. Social Science and Medicine, 46(2), 275286. Regier, D.A., Narrow, W.E., Rae, D.S., Manderscheid, R.W., Locke, B.Z., & Goodwin, F.K. (1993). The de facto U.S. mental health and addictive disorder service system: Epidemiologic catchment area prospective one-year prevalence rates of disorders and services. Archives of General Psychiatry, 59, 8594. Schulz, A., Israel, B., Williams, D., Parker, E., Becker, A., & James, S. (2000). Social inequality, stressors, and self-reported health status among AfricanAmerican women in the Detroit metropolitan area. Social Science and Medicine, 51, 16391653. Sclar, D.A., Robison, L.M., Skaer, T.L., & Galin, R.S. (1999). Ethnicity and the prescribing of antidepressant pharmacotherapy. Harvard Review of Psychiatry, 7, 2936. Shepard, D.S., Daley, M., Ritter, G.A., Hodgkin, D., & Beinecke, R.H. (2001). Effects of a statewide carve-out on spending and access to substance abuse treatment in Massachusetts, 1992 to 1996. Health Services Research, 36(6), 3244. Substance Abuse Mental Health Services Administration (SAMHSA). (2003). Presidents New Freedom Commission on Mental Health: Achieving the promise: Transforming mental health care in America. Rockville, MD: U.S. DHHS, SAMHSA (SAMHSA Publication Number SMA 03-3832). SAMHSA Center for Mental Health Services (CMHS). (2001a). Mental Health: A report of the Surgeon General. Rockville, MD: SAMHSA/CMHS. SAMHSA Ofce of Applied Studies. (1998). Services Research Outcomes Study. Rockville, MD: US DHHS SAMHSA. Publication No. (SMA) 98-3177. U.S. Census Bureau (2002). Census 2000 Summary File 1. Available at http://www.factnder.census.gov. Vega, W.A., Kolody, B., Aguilar-Gaxiolar, S., & Catalano, R. (1999). Gaps in service utilization by MexicanAmericans with mental health problems. American Journal of Psychiatry, 156(6), 928934. Weineck, R.M., Jacobs, E.A., Stome, L.C., Ortega, A.N., & Burstin, H. (2004). Hispanic healthcare disparities: Challenging the myth of a monolithic population. Medical Care, 42(4), 313320. Witherbee, D. (1994). History of the Mental Health/Substance Abuse Treatment Program in Massachusetts. Schneider Institute for Health Policy. Waltham, MA: Brandeis University, Unpublished report.