Deck DD, McFarland BH, Titus JM, Laws KE, Gabriel RM. Access to Substance Abuse Treatment Services Under the Oregon Health Plan. JAMA. 2000;284(16):2093–2099. doi:10.1001/jama.284.16.2093
Author Affiliations: RMC Research Corporation (Drs Deck, Titus, and Gabriel and Ms Laws) and Oregon Health Sciences University (Dr McFarland), Portland, Ore.
Context The shift to Medicaid managed care has raised numerous concerns about
access to publicly funded substance abuse treatment. The implementation of
a capitated chemical dependency benefit within the Oregon Health Plan in 1995
provided an opportunity to study the impact of funding mechanisms on access.
Objectives To determine to what extent access to publicly funded substance abuse
treatment changed following the shift to managed care in Oregon and to examine
factors associated with that change.
Design Analysis of statewide treatment and Medicaid eligibility data.
Setting and Patients All Medicaid-eligible persons aged 12 to 64 years who were enrolled
in the Oregon Health Plan during 1994 (88,320), 1996 (170,387), 1997 (160,929),
or 1998 (149,877).
Main Outcome Measures Access rates (the number of unique individuals admitted to treatment
during a calendar year divided by the average number of enrolled members)
computed before (1994) and after (1997) implementation of the capitated benefit.
Analyses were replicated with data for 1996 and 1998.
Results The percentage of Medicaid-eligible persons admitted to substance abuse
treatment programs during a calendar year increased from 5.5% of the average
number of enrolled members per month in 1994 to 7.7% in 1997, following the
shift to managed care. For 1996 and 1998, the rates were 6.9% and 7.7%, respectively.
Access rates varied considerably among the 7 largest prepaid health plans
after adjusting for case mix. Operating characteristics of these prepaid health
plans, such as the method of reimbursing treatment providers, were significant
predictors of access after controlling for member characteristics.
Conclusion According to our analyses, Medicaid-eligible persons in Oregon observed
an increase in access to substance abuse treatment after a shift to managed
The implementation of managed care has raised numerous concerns about
access to health services.1 The American Society
for Addiction Medicine has suggested that chemical dependency treatment programs
have experienced a greater impact from managed care than other sectors of
the health care industry. For example, an American Society for Addiction Medicine
working group reported a 75% reduction in the value of addiction insurance
coverage from 1988 to 1998 for mid-size to large-size companies.2,3
The Substance Abuse and Mental Health Services Administration found that chemical
dependency and mental health services declined from 9% of the national health
care expenditures in 1986 to 8.1% in 1996.4- 7
A key question concerns the impact of these trends on access to publicly funded
chemical dependency treatment.
Although these developments are under way in the larger health care
sector, states have been converting their Medicaid programs from fee-for-service
to managed care.8 Most Medicaid recipients
are now enrolled in managed care systems.8
At least 48 states have federal waivers allowing them to replace fee-for-service
Medicaid with managed care.6,9
Chang et al10 suggested that, at least in some
states, these changes have decreased access to chemical dependency treatment
services for Medicaid recipients. However, few quantitative data are available.
Although comparison between fee-for-service and managed care Medicaid
is of great theoretical interest, a return to unmanaged Medicaid seems unlikely.3 A more practical question concerns the difference
in access among various types of managed care systems, such as the contrast
between "integrated" health care systems (which provide both medical-surgical
and behavioral health care) and "carve out" programs (in which behavioral
health services are delivered by agencies distinct from the systems that provide
medical-surgical care).11 Tightly integrated
systems may facilitate the referral of patients with substance abuse disorders
from primary care to treatment programs, but integrated systems might direct
funds intended for substance abuse treatment into medical-surgical services.
Other dimensions of interest include the mechanism used to reimburse treatment
providers, preauthorization of treatment services, and for-profit vs not-for-profit
systems. What is needed to address these issues is a setting in which several
approaches to managed care are being implemented. Conveniently, Oregon has
been conducting just such a natural experiment.
In the early 1990s, the Oregon legislature initiated a demonstration
project to expand Medicaid coverage while controlling the rising public cost
of health care through a series of bills that became known as the Oregon Health
Although national attention focused on the controversial rationing of health
care through a prioritized list of covered services,18
the initiative incorporated other important innovations, such as expanded
eligibility and integration of chemical dependency services. The Health Care
Financing Administration granted Oregon a 5-year Medicaid waiver that enabled
the state to expand coverage to all adults and their dependents whose income
fell below the federal poverty limit.19 This
section 1115 waiver took effect in February 1994, more than doubling the Medicaid-eligible
population. To manage the delivery of covered medical services, Medicaid recipients
were required to enroll in one of the prepaid health plans with a contract
to serve the county in which they reside. By the third quarter of 1995, more
than 80% of the Medicaid-eligible population had enrolled in a prepaid health
In May 1995 the state added a capitated chemical dependency benefit
to improve the integration of substance abuse treatment with physical health
care. This benefit covered outpatient treatment services, including regular
outpatient, intensive outpatient, and methadone maintenance. This new program
placed responsibility for chemical dependency treatment services for Medicaid
recipients with 20 prepaid health plans, markedly altering the organization
and financing of these services. Under the preexisting fee-for-service system,
the traditional public sector agencies that delivered services were either
freestanding, not-for-profit agencies or components of county governments.
The state reimbursed treatment providers via several funding streams, including
Medicaid fee-for-service and "slot rate" payment programs for low-income participants.20 Little or no connection existed between the public
sector substance abuse treatment system and the mental health treatment system
or the larger medical-surgical health care sector.
Officials from Oregon's Office of Alcohol and Drug Abuse Programs and
other stakeholders were concerned that the rate of participation in substance
abuse treatment might be low under managed care if health plan officials believed
that promoting such services was not in their best interest. Therefore, the
state increased capitation payments to encourage primary care practitioners
to screen their patients for substance abuse disorders.
Not-for-profit managed care organizations operated 5 of the 7 largest
health plans that covered Oregon Health Plan participants during the study
period (Table 1). Most of these
plans subcontracted with traditional community providers for substance abuse
treatment services, but 2 integrated plans operated their own programs. However,
the state required even these plans to refer at least half of their substance
abuse patients to traditional community treatment providers. The prepaid health
plans reimburse treatment providers using 1 of 3 approaches: (1) modified fee-for-service, whereby the plan pays a certain percentage
of the state rate for each billable service provided to the participant; (2) subcapitation, whereby the plan pays a fixed fee each month
based on the number of enrolled members in that service area; or (3) case rate, whereby the plan pays a fixed fee for each new
treatment admission, regardless of the level of care provided. Typically,
the plans reimburse methadone maintenance services under fee for service regardless
of the method used for other outpatient services.
During the study period, the health plans increasingly relied on intermediate
organizations to administer the chemical dependency benefit. Health plan representatives
stated that contracting with such agencies was more efficient because the
plans lacked personnel with expertise in managing substance abuse treatment.
In contrast, an intermediate organization could manage treatment services
for several health plans.
This study examined access to substance abuse treatment services following
the implementation of a capitated chemical dependency benefit under the Oregon
Health Plan and identified plan characteristics that contributed to differences
in access among plans.
We calculated access rates by dividing the number of substance abuse
treatment users by the number in the population eligible for those services.
For this study, we defined a treatment user as a person aged 12 to 64
years entering publicly funded substance abuse treatment who was enrolled
in the Oregon Health Plan within 30 days of the admission date. The counts
included only unique individuals admitted to treatment at least once in a
calendar year and excluded Medicare beneficiaries who were also eligible for
The Client Process Monitoring System managed by Oregon's Office of Alcohol
and Drug Abuse Programs identified treatment users. Each record in this database
represents an episode of care with an admission date, discharge date, and
the modality of service. Merging this file with the statewide Medicaid eligibility
and enrollment files from Oregon's Office of Medical Assistance Programs verified
each member's eligibility category and enrollment status within the first
30 days of each treatment episode. Quality control processes removed duplicate
records and services other than treatment. Counting only the first admission
for each calendar year identified unique individuals. Residential admissions
(including detoxification) were counted in addition to outpatient and methadone
admissions since alternative funding was available for persons meeting placement
criteria for residential services. Approximately 75% of persons admitted to
treatment in a calendar year have at least 1 outpatient or methadone admission.
A comparison of Client Process Monitoring System data with treatment
provider records as part of 2 parallel projects16,21,22
suggested that reporting to the system is generally complete and accurate
but also revealed that Medicaid identifiers were missing for some recipients,
resulting in an underestimate of access.
We averaged the number of Medicaid recipients aged 12 to 64 years enrolled
in a prepaid health plan on the first day of each month during a calendar
year. This count excluded Medicare recipients, persons eligible under programs
exempt from the Oregon Health Plan, and persons living in areas lacking a
prepaid plan. Because the 7 largest health plans enrolled more than three
quarters of all Oregon Health Plan members, the analysis collapsed smaller
health plans into a single category.
For this study, the analysis lumped Medicaid eligibility programs into
5 categories, which remained relatively stable over time and, where possible,
have counterparts in other states: disabled (Supplemental Security Income
recipients), welfare (Aid for Families With Dependent Children or Temporary
Aid for Needy Families recipients), other poverty programs, expansion (single
individuals and childless couples newly eligible under the section 1115 waiver),
and all other programs.
We adjusted each plan's access rates for differences on 4 demographic
variables: age group (12-17, 18-30, and 31-64 years), sex (male, female, unknown),
ethnicity (white, minority, unknown), and eligibility category.
We used logistic regression to examine the relative contribution of
member characteristics and health plan operating characteristics in predicting
access rates. We dummy coded plan characteristics using the results of structured
interviews with officials from each of the 7 largest health plans.
Access rates for 1994 serve as a baseline, reflecting the expanded population
of Medicaid recipients but not the capitation of chemical dependency services.
Access rates for 1996 through 1998 reflect the added impact of the implementation
and maturation of the chemical dependency benefit.
Statewide, the access rate increased substantially between 1994 and
1997 (Figure 1) for Oregon Health
Plan recipients aged 12 to 64 years. In 1994, when Medicaid reimbursed substance
abuse treatment providers under fee-for-service, the rate was 5.5%. Following
capitation, the access rate increased to 7.7% by 1997. For 1996 and 1998,
the rates were 6.9% and 7.7%, respectively.
Access differed by subgroup in 1997 and exhibited differential change
by subgroup between 1994 and 1997 under the Oregon Health Plan. Males in 1997
were more likely to participate in substance abuse treatment than females
and experienced a greater increase in access between 1994 and 1997. Whites
were slightly less likely to participate than minorities in 1997. Adolescents
(aged 12 to 17 years) were about a third as likely to be admitted to treatment
as adults older than 30 years in 1997. Disabled persons were the least likely
to be admitted to treatment during 1997. This category was exempt from enrollment
during 1994. Members in the welfare and expansion categories, which together
compose most of the statewide enrollment, experienced much higher access rates.
The eligibility category labeled "other" incorporates several programs that
serve persons who are very likely to participate in substance abuse treatment,
but this category only represents a small proportion (3%) of the state enrollment
within this age range.
The access rates for 3 of the 7 largest prepaid health plans increased
dramatically between 1994 and 1997 (Figure
2). Plan F and plan G, operated by integrated managed care organizations,
had the lowest rates in 1997, representing a decrease or no change from 1994.
Plan D was not operating in 1994. The access rates in 1997 varied widely across
plans, even after adjusting for differences in case mix (Table 2). The plan with the greatest penetration of substance abuse
treatment services admitted more than twice the percentage of its enrolled
members as the plan with the least service use. In 1997 the lowest rate among
the 7 plans was about 6% of enrolled members and the highest rate was nearly
16%. The rates for each plan varied somewhat in 1996 and 1998, but the ranking
among plans remained fairly constant.
The health plan with the highest overall access rate in 1997 (plan A)
has a reputation for more extensive outreach among potential treatment users.
An interview with an official from this plan revealed that the managed care
organization has no commercial plan and perceives its mission as serving low-income
families. Its medical clinics routinely screen for substance abuse, and the
plan maintains strong ties with social service providers in the metropolitan
area it serves. Plan B, which had the second highest participation rate, has
some business ties with plan A. Both plans contract with the same intermediate
organization to manage the chemical dependency benefit for their Medicaid
members. These 2 plans also had the highest percentage of admissions to residential
substance abuse treatment, a service that is covered by state funds other
The results of a logistic regression analysis show that, as we expected,
member characteristics—except the stability of plan enrollment (ie,
whether the member was enrolled in the same plan after 12 months)—significantly
predicted access (Table 3). Adults
were more than twice as likely to access substance abuse atment as adolescents.
Welfare recipients were more than twice as likely to access treatment as disabled
persons. Males were more likely to access than females, whites more likely
than minorities, urban or town dwellers more likely than rural residents,
and expansion members more likely than disabled persons.
More importantly, 3 of 4 health plan operating characteristics were
also significant predictors. Members enrolled in plans that reimbursed substance
abuse treatment providers using a modified fee-for-service approach were 55%
more likely to access substance abuse treatment than those in integrated plans
that used case rates. Members enrolled in plans that reimbursed providers
through subcapitation were 35% more likely to access treatment than those
in plans that used case rates. Members enrolled in for-profit plans were somewhat
more likely to access treatment than those in not-for-profit plans. Prior
authorization was not a significant predictor. Similar results were observed
for the 1996 and 1998 cohorts.
Contrary to expectations, access to substance abuse treatment substantially
increased for Medicaid recipients when Oregon integrated a capitated chemical
dependency benefit with the primary care benefit under the Oregon Health Plan.
In 1994, before prepaid health plans were responsible for substance abuse
treatment, 5.5% of those enrolled for the primary care benefit accessed treatment
services, consistent with access observed in previous years. After integration,
approximately 40% more enrollees accessed treatment each year. Analyses controlling
for member characteristics suggest that the high access rates observed were
not simply due to the expansion of the population eligible for Medicaid and
do appear to be associated with implementation of an innovative, carefully
designed managed care system.
Although these data show that access to substance abuse treatment improved
for Oregon Medicaid recipients following the implementation of managed care,
one might hypothesize that access to such services in Oregon might have been
low compared with that in other states. National treatment data from the Treatment
Episode Data Set23 and eligibility data from
the Health Care Financing Administration24
appear to show that access to treatment services had been somewhat greater
in Oregon. Thus, managed care did not simply elevate utilization from a low
preexisting baseline. Rather, the access rate increased above levels that
already were considerable, at least when compared with national data.
One might also hypothesize that other concurrent events might explain
these results. For example, the economy of Oregon was strong during this period,
so perhaps individuals with less need for treatment tended to find employment
and lose eligibility, thus changing the risk profile of the remaining enrollees.
However, the empirical evidence does not support this argument. On the other
hand, Oregon did initiate welfare reforms25
as early as 1992, before the integrated benefit. Under a federal waiver, substance
abuse treatment could be treated as a work activity. Although this initiative
may account for some of the change observed with welfare recipients, it does
not account for the change observed with other eligibility groups.
Thus, unlike the experiences of managed care members in other states,
Oregon Medicaid recipients experienced an increase rather than a decrease
in access to substance abuse treatment services when capitation replaced fee-for-service
financing. A rather dramatic rise in admissions to publicly funded substance
abuse treatment occurred. Even though health plans were at financial risk
for providing substance abuse treatment (as well as medical-surgical care),
no evidence indicated that the plans limited their Medicaid recipients' access
to substance abuse treatment. This estimate is still conservative, because
the definition of access excludes persons continuing in treatment from a prior
year. In addition, missing Medicaid identifiers for some recipients who entered
treatment resulted in an undercount of treatment users.
A substantial gap still exists, however, between access and the level
of need for substance abuse treatment. Feyerherm and Skokan26
conducted a household survey in 1994 to estimate need for treatment in the
expanded Medicaid population. They estimated that 16.5% of the respondents
likely to be found eligible for Medicaid under the expanded requirements met
diagnostic criteria for substance abuse or dependence.27
The findings of this study suggest that less than half of those with a substance
abuse disorder enrolled in the Oregon Health Plan access publicly funded substance
abuse treatment during a single year.
Although the statewide access rate is relatively high, substantial differences
exist among prepaid health plans and types of Medicaid recipients. The exact
mechanisms responsible for these high access rates or the differences among
plans are not entirely clear. The high level of integration of primary care
and substance abuse services expected under the Oregon Health Plan has not
been achieved overall, although some evidence suggests that such integration
may have been partially responsible for the high participation rates achieved
by at least 1 plan.
Contrary to expectations, the most tightly integrated health care system
had the lowest access rate. Indeed, access to substance abuse treatment by
this health plan's Medicaid recipients actually decreased slightly when capitation
replaced the preexisting fee-for-service financing system. The plan was, however,
required to direct at least half of its treatment referrals to traditional
community treatment facilities. Conversely, other health plans experienced
dramatic increases in the substance abuse treatment access rates of their
Medicaid recipients following implementation of the capitated managed care
system. Especially striking was the performance of a public sector health
plan formed solely to deal with the Medicaid managed care system instituted
under the Oregon Health Plan. This health plan apparently retained, and perhaps
strengthened, the preexisting linkages among public sector primary care and
substance abuse treatment providers. The plan also encouraged its clinicians
to engage in outreach to Medicaid recipients with substance abuse disorders.
The 3 plans with the highest access rates all reimbursed providers using a
modified fee-for-service system. Even after adjusting for the influence of
member characteristics, this provider payment method was still a significant
predictor of access to treatment.
The data from the rural health plans are also interesting. The intermediate
agency that managed substance abuse treatment services for these plans emphasized
outreach to Medicaid recipients with substance abuse disorders. The close
physical proximity of treatment services to social service programs in rural
areas may have been an important factor in the access rates of these plans.
The results of this study should be interpreted in light of certain
limitations. Other factors not represented in the data sets available for
this study could have influenced treatment access,28,29
although no plausible alternative explanations surfaced in discussions with
state officials, plan representatives, or treatment providers. The study relied
on reporting from substance abuse treatment providers to the state database.
This reporting could have been inaccurate, although several validation strategies
suggested that the state data were reasonably accurate and that reporting
was reasonably consistent across health plans. This study focused only on
access to services (ie, admissions to treatment) and did not examine treatment
outcomes, other measures of service utilization (eg, intensity or length of
stay), or satisfaction.
In summary, Oregon Medicaid recipients experienced increased access
to substance abuse treatment services following replacement of a fee-for-service
financing system with capitated financing. Substantial differences among health
plans suggest that appropriate organizational and financial arrangements can
facilitate Medicaid recipients' access to substance abuse programs. We are
conducting further research to study the outcomes of treatment in this pluralistic