Findings from the SAMHSA Managed Behavioral Health Care in the Public Sector Study

1
Findings from the SAMHSA Managed Behavioral
Health Care in the Public Sector Study

• Judith A. Cook, Ph.D.
• Professor and Director
• Center for Mental Health Services Research
  Policy
• Department of Psychiatry, University of Illinois
  at Chicago
• Presented at Using Research to Move Forward A
  Consensus Conference on Publicly Funded Managed
  Care for Children Adolescents with Behavioral
  Health Disorders and Their Families
• September 29 30, 2003, Washington, DC

2
Study Locations, Site and Coordinating Center
PIs, Family Representative

• Rural Counties in NW Oregon
• Portland State University
• Robert I. Paulson, Ph.D.
• Tennessee and Mississippi
• Vanderbilt University
• Craig Anne Heflinger, Ph.D.
• Westchester County, New York
• Columbia University
• Christina Hoven, Dr.P.H.
• Rural Counties in Central Pennsylvania
• University of Pittsburgh
• Kelly Kelleher, M.D.

• Hamilton Summit Counties, Ohio
• Pacific Institute for Research Evaluation,
• Al Stein-Seroussi, Ph.D.
• Coordinating Center
• University of Illinois at Chicago
• Judith A. Cook, Ph.D.
• Family Representative
• Federation for Families
• Valerie Burrell-Mohammed
• Funded by CMHS CSAP of SAMHSA

3
Focus of the Study Children with Severe
Emotional Disorders (SED)

• Inclusion Criteria
• DSM-IV Diagnosis
• Intensive Service Use (defined as use of any of
  the following inpatient, residential, day
  treatment, partial hospitalization, in-home
  support, rehabilitation, therapeutic foster care,
  special school, crisis services, intensive case
  management, or use of outpatient services 3 or
  more days/week)
• Age 4-17 years at time of sampling
• Medicaid-eligible
• In managed care or fee-for-service plan at
  baseline interview
• Exclusion Criteria
• DSM-IV Diagnosis of solely MR, SA, or adjustment
  disorder
• Children with severe/profound MR/DD or those
  served primarily through the MR/DD system(s)

4
Study Methodology

• Parents and children were recruited through
  mailings to households containing children with
  SED being served through MC and FFS plans one
  site (OR) also used newspaper advertisements
• Response rates ranged from 10 to 98
• Consenting caregivers and children (age 11
  years) were interviewed at study baseline (T1)
  and six month followup (T2)
• Followup rate was 88 (N1517) there were no
  attrition differences re childs age, gender,
  functional impairment, health status,
  symptomatology, or caregiver strain only
  significant difference was in race/ethnicity.

5
The Adult Respondent
The most knowledgeable caretaker of the child,
including relatives (if available) and
professional caregivers (if not).
6
Managed Care Arrangements Variations at
Different Sites

• Who pays?
• For which services?
• For which children/adolescents?
• How is risk shifted?

7
Who Pays?
8
For Which Services?
9
For Whom?
10
How is Risk Shifted?
11
Research Questions Addressed Today

• Did psychiatric status, level of functional
  impairment, likelihood of mental health service
  utilization differ significantly between children
  in managed care vs. fee-for-service arrangements?
• Did satisfaction with the childs provider
  organization and behavioral health care plan
  differ significantly between caregivers of
  children in the two types of plans?
• Did caregivers ratings of provider service
  coordination differ for children in the two types
  of plans?

12
Description of the Sample
13
1st Research Question Childrens Statuses
Service Outcomes

• Does the psychiatric status, level of functional
  impairment, and likelihood of mental health
  service utilization differ significantly between
  children with SED served under managed care
  versus fee-for-service arrangements?

14
Dependent Variables

• Psychiatric Status (Child Behavior Checklist
  -CBCL)
• Functional Impairment (Columbia Impairment Scale
  - CIS)
• Service Utilization (Services Utilization
  Instrument - SUI)
• Inpatient/Residential
• Traditional Outpatient
• Psychotropic Medication
• Non-Traditional Services (i.e., day treatment,
  partial hospitalization, in-home treatment,
  school-based services, case management, or group
  home care)

15
Levels of Functional Impairment and Psychiatric
Symptomatology
CIS baseline 79 scored at or higher than the
clinical cutoff of 16. CBCL Total baseline
over 50 scored above the clinical mean,
indicating the presence of psychiatric symptoms
characteristics of children being treated for
mental health disorders
16
Proportion of Children Using Each Type of Service
between T1 T2
plt.05 plt.01 plt.001
17
Model Tested - Symptoms and Functioning

• Block 1 T1 Score for Dependent Variable (CIS
  or CBCL)
• Block 2 Child Characteristics (age, gender,
  minority status, juvenile justice involvement,
  health)
• Block 3 Caregiver Characteristics (education,
  gender, age, caregiver strain, physical health,
  mental health, satisfaction with behavioral
  health plan)
• Block 4 Household/Neighborhood Characteristics
  (income, number of co-residents, urban
  neighborhood, rural neighborhood)
• Block 5 Study Condition (managed care versus
  fee-for-service)
• Block 6 Site (TN/MS, OR, PA)

18
Model Tested - Service Utilization

• Block 1 Childs Need Variables (level of
  functional impairment, level of psychiatric
  symptomatology, substance use ever)
• Block 2 Child Characteristics (age, gender,
  minority status, juvenile justice involvement,
  health)
• Block 3 Caregiver Characteristics (education,
  gender, age, caregiver strain, physical health,
  mental health, satisfaction with behavioral
  health plan)
• Block 4 Household/Neighborhood Characteristics
  (income, number of co-residents, urban
  neighborhood, rural neighborhood)
• Block 5 Study Condition (managed care versus
  fee-for-service)
• Block 6 Site (TN/MS, OR, PA)

19
Results Symptoms, Functioning, Serice Use

• There were no significant differences in the
  functional status of children served in MC versus
  FFS arrangements
• There were no significant differences in the
  psychiatric status of children served in MC
  versus FFS arrangements, although a trend toward
  significance indicated somewhat poorer mental
  health status among children in the MC condition
• There were significant differences in the
  likelihood of some types of mental health service
  utilization but not others
• Children in MC arrangements were significantly
  less likely to receive inpatient/residential
  treatment
• Children in MC were significantly less likely to
  receive non-traditional mental health services
• There was a trend toward significance in which
  children in MC were somewhat less likely to
  receive psychopharmacologic treatment
• There was no significant difference in the
  likelihood of receiving traditional outpatient
  mental health services

20
2nd Research Question - Satisfaction

• Does caregiver satisfaction with the childs
  provider organization, and the childs behavioral
  health care plan differ significantly between
  children served under managed care versus
  fee-for-service arrangements?

21
Caregiver Satisfaction with Behavioral Health
Care Provider Agency

• Using any number on a scale from 0 to 10, where
  0 is the worst possible care and 10 is the best
  possible care, what is your overall rating of the
  care childs name has received from the agency
  providing the most hours of service in the past
  six months.
• MC FFS Total Group
• Average Score 7 7 7
• (difference non-significant)

22
Caregiver Satisfaction with Behavioral Health
Care Plan
Overall, what is your rating of health care
plan name now? Use any number on a scale from 0
to 10, where 0 is as bad as a health insurance
plan can be, 5 is okay or average, and 10 is as
good as a health insurance plan can
be. MC FFS Total Group Average Score
7 8 7.5 p lt.001, difference remains
significant controlling for site
23
Proportion Reporting Different Types of Provider
Agency Satisfaction and Relationship to 0-10
Rating

• Usually/Always
• Got appointment promptly 80
• Would recommend agency 83
• Agency explained things well 86
• Agency listed carefully 85
• Agency aware of services 87
• Involved caregiver in decisions 79
• Caregiver treated with respect 91
• Significant relationship with 0-10 Provider
  Agency rating p lt.05

24
Proportion Reporting Different Types of Health
Care Plan Satisfaction/Dissatisfaction
Relationship to 0-10 Satisfaction Rating
Significant relationship with 0-10 Provider
Agency Rating, plt.05
25
Model Tested - Provider/Plan Satisfaction

• Block 1 Child Characteristics (age, gender,
  minority status, juvenile justice involvement,
  health)
• Block 2 Caregiver Characteristics (education,
  gender, age, caregiver strain, physical health,
  mental health)
• Block 3 Household/Neighborhood Characteristics
  (income, number of co-residents, urban
  neighborhood, rural neighborhood)
• Block 4 Childs Behavioral Health Need
  Variables (level of psychiatric symptomatology,
  level of functional impairment)
• Block 5 Childs Service Utilization
  (inpatient/residential treatment, outpatient
  treatment, psychotropic medication,
  nontraditional services)
• Block 6 Study Condition (managed care versus
  fee-for-service)
• Block 7 Site (TN/MS, OR, PA, OH)

26
Results Provider Plan Satisfaction

• There were no significant differences in level of
  satisfaction with the childs provider agency (as
  rated by adult caregivers) between children
  served in managed care versus fee-for-service
  arrangements.
• Satisfaction with the childs behavioral health
  care plan was significantly lower among
  caregivers whose children were enrolled in
  managed care versus fee-for-service plans. This
  was rue even controlling for characteristics of
  the child, caregiver, household/ neighborhood,
  childs level of need, recent service
  utilization, and study site.

27
3RD Research Question Service Coordination

• Does the caregivers rating of degree of
  service coordination vary by whether the child
  was enrolled in a managed care plan versus a
  fee-for-service plan?

28
Service Coordination Scale (SCC)

• A set of 9 Likert-scaled responses to items
  asking caregivers about the degree to which the
  childs service providers communicate
  coordinate their service delivery efforts
• Administered to 266 caregivers of children
  adolescents with SED, the scale had good
  psychometrics (high internal consistency, good
  construct validity with measures of satisfaction
  and family participation)
• (Koren, Paulson, Kinney et al., 1997)

29
Degree of Service Coordination Among Providers as
Assessed by Caregivers
30
Model Tested - Service Coordination

• Block 1 Child Characteristics (age, gender,
  minority status)
• Block 2 Caregiver Characteristics (caregiver
  education, caregiver gender)
• Block 3 Caregiver Stressors (level of caregiver
  strain, caregiver health, caregiver
  depression)
• Block 4 Child Need (childs mental health
  symptoms)
• Block 5 Site (TN/MS, OR, PA)
• Block 6 Study Condition (MC vs. FFS)

31
Results Service Coordination

• Most caregivers are fairly satisfied with the
  degree of service coordination occurring on
  behalf of children and youth with SED.
• As perceived by their caregivers, children in MC
  behavioral health plans experience lower levels
  of service coordination than do children in FFS
  plans.
• This difference remained significant in
  multivariate models, even controlling for study
  site, caregiver strain, and caregiver physical
  health. Other significant predictors of service
  coordination include caregivers education,
  caregivers level of depression, and severity of
  childs psychiatric symptoms.

32
Conclusions

• While there were no differences between the
  functional status psychiatric symptom severity
  of children enrolled in MC vs. FFS plans, there
  was significantly lower utilization of some
  mental health services.
• There was lower satisfaction with the childs
  behavioral health care plan among caregivers of
  children in MC arrangements compared to FFS.
• There was significantly lower service
  coordination among providers of children served
  in MC vs. FFS plans.

33
For further information

• Visit the website
• www.psych.uic.edu/mhsrp
• study description
• downloadable protocols
• research presentations
• link to larger study