VAL ALUE UE Implementing Behavioral Health Homes to Improve - - PowerPoint PPT Presentation

INNOVATION

VAL ALUE UE

June 25, 2017

Cara Nikolajski, MPH; James Schuster, MD, MBA; Patricia Schake, MSW, LSW; Tracy Carney; Charles Reynolds, MD; Chaeryon Kang, PhD; Jane Kogan, PhD; Michele Mesiano, MSW

Implementing Behavioral Health Homes to Improve Patient- Centered Outcomes for Adults with Serious Mental Illness

• The role of a payer in supporting changes in care delivery

across a network of community mental health centers

• Overview of two system-level models of care aimed at

improving the health of adults with serious mental illness (SMI) who are at high risk for chronic disease

• Patient-centered comparative effectiveness research (PC-

CER) study conducted to evaluate the impact of the two models on patient-centered outcomes

• Next steps: analyses and dissemination

Presentation overview

2

• UPMC Center for High-

Value Health Care

– Non-profit research division of the UPMC ISD – Evaluates and translates UPMC’s work into evidence- based practice and policy change – Work supported through grants/contracts and conducted in partnership with patients and family members, community organizations, researchers, and government agencies

UPMC Insurance Services Division (ISD) & UPMC Center for High-Value Health Care (Center)

3

• Largest nonprofit behavioral health managed care organization in the

country and largest insurance provider of Pennsylvania Medicaid beneficiaries for behavioral health services

• Manages behavioral health services for nearly 1 million members in 39
• f 67 PA counties
• Works closely with their large provider network including community

mental health centers (CMHC) to deliver comprehensive, recovery-

• riented, services

Community Care Behavioral Health

4

• Resource and care delivery limitations due to financial-,

personnel-, and policy-related barriers

• Often a frequent or the only point of contact with healthcare

system for vulnerable individuals with serious mental illness (SMI)

• Individuals with SMI experience or are at risk for a range of

challenges, including physical health comorbidities, making behavioral/physical healthcare integration critically important within the CMHC setting Challenges faced by community mental health centers

5

• Impacts approximately 4% of the U.S. adult population (9.8 million

individuals) annually

• Decreased life expectancy of up to 25 years compared to general

population

• Physical comorbidities greatly contribute to this disparity

– Increased rates of cardiovascular disease and metabolic syndrome compared to general population

• Key contributors to morbidity/mortality

– Heavy tobacco use – Metabolic effects of atypical antipsychotic medications – Access to care – Untreated health conditions – Poor diet – Sedentary lifestyle

Morbidity and mortality among individuals with SMI

6

References: NIMH, 2015: https://www.nimh.nih.gov/health/statistics/prevalence/serious-mental-illness-smi-among-us-adults.shtml Parks, 2006 https://www.nasmhpd.org/sites/default/files/Mortality%20and%20Morbidity%20Final%20Report%208.18.08.pdf

• Community Care, providers, and other stakeholders developed behavioral

health home (BHH) model in 2010 with a focus on:

– Enhancing capacity of behavioral health providers to serve as health homes – Comprehensive care management – Care coordination and health promotion – Linkage of service-users to community resources

• To promote model scaling, needed to understand how a less resource

intensive BHH focusing on disease self-management resources compared to a nurse-supported BHH with a more formalized consultation and care coordination focus

– Provided the basis for our CER study supported with funding from the Patient-Centered Outcomes Research Institute (PCORI)

• Stakeholder input obtained to inform all elements of the research process

(research questions, study design, implementation, analysis, dissemination)

Creating behavioral health home models to support integrated care delivery in CMHCs

7

Using CER to Examine Impact of the Behavioral Health Home Models

8

• Cluster-randomized design with mixed methods approach
• Models implemented in 11 community mental health centers

(CMHC) over 2 years starting in 2013

• Research participant inclusion criteria:

– Medicaid enrolled – 21+ years of age – Diagnosed with a serious mental illness – Receives services at community mental health center within Community Care’s network

• Institute for Healthcare Improvement’s Learning

Collaborative Model used to support implementation Study Methods and Design

9

Reference: Institute for Healthcare Improvement Breakthrough Series: http://www.ihi.org

Enrollment in CER Study

10

11 provider sites randomized

Provider- Supported Care (5 sites) Self-Directed Care (6 sites) Eligible: n=632 Site 1: 97 (15.3%) Site 2: 127 (20.1%) Site 3: 91 (14.4%) Site 4: 67 (10.6%) Site 5: 134 (21.2%) Site 6: 116 (18.4%) Eligible: n=811 Site 1: 87 (10.7%) Site 2: 135 (16.6%) Site 3: 49 (6.0%) Site 4: 350 (43.2%) Site 5: 190 (23.4%) Enrolled: n=516 Site 1: 55 (57.6%) Site 2: 112 (88.2%) Site 3: 78 (85.7%) Site 4: 40 (59.7%) Site 5: 133 (99.3%) Site 6: 98 (84.5%) Enrolled: n=713 Site 1: 83 (95.4%) Site 2: 114 (84.4%) Site 3: 27 (55.1%) Site 4: 313 (89.4%) Site 5: 176 (92.6%)

Who participated in the research?

11

Characteristic Provider-Supported Self-Directed Total N % N % N % Total 713 58.0% 516 42.0% 1,229 100% Age (mean/range) 43.47 19-72 42.37 18-76 43.01 18-76 Gender Female Male 428 285 60.0% 40.0% 341 175 66.1% 33.9% 769 460 62.6% 37.4% Race White Black Other 622 72 19 87.2% 10.1% 2.7% 487 21 8 94.4% 4.1% 1.6% 1104 93 27 90.2% 7.6%% 2.2% Ethnicity Non-Hispanic Hispanic 710 3 99.6% 0.4% 512 4 99.2% 0.9% 1222 7 99.4% 0.6% Diagnosis MDD Bipolar Schizoaffective Schizophrenia Other None 227 193 131 86 67 9 31.8% 27.1% 18.4% 12.1% 9.4% 1.3% 234 137 64 40 31 10 45.3% 26.6% 12.4% 7.8% 6.0% 1.9% 461 330 195 126 98 19 37.5% 26.9% 15.9% 10.3% 8.0% 1.5%

Patient-centered outcomes & data sources

12

Data collection & completion

13

11 provider sites randomized

Provider- Supported Care (5 sites) Self-Directed Care (6 sites) Eligible: n=632 Eligible: n=811 Enrolled: n=516 Enrolled: n=713

Data Collection Loss to Follow-Up (Cumulative) TP1: self-report n=514; claims n=441 N/A TP2: self-report n=337; claims n=456 15 (2.9%) TP3: self-report n=282; claims n=448 41 (7.9%) TP4: self-report n=269; claims n=420 71 (13.8%) TP5: self-report n=220; claims n=310 88 (17.1%) Data Collection Loss to Follow-Up (Cumulative) TP1: self-report n=713; claims n=649 N/A TP2: self-report n=552; claims n=652 23 (3.2%) TP3: self-report n=489; claims n=619 49 (6.9%) TP4: self-report n=437; claims n=601 73 (10.2%) TP5: self-report n=386; claims n=582 102 (14.3%)

Overview of study findings

14

• Why patient activation?

– Measures an individuals level of engagement in their own health care using Patient Activation Measure – A single point increase in PAM score correlates to a 2% decrease in hospitalization and 2% increase in medication adherence

• Our findings:

– Provider-Supported led to more immediate and stable improvement in patient activation (significant treatment X time interaction observed; p<0.0001) – Subgroup analysis revealed that male gender associated with a greater improvement in Self-Directed arm and female gender associated with faster and greater improvement in patient activation in the Provider-Supported arm

Patient activation improved in both study arms but at different times and differed by gender

15

Reference: Insignia Health: http://www.insigniahealth.com/

• Why health status?

– SF-12 used to measure perceived physical and mental health status – Even a small score change can impact mortality rates and other health-related factors

• Our findings:

– Mental health status score increased, particularly at month 6 (p<0.0001) – Physical health status score decreased over time, particularly after month 12 (p<0.0001)

Perceived mental health status improved and perceived physical health status declined

16

Reference: Bjorner, 2013

• Why engagement in primary/specialty care?

– Annual mean # of primary/specialty care visits measured using claims data – Increased rates of primary/specialty care utilization can lead to increased receipt of preventive and treatment measures

• Our finding:

– While the two interventions did not differ significantly in their impact on this outcome, both showed improvement

• ver time (p<0.0001)

Engagement in primary/specialty care increased significantly in both arms

17

References: Druss, 2001; Banta, 2009

• Shift in definition of health and wellness, away from vague to more personalized
• Increased awareness of interconnectedness of mental and physical health
• Overall favorable intervention experiences
• No major distinctions between arms – no evident differences in engagement in or satisfaction

with interventions

• Most important factor leading to intervention participation was relationship with wellness coach

Qualitative findings: Service user perspective

18 “I’ve actually exceeded my goal…the weight I am now, I haven’t been since I was a young teenager. I lost 25 pounds in the beginning and I’ve actually lost closer to 45. I feel like a have a lot more energy.” “I’m eating a lot more fruits and vegetables…My case manager knows that I’m trying to watch my calories and she’ll ask me how the calorie counting is going. She encourages me.” “I didn’t realize there’s a lot of stuff that had sodium. I mean I was really surprised about the amount of sodium in foods; never really paid much

• attention. Also [I learned

about] nutritional values in foods and food that I have to avoid for my medications. So it was a big help.”

• High degree of agency support for wellness coaching
• Culture of wellness that benefitted both service users and providers
• Models integrated into routine practice
• Robustly positive impact on service user’s health/wellness
• Acute needs sometimes trumped wellness coaching
• Structural barriers (e.g. healthcare access, community resources, transportation) can limit

success

Qualitative findings: Provider perspective

19 “We’re putting in the works…and trying to make a dynamic change; a culture shift with the group that we’re working with.” “He went from three packs a week to one pack a month, and he did that within six month.” “You know, when we do break it down and they do start to feel good about themselves; like when we can get them to break down smaller, and they do see like 'Oh, I lost that five

• pounds. That was pretty easy. I want

to lose five more or ten more.’ When they start to build up their confidence and see that they can do it, then I think it's successful, for sure.”

• Additional analyses:

– Identify comparison group to assess additional cost, utilization, and quality outcomes. Preliminary analyses suggest:

• Lower total spending and decreased physical and behavioral health inpatient utilization in

Provider-Supported vs Self-Directed – Analysis of Provider-Support/Self-Directed vs post-trial comparison group yielded similar findings

• Dissemination:

– Implementation resource website underway – Manuscripts & conference presentations – Behavioral Health Home Plus (BHHP) implemented in additional community mental health centers – PCORI proposal submitted to scale (BHHP) implementation in adolescent residential treatment facilities and opioid treatment facilities

Next steps

20

• Contact Information:

– Cara Nikolajski, MPH – nikolajskice@upmc.edu

• UPMC Center for High-Value Health Care

– http://www.upmchighvaluehealthcare.com/

Thank You

21

All quantitative outcomes at a glance

22 Outcome Significant Treatment X Time Interaction Significant Change Over Time Gender as a Moderator No Significant Findings

PRIMARY OUTCOMES Patient Activation X X Health Status – Mental X X Health Status – Physical X Engagement in Primary/Specialty Care X SECONDARY OUTCOMES Hope X Quality of Life X Emergent Care Use X Lab Monitoring* X (total services; glucose) X (lipids) X (EKG) Medication Adherence* X (antidepressants; hypertension; diabetes) X (antipsychotics) Satisfaction with Care X Functional Status X