Collaborative care models for management of mental and
behavioral health concerns in Federally Qualified Health Centers
1Bethany M. Kwan, 1Benjamin F. Miller, 1,2Marion R. Sills, 1Mika K. Hamer, 1Lisa M. Schilling, & the SAFTINet Consortium
1University of Colorado School of Medicine, 2Children’s Hospital Colorado
Background & Objectives
Collaborative care (also known as integrated behavioral health in primary care) is a model of
care in which primary care medical clinicians and mental and behavioral health clinicians (e.g.,
psychiatrists, psychologists, social workers) work together with patients to identify and treat
mental and behavioral health concerns, from a “whole person” perspective.
Although models vary, basic elements of Collaborative Care (CC) include: (Peek, 2011)
• A collaborative team comprised of mental/behavioral health and medical clinicians
• Protocols for identifying, triaging, treating, and tracking mental health concerns from within primary care
• Supporting information technology infrastructure
Dimensions on which CC models may vary: (Miller, Mendenhall, & Malik, 2009)
• Type of collaboration between medical and behavioral health clinicians
• Type of spatial arrangement in which behavioral health clinicians interact with patients and
medical clinicians
• The range of behavioral health concerns addressed in primary care settings
Collaborative care is an effective model for improving access to behavioral health services and
health outcomes, especially in the context of depression and anxiety (Butler et al., 2008).
There is a range of mental & behavioral health (MBH) concerns for which behavioral health
providers (BHPs) can be useful in primary care, such as counseling for health behavior change
and psychosocial issues (Miller et al., under review). There is limited evidence on translation
and implementation of CC models in the real world, especially in low-resource settings (Kwan &
Nease, 2013).
Research Objective:
• To explore the variety of CC models used by Federally Qualified Health Centers (FQHCs) to
provide services for a range of mental and behavioral health (MBH) concerns of primary care
patients.
Research Questions:
1. What CC models do FQHCs use (type of collaboration and spatial arrangement: referral-
based, co-location, full integration) and for which MBH concerns (range of functions and
expertise: health behavior change, psychosocial concerns, mental health, serious mental
illness)?
2. How many and what types of MBH providers work in FQHCs? How many full time equivalents
(FTEs) are employed given practice size?
3. In what practice settings and patient populations are various CC models used?
Methods
Setting: SAFTINet (Scalable Architecture for
Federated Translational Inquiries Network)
• Practice-based research network (PBRN) of
Federally Qualified Health Centers (FQHCs)
and FQHC look-alikes
• 100 member practices in Colorado,
Tennessee, Vermont, and California
• Founded in 2010, in part based on a
collective interest in understanding the
impact of models of care such as the Patient
Centered Medical Home (PCMH) and
Collaborative Care
• Using practice-level surveys, we characterized member practices’ existing models of collaborative care in
terms of staffing, service delivery mechanisms, and the range of mental and behavioral health concerns
systematically addressed within the practice
Measures: Representatives from 47 member FQHCs and FQHC look-alikes from 4 health care organizations in
Colorado and Tennessee completed practice-level surveys in summer 2013
Practice Characteristics
• Designation as FQHC
• Specialties (Adult internal medicine, family medicine, pediatrics, geriatrics)
• Total # of unique patients and patient encounters/year
• Demographics of patient population (ages, gender, race/ethnicity, preferred language, and payer mix)
• Staffing: # and full-time equivalents (FTE) of MBH providers
• LCSW/MSW: master’s level mental health clinicians
• PsyD/PhD or doctoral candidates
• PCMH recognition status (currently recognized, pursuing recognition, not pursuing recognition)
• Participation in reimbursement programs (e.g., Bundled payments, pay for performance)
Results
Distribution of CC models within practices
Mechanism(s) by which any MBH concern is addressed within primary care:
Distribution of CC models by MBH concern
For each MBH concern, to what extent do practices offer services using a systematic
mechanism, and with what CC model(s)?
CC models by practice characteristics
• Practices with family medicine providers were more likely to have full integration (26/34)
than practices without family medicine (7/13), although the difference was not statistically
significant (OR = 2.79, p = .14), likely due to small sample sizes.
• Practices with full integration care for more unique pts than practices without full
integration (M = 6168 vs 3744; p = .08)
Staffing of CC models in FQHCs
• On average, practices with on-site providers employed 1.35 FTEs (SD = 2.1, median = 0.50;
range 0-8.5) of a master’s level provider (LCSW/MSW) and 1.25 FTEs (SD = 2.6, median =
0.50, range 0-14.7) of a doctoral level psychologist (PsyD/PhD) per 5000 unique patients.
• Those with fully integrated models reported a median of 0.58 FTEs for LCSW/MSW and 0.52
FTEs for PsyD/Phd, compared to .16 and .25, respectively, for those without full integration.
Conclusions
• A significant proportion of SAFTINet practices are engaged in integrated care
• It is feasible for FQHCs to provide on-site MBH services for a range of mental and behavioral
health concerns. These models are staffed by both masters and doctoral level MBH providers.
• There remains variability in how practices address mental health
• Multiple CC models can be used within a practice across the range of MBH concerns
• There appear to be opportunities to better measure and test the effects of integration
through networks like SAFTINet.
Policy Implications
• While SAFTINet practices may not be typical of FQHCs in the U.S., it appears collaborative care is gaining
ground as a valued model for addressing behavioral concerns in primary care.
• Meeting the needs of front line primary care practices will require a better understanding of the workforce
skills and training required (Blount & Miller, 2009; Burke et al., 2013) and the best ways to allocate resources
and pay for these services (Kathol, DeGruy, & Rollman, 2014).
• In order to make a stronger policy case for integration, we must have more consistency in how we define,
evaluate, and advocate for CC models
Future Plans
SAFTINet partners plan to conduct comparative effectiveness research on the effects of collaborative care
models on health outcomes for patients with multiple chronic conditions. We also plan to investigate the role of
payment models and reimbursement structures on implementation and sustainability of CC models. Ultimately,
we will disseminate our findings and “best practices” to support more widespread implementation of the model.
Health behavior
change
Psychosocial Mental health Serious mental
illness or
substance abuse
Any MBH
concern
Referral 6 (12.8%) 6 (12.8%) 7 (14.9%) 19 (40.4%) 19 (40.4%)
Co-location 12 (25.5%) 15 (31.9%) 17 (36.2%) 14 (29.8%) 17 (36.2%)
Full integration 33 (70.2%) 33 (70.2%) 33 (70.2%) 26 (55.3%) 33 (70.2%)
No systematic
mechanism
1 (2.1%) 1 (2.1%) 1 (2.1%) 3 (6.4%) 3 (6.4%)
Table 1. #/% Practices using various CC models to address range of MBH concerns
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
ORG1_1
ORG3_9
ORG3_7
ORG2_11
ORG2_12
ORG2_17
ORG2_20
ORG2_4
ORG2_6
ORG2_9
ORG3_1
ORG3_2
ORG3_4
ORG3_6
ORG1_10
ORG1_9
ORG2_13
ORG2_18
ORG1_2
ORG1_4
ORG1_5
ORG1_6
ORG1_8
ORG2_1
ORG2_10
ORG2_16
ORG2_5
ORG2_7
ORG3_5
ORG3_1.1
ORG3_3
ORG1_11
ORG1_7
ORG3_8
ORG3_9.1
ORG4_1
ORG4_2
ORG4_4
ORG4_5
ORG4_6
ORG4_7
ORG4_8
ORG4_9
ORG2_2
ORG2_3
ORG3_10
ORG4_3
MixofCCmodelswithinpractice
none
full
colocate
refer
• Most practices reported some form of collaborative care for
providing care to patients across the full range of MBH concerns.
• 70% of practices (33/47) reported the use of a fully integrated
model (on-site BHPs with shared clinic space with medical
providers) to address health behavior, psychosocial, and/or
mental health concerns.
• 55% (26/47) reported using a fully integrated model for SMI/SA,
while 40% (19/47) reported referral to off-site providers for
SMI/SA.
% of practices offering services for
HEALTH BEHAVIOR
Referral
Co-Location
Full Integration
R+C
R+F
C+F
R+C+F
No system
% of practices offering services for
PSYCHOSOCIAL CONCERNS
Referral
Co-Location
Full Integration
R+C
R+F
C+F
R+C+F
No system
% of practices offering services for
MENTAL HEALTH
Referral
Co-Location
Full Integration
R+C
R+F
C+F
R+C+F
No system
% of practices offering services for
SMI/SA
Referral
Co-Location
Full Integration
R+C
R+F
C+F
R+C+F
No system
Measures continued…
Mental & Behavioral Health (MBH) Services
• By what mechanisms (if any) do this practice’s patients receive MBH services?
• Referral-based model: Referral to off-site MBH providers
• Co-location model: Co-located MBH providers, who spend some but not all their time in the same clinic
space as medical providers
• Full integration model: On-site behavioral health providers, who share clinic space with medical
providers
• No CC model: No systematic mechanism for delivering MBH services
• For each existing mechanism, for which types of MBH concerns are those services available?
• Health behavior change counseling (e.g., diet, exercise, smoking cessation)
• Counseling for psychosocial/social determinants of health issues (e.g., domestic violence, social
support)
• Diagnosis and/or treatment of mental health conditions (e.g., depression, anxiety)
• Diagnosis and/or treatment of serious mental illness or substance abuse (SMI/SA)
• Note: Practices could report more than one mechanism overall and for each type of MBH concern
• Information exchanged periodically with minimally shared care plans or workflows
Referral-triggered periodic exchange
• Regular communication and coordination, usually via separate systems and workflows, but with care plans
coordinated to a significant extent
Regular communication/coordination
• Fully shared treatment plans and documentation, regular communication facilitated and/or clinical
workflows that ensure effective communication and coordination
Full collaboration/integration
Separate
space
Behavioral health and
medical clinicians spend
little time with each
other practicing in same
clinic space.
Patient has to see
providers in at least two
buildings.
Co-located
space
Behavioral health and
medical clinicians in
different parts of the
same building, spending
some but not all their
time in same medical
clinic space.
Patient typically has to
move from primary care
to behavioral health.
space
Fullyshared
space
Behavioral health and medical
clinicians share the same
provider rooms, spending all or
most of their time seeing
patients in that shared space.
Typically, both clinicians see
the patient in same exam
room.
• Identifying and
treating health
behavior needs
• Health behavior
activation & self-
management support
• E.g., Diet, physical
activity, smoking
cessation
Basic health
behavior
• Identifying social
barriers for patients
in receiving their
health care
• Connecting patients
with community
resources
• Providing social and
community support
for patients
Psychosocial
concerns
• Screening, assessment,
and management of
acute mental health
problems, with triage,
referral and
consultation when
necessary
• Appropriate
psychological and
pharmaceutical
interventions
• E.g., Depression,
anxiety
Mental health
• Screening,
assessment, and
management of
acute SMI/SA
problem, with triage,
referral and
consultation when
necessary
• Appropriate
psychological and
pharmaceutical
interventions
Serious mental
illness & substance
abuse (SMI/SA)
*CC models and MBH concern categories are not mutually exclusive
Acknowledgments
We would like to thank the entire SAFTINet Consortium for their contributions to this work. We would especially like to
acknowledge the important contributions of current and past leadership at network partner sites, including Parinda Khatri, Jena Saporito,
Andrea Auxier, Arthur Davidson, Jeanne Rozwadowski, Maria DeJesus Diaz-Perez, and Alicyn Kaiser.
Funding provided by AHRQ 1R01HS019908 (Scalable Architecture for Federated Translational Inquiries Network) and AHRQ R01HS022956
(SAFTINet: Optimizing Value and Achieving Sustainability)
References available upon request

Collaborative care models for management of mental and behavioral health concerns in Federally Qualified Health Centers

  • 1.
    Collaborative care modelsfor management of mental and behavioral health concerns in Federally Qualified Health Centers 1Bethany M. Kwan, 1Benjamin F. Miller, 1,2Marion R. Sills, 1Mika K. Hamer, 1Lisa M. Schilling, & the SAFTINet Consortium 1University of Colorado School of Medicine, 2Children’s Hospital Colorado Background & Objectives Collaborative care (also known as integrated behavioral health in primary care) is a model of care in which primary care medical clinicians and mental and behavioral health clinicians (e.g., psychiatrists, psychologists, social workers) work together with patients to identify and treat mental and behavioral health concerns, from a “whole person” perspective. Although models vary, basic elements of Collaborative Care (CC) include: (Peek, 2011) • A collaborative team comprised of mental/behavioral health and medical clinicians • Protocols for identifying, triaging, treating, and tracking mental health concerns from within primary care • Supporting information technology infrastructure Dimensions on which CC models may vary: (Miller, Mendenhall, & Malik, 2009) • Type of collaboration between medical and behavioral health clinicians • Type of spatial arrangement in which behavioral health clinicians interact with patients and medical clinicians • The range of behavioral health concerns addressed in primary care settings Collaborative care is an effective model for improving access to behavioral health services and health outcomes, especially in the context of depression and anxiety (Butler et al., 2008). There is a range of mental & behavioral health (MBH) concerns for which behavioral health providers (BHPs) can be useful in primary care, such as counseling for health behavior change and psychosocial issues (Miller et al., under review). There is limited evidence on translation and implementation of CC models in the real world, especially in low-resource settings (Kwan & Nease, 2013). Research Objective: • To explore the variety of CC models used by Federally Qualified Health Centers (FQHCs) to provide services for a range of mental and behavioral health (MBH) concerns of primary care patients. Research Questions: 1. What CC models do FQHCs use (type of collaboration and spatial arrangement: referral- based, co-location, full integration) and for which MBH concerns (range of functions and expertise: health behavior change, psychosocial concerns, mental health, serious mental illness)? 2. How many and what types of MBH providers work in FQHCs? How many full time equivalents (FTEs) are employed given practice size? 3. In what practice settings and patient populations are various CC models used? Methods Setting: SAFTINet (Scalable Architecture for Federated Translational Inquiries Network) • Practice-based research network (PBRN) of Federally Qualified Health Centers (FQHCs) and FQHC look-alikes • 100 member practices in Colorado, Tennessee, Vermont, and California • Founded in 2010, in part based on a collective interest in understanding the impact of models of care such as the Patient Centered Medical Home (PCMH) and Collaborative Care • Using practice-level surveys, we characterized member practices’ existing models of collaborative care in terms of staffing, service delivery mechanisms, and the range of mental and behavioral health concerns systematically addressed within the practice Measures: Representatives from 47 member FQHCs and FQHC look-alikes from 4 health care organizations in Colorado and Tennessee completed practice-level surveys in summer 2013 Practice Characteristics • Designation as FQHC • Specialties (Adult internal medicine, family medicine, pediatrics, geriatrics) • Total # of unique patients and patient encounters/year • Demographics of patient population (ages, gender, race/ethnicity, preferred language, and payer mix) • Staffing: # and full-time equivalents (FTE) of MBH providers • LCSW/MSW: master’s level mental health clinicians • PsyD/PhD or doctoral candidates • PCMH recognition status (currently recognized, pursuing recognition, not pursuing recognition) • Participation in reimbursement programs (e.g., Bundled payments, pay for performance) Results Distribution of CC models within practices Mechanism(s) by which any MBH concern is addressed within primary care: Distribution of CC models by MBH concern For each MBH concern, to what extent do practices offer services using a systematic mechanism, and with what CC model(s)? CC models by practice characteristics • Practices with family medicine providers were more likely to have full integration (26/34) than practices without family medicine (7/13), although the difference was not statistically significant (OR = 2.79, p = .14), likely due to small sample sizes. • Practices with full integration care for more unique pts than practices without full integration (M = 6168 vs 3744; p = .08) Staffing of CC models in FQHCs • On average, practices with on-site providers employed 1.35 FTEs (SD = 2.1, median = 0.50; range 0-8.5) of a master’s level provider (LCSW/MSW) and 1.25 FTEs (SD = 2.6, median = 0.50, range 0-14.7) of a doctoral level psychologist (PsyD/PhD) per 5000 unique patients. • Those with fully integrated models reported a median of 0.58 FTEs for LCSW/MSW and 0.52 FTEs for PsyD/Phd, compared to .16 and .25, respectively, for those without full integration. Conclusions • A significant proportion of SAFTINet practices are engaged in integrated care • It is feasible for FQHCs to provide on-site MBH services for a range of mental and behavioral health concerns. These models are staffed by both masters and doctoral level MBH providers. • There remains variability in how practices address mental health • Multiple CC models can be used within a practice across the range of MBH concerns • There appear to be opportunities to better measure and test the effects of integration through networks like SAFTINet. Policy Implications • While SAFTINet practices may not be typical of FQHCs in the U.S., it appears collaborative care is gaining ground as a valued model for addressing behavioral concerns in primary care. • Meeting the needs of front line primary care practices will require a better understanding of the workforce skills and training required (Blount & Miller, 2009; Burke et al., 2013) and the best ways to allocate resources and pay for these services (Kathol, DeGruy, & Rollman, 2014). • In order to make a stronger policy case for integration, we must have more consistency in how we define, evaluate, and advocate for CC models Future Plans SAFTINet partners plan to conduct comparative effectiveness research on the effects of collaborative care models on health outcomes for patients with multiple chronic conditions. We also plan to investigate the role of payment models and reimbursement structures on implementation and sustainability of CC models. Ultimately, we will disseminate our findings and “best practices” to support more widespread implementation of the model. Health behavior change Psychosocial Mental health Serious mental illness or substance abuse Any MBH concern Referral 6 (12.8%) 6 (12.8%) 7 (14.9%) 19 (40.4%) 19 (40.4%) Co-location 12 (25.5%) 15 (31.9%) 17 (36.2%) 14 (29.8%) 17 (36.2%) Full integration 33 (70.2%) 33 (70.2%) 33 (70.2%) 26 (55.3%) 33 (70.2%) No systematic mechanism 1 (2.1%) 1 (2.1%) 1 (2.1%) 3 (6.4%) 3 (6.4%) Table 1. #/% Practices using various CC models to address range of MBH concerns 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% ORG1_1 ORG3_9 ORG3_7 ORG2_11 ORG2_12 ORG2_17 ORG2_20 ORG2_4 ORG2_6 ORG2_9 ORG3_1 ORG3_2 ORG3_4 ORG3_6 ORG1_10 ORG1_9 ORG2_13 ORG2_18 ORG1_2 ORG1_4 ORG1_5 ORG1_6 ORG1_8 ORG2_1 ORG2_10 ORG2_16 ORG2_5 ORG2_7 ORG3_5 ORG3_1.1 ORG3_3 ORG1_11 ORG1_7 ORG3_8 ORG3_9.1 ORG4_1 ORG4_2 ORG4_4 ORG4_5 ORG4_6 ORG4_7 ORG4_8 ORG4_9 ORG2_2 ORG2_3 ORG3_10 ORG4_3 MixofCCmodelswithinpractice none full colocate refer • Most practices reported some form of collaborative care for providing care to patients across the full range of MBH concerns. • 70% of practices (33/47) reported the use of a fully integrated model (on-site BHPs with shared clinic space with medical providers) to address health behavior, psychosocial, and/or mental health concerns. • 55% (26/47) reported using a fully integrated model for SMI/SA, while 40% (19/47) reported referral to off-site providers for SMI/SA. % of practices offering services for HEALTH BEHAVIOR Referral Co-Location Full Integration R+C R+F C+F R+C+F No system % of practices offering services for PSYCHOSOCIAL CONCERNS Referral Co-Location Full Integration R+C R+F C+F R+C+F No system % of practices offering services for MENTAL HEALTH Referral Co-Location Full Integration R+C R+F C+F R+C+F No system % of practices offering services for SMI/SA Referral Co-Location Full Integration R+C R+F C+F R+C+F No system Measures continued… Mental & Behavioral Health (MBH) Services • By what mechanisms (if any) do this practice’s patients receive MBH services? • Referral-based model: Referral to off-site MBH providers • Co-location model: Co-located MBH providers, who spend some but not all their time in the same clinic space as medical providers • Full integration model: On-site behavioral health providers, who share clinic space with medical providers • No CC model: No systematic mechanism for delivering MBH services • For each existing mechanism, for which types of MBH concerns are those services available? • Health behavior change counseling (e.g., diet, exercise, smoking cessation) • Counseling for psychosocial/social determinants of health issues (e.g., domestic violence, social support) • Diagnosis and/or treatment of mental health conditions (e.g., depression, anxiety) • Diagnosis and/or treatment of serious mental illness or substance abuse (SMI/SA) • Note: Practices could report more than one mechanism overall and for each type of MBH concern • Information exchanged periodically with minimally shared care plans or workflows Referral-triggered periodic exchange • Regular communication and coordination, usually via separate systems and workflows, but with care plans coordinated to a significant extent Regular communication/coordination • Fully shared treatment plans and documentation, regular communication facilitated and/or clinical workflows that ensure effective communication and coordination Full collaboration/integration Separate space Behavioral health and medical clinicians spend little time with each other practicing in same clinic space. Patient has to see providers in at least two buildings. Co-located space Behavioral health and medical clinicians in different parts of the same building, spending some but not all their time in same medical clinic space. Patient typically has to move from primary care to behavioral health. space Fullyshared space Behavioral health and medical clinicians share the same provider rooms, spending all or most of their time seeing patients in that shared space. Typically, both clinicians see the patient in same exam room. • Identifying and treating health behavior needs • Health behavior activation & self- management support • E.g., Diet, physical activity, smoking cessation Basic health behavior • Identifying social barriers for patients in receiving their health care • Connecting patients with community resources • Providing social and community support for patients Psychosocial concerns • Screening, assessment, and management of acute mental health problems, with triage, referral and consultation when necessary • Appropriate psychological and pharmaceutical interventions • E.g., Depression, anxiety Mental health • Screening, assessment, and management of acute SMI/SA problem, with triage, referral and consultation when necessary • Appropriate psychological and pharmaceutical interventions Serious mental illness & substance abuse (SMI/SA) *CC models and MBH concern categories are not mutually exclusive Acknowledgments We would like to thank the entire SAFTINet Consortium for their contributions to this work. We would especially like to acknowledge the important contributions of current and past leadership at network partner sites, including Parinda Khatri, Jena Saporito, Andrea Auxier, Arthur Davidson, Jeanne Rozwadowski, Maria DeJesus Diaz-Perez, and Alicyn Kaiser. Funding provided by AHRQ 1R01HS019908 (Scalable Architecture for Federated Translational Inquiries Network) and AHRQ R01HS022956 (SAFTINet: Optimizing Value and Achieving Sustainability) References available upon request