Integration of Primary Care and Behavioral Health Services in Midwestern Community Health Centers: A Mixed Methods Study

Abstract

Introduction:

Integrating behavioral health (BH) and primary care is an important strategy to improve health behaviors, mental health, and substance misuse, particularly at community health centers (CHCs) where disease burden is high and access to mental health services is low. Components of different integrated BH models are often combined in practice. It is unknown which components distinguish developing vs. established integrated BH programs.

Method:

A survey was mailed to 128 CHCs in 10 Midwestern states in 2016. Generalized estimating equation models were used to assess associations between program characteristics and stage of integration implementation (pre-contemplation, contemplation, preparation, action, or maintenance). Content analysis of open-ended responses identified integration barriers.

Results:

Response rate was 60% (N=77). Most CHCs had co-located BH and primary care services, warm hand-offs from primary care to BH clinicians, shared scheduling and EHR systems, and depression and substance use disorder screening. Thirty-two CHCs (42%) indicated they had completed integration and were focused on quality improvement (maintenance). Being in the maintenance stage was associated with having a psychologist on staff (OR=7.16, 95% CI 2.76–18.55), a system for tracking referrals (OR=3.42, 95% CI 1.03–11.36), a registry (OR=2.71, 95% CI 1.86–3.94), PCMH designation (OR=2.82, 95% CI 1.48–5.37), and a lower proportion of Black/African American patients (OR=0.82, 95% CI 0.75–0.89). The most common barriers to integration were difficulty recruiting and retaining BH clinicians and inadequate reimbursement.

Discussion:

CHCs have implemented many foundational components of integrated BH. Future work should address barriers to integration and racial disparities in access to integrated BH.

Nearly one in five adults in the U.S. experiences a mental health or substance use disorder each year (National Institute of Mental Health, 2019). An important strategy to address the burden of mental health is integrated behavioral health (BH), in which medical and BH providers collaborate to address patients’ overall health. Studies have shown that integrated BH improves quality of care and patient outcomes while reducing costs (Crowley & Kirschner, 2015). Uptake of integrated BH in real-world practice, however, has been variable (Davis et al., 2013). Clinics have found that implementation requires extensive training, effective leadership, and changes to infrastructure, attitudes, clinical practice, and funding (Clark et al., 2017; Hall et al., 2015; Vogel, Kanzler, Aikens, & Goodie, 2016; Wakida et al., 2018).

Despite the challenges, the majority of community health centers (CHCs) in the U.S. have integrated BH and primary care to some degree (Jones & Ku, 2015). CHCs are community-based organizations that provide comprehensive primary health care to patients regardless of ability to pay (Health Resources & Services Administration, 2018). In a national survey of CHCs in 2010, 65% met the authors’ criteria for being fully integrated, i.e., co-location, communication and collaboration, shared access to information, and joint decision-making (Lardiere, Jones, & Perez, 2011). In a follow-up study of the CHCs deemed fully integrated in the 2010 survey, over 80% reported they had BH clinicians available for warm hand-offs, but less than a third used a registry. Half reported barriers related to recruitment and retention of BH clinicians and billing restrictions for same-day physical and mental health services (NORC, 2011).

Since the 2010 survey, major healthcare policy changes have gone into effect. The Affordable Care Act (ACA) allocated more funding to support CHCs, including grants focused on BH integration (Health Resources & Services Administration, 2012, 2014a, 2014b). In addition, the expansion of Medicaid offered CHCs in some states additional revenue. Three-quarters of CHCs in expansion states and two-thirds in non-expansion states increased mental health services after implementation of the ACA (Rosenbaum et al., 2017). The ACA also directed funds to the National Health Service Corps loan repayment programs to encourage BH clinicians to work in underserved areas (Office for the Assistant Secretary for Planning and Evaluation, 2014). Additionally, same-day billing restrictions were loosened in many states (Ertle & McKinney, 2015; Roby & Jones, 2016). Thus, CHCs’ capacity for integrated BH has likely increased in recent years.

Efforts to promote BH integration would benefit from understanding the experiences of CHCs across different stages of implementation, particularly what differentiates developing vs. well-established programs. Factors associated with stage of integration implementation have not been well elucidated (Guerrero, Heslin, Chang, Fenwick, & Yano, 2015). The purpose of this study was to provide an updated, comprehensive analysis of BH integration at CHCs across the Midwest and to identify factors that distinguished CHCs in the maintenance stage of integration.

Method

Design

In summer 2016, we surveyed BH leaders at CHCs affiliated with Midwest Clinicians’ Network (MWCN), a nonprofit corporation that provides education, research, and networking for CHCs in 10 states. Surveys were mailed to BH directors (or another knowledgeable leader) at all 132 MWCN-affiliated CHCs. Questionnaires were mailed up to three times approximately one month apart with $5 included in the first mailing. The study was determined to be exempt by the University of Chicago Biological Sciences Division Institutional Review Board.

Measures

The survey (Appendix 1) was developed based on a review of major integrated BH models and existing survey measures (Agency for Healthcare Research and Quality; AIMS Center, 2019; Fauth & Tremblay, 2011; Heath, Wise Romero, & Reynolds, 2013; Lardiere et al., 2011; Peek, 2013). Results related to substance use disorder treatment and integration of depression and diabetes care have been previously published (Jones et al., 2020; Laiteerapong et al., 2021).

BH Program Characteristics

Survey questions covered various aspects of CHCs’ BH programs: access to BH care (e.g., how often BH providers were available, referral arrangements); types of BH services (e.g., short-term therapy, long-term therapy, groups); integration of BH and primary care (e.g., co-location, warm hand-offs, shared systems); population health (e.g., systematic screening, use of a registry); personnel (e.g., FTE by clinician type); and funding (e.g., grants, insurance payments).

Stage of Implementation for BH Integration

To assess stage of integration implementation, we asked, “Overall, how would you describe the status of your CHC’s efforts to integrate physical and BH care,” adapted from Sanchez, et al. (Sanchez, Thompson, & Alexander, 2010). Response options corresponded to the Transtheoretical Model’s stages of change: “We have not started thinking about it” (pre-contemplation); “We are thinking about it, but are unsure how to proceed” (contemplation); “We are in the planning stage” (preparation); “We have implemented some strategies, but have more work to do” (action); and “Our implementation is complete, and we are focused on maintenance and quality improvement” (maintenance) (Prochaska, Redding, & Evers, 2015).

Barriers to BH and Primary Care Integration

Respondents were asked to rate their agreement with statements about potential barriers identified in the literature (e.g., “We have difficulty recruiting or retaining BH counselors”). An open-ended question asked, “What other barriers do you perceive regarding BH integration at your CHC?”

CHC Characteristics

Respondents were asked to report the number of clinic sites at their CHC and whether their CHC had Patient Centered Medical Home (PCMH) designation. Publicly available data about CHCs’ patient populations and locations were accessed through the Bureau of Primary Health Care Uniform Data System.

Analysis

The dependent variable was stage of integration implementation, dichotomized as maintenance vs. earlier stages. Independent variables were CHC characteristics, BH program characteristics, and barriers to integration. We constructed generalized linear mixed models via generalized estimating equations (GEE) with the logit link function to model the probability of having reached the maintenance stage, taking within-state associations into account. We conducted bivariate analyses for each independent variable. Variables with p-values <0.1 in bivariate analyses were considered in multivariate analyses. We used backward selection to eliminate variables starting with the largest p-value until the GEE model fit criteria, QIC and QICu, were minimized. We calculated adjusted odds ratios (OR) and determined significance based on 95% confidence intervals (CI).

We conducted content analysis of responses to the open-ended question about barriers (Hsieh & Shannon, 2005). A codebook was generated based on initial review of responses. Three reviewers (ES, ML, NL) independently coded responses and discussed to consensus. Similar codes were combined into overarching themes.

Results

CHC Characteristics

Four CHCs were deemed ineligible: three offered limited primary care services and mail for one was returned undeliverable. Of the 128 eligible CHCs, 77 from nine states completed surveys (60% response). States with the largest number of respondents were Illinois, Michigan, and Ohio (Appendix 2). Characteristics of CHCs’ patient populations are shown in Table 1.

Table 1.

Community Health Center Characteristics (N=77)

	Mean (SD) / N (%)
Patient population, mean (SD)*
Adults (18–64 years)	60% (10%)
Older adults (≥65 years)	8% (5%)
Black/African American	26% (25%)
Hispanic/Latino	17% (21%)
Asian	2% (5%)
Native American/American Indian	1% (1%)
< 100% Federal poverty limit	68% (17%)
Medicaid	52% (16%)
Uninsured	19% (13%)
Number of clinic sites, mean (SD)	5 (4)
Location, N (%)
Rural	26 (34%)
Urban	34 (44%)
Both	17 (22%)
Mental Health Professional Shortage Area, N (%)	56 (73%)
Patient Centered Medical Home, N (%)
All clinic sites	43 (57%)
Some clinic sites	25 (33%)
No clinic sites	7 (9%)

^*Mean (SD) at the health center level (e.g., on average, 26% of the patients at a health center were Black/African American)

BH Program Characteristics

Access

Nearly all CHCs offered BH services at one or more clinic sites (N=75, 97%) and the majority had BH services at all sites (N=47, 61%) (Table 2). Most had BH clinicians in clinic every day (N=61, 79%). Less than half of CHCs had formal referral arrangements with external BH organizations (N=32, 42%).

Table 2.

Characteristics of Community Health Centers’ Behavioral Health Programs (N=77)

	Health Centers N (%)
Access
BH services available at
All clinic sites	47 (61%)
Some clinic sites	28 (36%)
No clinic sites	2 (3%)
Appointment types
Scheduled	72 (94%)
Same day	57 (74%)
Walk in	44 (57%)
Wait list	23 (30%)
BHPs in clinic every day	61 (79%)
Track if BH referrals completed	61 (81%)
Internal referrals	46 (60%)
External referrals	38 (49%)
Referral arrangements with external BH organizations
Formal agreement	32 (42%)
Informal arrangement	24 (31%)
No arrangements with specific organizations	9 (12%)
Services
Brief interventions (<30 minutes of education/counseling)	65 (84%)
Educational materials (e.g., pamphlets, books, videos)	64 (83%)
Health behavior counseling	61 (79%)
Short-term individual therapy (<8 visits)	61 (79%)
Long-term individual therapy (≥8 visits)	52 (68%)
Case management	38 (49%)
Medication management by BHP	33 (43%)
Group classes	24 (31%)
Telehealth	17 (22%)
Integration
BH and PC use the same appointment scheduling system	73 (96%)
BH and PC use the same EHR	72 (95%)
PCPs and BHPs can both access BH treatment plan	67 (91%)
At least one champion for BH integration	65 (86%)
Warm hand-offs from PCPs to BHPs	63 (82%)
BH and PC located in
Same offices	62 (81%)
Separate offices within same building	13 (17%)
Population Health
Systematic screening
Depression	74 (96%)
Substance use	54 (70%)
Anxiety	38 (49%)
Suicidality	34 (44%)
Trauma	16 (21%)
Other	12 (16%)
Systematic tracking
Depression	54 (70%)
Anxiety	19 (25%)
Other	9 (12%)
EHR-based BH registry	38 (51%)
Frequency of depression screening
Initial visit	1 (1%)
Annual	45 (62%)
Every 3–6 month	10 (14%)
Every visit	17 (23%)
Depression triage algorithm based on severity	44 (59%)
Personnel
BHPs on staff, N (%) / Total FTE, mean (SD)	74 (96%) / 5.33 (4.74)
Licensed social worker	59 (77%) / 2.94 (2.43)
Psychiatrist	29 (38%) / 1.18 (1.16)
Licensed professional counselor	26 (34%) / 2.55 (1.60)
Licensed psychologist	19 (25%) / 2.52 (2.62)
Masters prepared nurse with BH specialty	11 (14%) / 0.94 (0.74)
Licensed marriage and family therapist	5 (6%) / 1.40 (0.55)
Training site for BHPs
Licensed social worker	20 (26%)
Psychiatrist	7 (9%)
Licensed professional counselor	8 (10%)
Licensed psychologist	9 (12%)
Masters prepared nurse with BH specialty	8 (10%)
Licensed marriage and family therapist	3 (4%)
BH training required for other staff
PCPs	30 (39%)
Clinical staff	25 (32%)
Non-clinical staff	15 (19%)
Funding
Insurance payments	58 (75%)
Patient self-pay	44 (57%)
Federal, state, or local grants	43 (56%)
Private grants	12 (16%)
Partially or completely unfunded	15 (19%)

BH=behavioral health, BHP=behavioral health provider, FTE=full-time equivalent. PC=primary care, PCP=primary care provider. Systematic screening is defined as having a protocol that applies for every patient and is actively used by staff. Systematic tracking is defined as having a system that allows care managers and providers to access information for a specified patient population, monitor progress toward meeting clinical targets, proactively identify patients who are not improving, and change treatment plans. For total FTE, mean is calculated among CHCs with each type of provider, not including CHCs with zero FTE.

Services

The most widely reported services were brief interventions (N=65, 84%), educational materials (N=64, 83%), health behavior counseling (N=61, 79%), and short-term therapy (< 8 visits; N=61, 79%).

Population Health

Most CHCs systematically screened adult (N=73, 95%) and adolescent (N=66, 86%) patients for depression at least once per year (N=72, 94%) using the Patient Health Questionnaire (PHQ) (PHQ-2 (N=2, 3%), PHQ-9 (N=20, 26%), both (N=52, 68%)) (Kroenke, Spitzer, & Williams, 2001, 2003). Many CHCs also screened for substance use disorder (N=54, 70%). CHCs indicated a need for better screening tools for patients with low literacy (N=47, 61%).

Systematic tracking (defined as a system that allowed care managers and clinicians to access information for a patient population, monitor progress, and change treatment plans) was typically done for patients with depression (N=54, 70%). Half of CHCs used a BH registry within the electronic health record (EHR) system (N=38, 51%) to track patients; some of these CHCs had an additional registry outside the EHR (N=10, 19%).

Integration

BH and primary care services were located in the same offices at most CHCs (N=62, 81%). BH and primary care clinicians almost always used the same scheduling (N=73, 96%) and EHR (N=72, 95%) systems. Primary care clinicians were able to do warm hand-offs to BH clinicians during appointments at 82% of CHCs (N=63).

Personnel

Most CHCs had at least one licensed social worker on staff (N=59, 77%). Less than half had a psychiatrist (N=29, 38%) and only a quarter had a psychologist (N=19, 25%). About a third of CHCs (N=25, 32%) served as training sites for BH clinicians.

Funding

Seventy-five percent (N=58) of CHCs funded their BH services through payments from insurance. About 60% also relied on self-pay and/or government grants. Sixty percent (N=46) received start-up or service expansion funding to help launch their BH services, and a third (N=26, 34%) received grant funding for BH integration through the ACA. Nineteen percent (N=15) reported that a portion of their BH services were unfunded.

Stage of Implementation for BH Integration

Efforts to integrate physical and BH care were underway at all of the CHCs: two (3%) in preparation, 43 (56%) in action, and 32 (42%) in maintenance. CHC characteristics, BH program characteristics, and barriers to integration for CHCs in maintenance vs. preparation/action are shown in Appendix 3, along with bivariate associations between each independent variable and stage of integration. In the final multivariate model (Table 3), CHCs that had a psychologist on staff were seven times more likely to be at the maintenance stage (OR=7.16, 95% CI 2.76–18.55). CHCs that tracked whether or not BH referrals were completed (OR=3.42, 95% CI 1.03–11.36), maintained a BH registry (OR=2.71, 95% CI 1.86–3.94), and had PCMH designation at all clinic sites (OR=2.82, 95% CI 1.48–5.37) were also more likely to be at the maintenance stage. CHCs with a higher proportion of Black/African American patients were less likely to be at the maintenance stage: odds were 18% lower per 10% difference in the proportion of Black/African American patients.

Table 3.

Characteristics Associated with Maintenance Stage of Behavioral Health Integration at Health Centers

	Multivariate Model		Final Multivariate Model
	OR (95% CI)	p-value	OR (95% CI)	p-value
% Black/African American patients	0.74 (0.68–0.81)	0.06	0.82 (0.75–0.89)	0.07
% Medicaid patients	1.22 (0.99–1.51)	0.13	--	--
Mental Health Professional Shortage Area	3.24 (0.84–12.52)	0.13	--	--
Patient Centered Medical Home designation at all clinic sites	2.77 (0.93–8.29)	0.09	2.82 (1.48–5.37)	0.02
BH services at all sites	2.04 (0.32–12.87)	0.43	--	--
Same day BH appointments	2.65 (0.72–9.74)	0.14	--	--
BHPs in clinic every day	0.89 (0.23–3.45)	0.87	--	--
Track if BH referrals completed	2.47 (1.19–5.15)	0.05	3.42 (1.03–11.36)	0.048
Formal referral arrangement with external BH organization	0.18 (0.05–0.68)	0.15	--	--
Warm hand-offs from PCPs to BHPs	1.24 (0.56–2.74)	0.63	--	--
EHR-based BH registry	3.07 (1.03–9.11)	0.20	2.71 (1.86–3.94)	0.02
Psychologist on staff	7.29 (1.45–36.57)	0.13	7.16 (2.76–18.55)	0.048
Financial barriers to patients seeing PCPs and BHPs on same day	0.50 (0.17–1.45)	0.17	--	--

BH=behavioral health, BHP=behavioral health provider, PCP=primary care provider. Multivariate model includes variables with p-values <0.1 in bivariate analyses (see Appendix 3). For the final multivariate model, we used backward selection to eliminate variables starting with the largest p-value until model fit criteria, QIC and QICu, were minimized.

Barriers to BH and Primary Care Integration

Many CHCs had difficulties recruiting or retaining BH clinicians, especially psychiatrists (N=55, 75%) (Table 4). Half of CHCs indicated that reimbursement rates were not sufficient to cover costs of providing BH services or integrating BH and medical care. About a third of CHCs reported financial barriers to patients being seen for primary care and BH on the same day.

Table 4.

Barriers to Behavioral Health Integration at Community Health Centers (N=77)

	Strongly Disagree / Disagree	Undecided	Strongly Agree / Agree
Reimbursement rates from payers adequately cover the cost of providing BH services	38 (51%)	16 (21%)	21 (28%)
Reimbursement rates from payers adequately cover the cost of integrating BH services and medical care	38 (51%)	22 (30%)	14 (19%)
Financial barriers prevent patients from being seen by primary care and BH providers on the same day	46 (61%)	3 (4%)	27 (36%)
Certification or licensure requirements prevent trained providers from providing BH services	54 (70%)	9 (12%)	14 (18%)
We have difficulty recruiting or retaining BH counselors	28 (36%)	14 (18%)	35 (45%)
We have difficulty recruiting or retaining psychiatrists	8 (11%)	10 (14%)	55 (75%)

BH=behavioral health.

Fifty-three CHCs (69%) responded to the open-ended question about barriers, with responses falling within four major themes (Table 5): limited resources, challenges changing systems and practices, inadequate payment, and individual patient barriers.

Table 5.

Additional Barriers to Integration Reported by Community Health Centers (N=53)

Theme	N (%)	Example Quotes
Limited health center resources	18 (34%)	“We are a young organization working towards integration. Our barrier is facility capacity and staff capacity.” “[Having] enough staff (funding) at each site to meet the demand.” “No strong long-term mental health options for referral.”
Challenges changing systems and practices	17 (32%)	“Changing leadership [requires] continuous efforts to seek buy-in.” “Communication with primary care providers. We are considering scheduling a monthly meeting but it’s difficult getting everyone together.” “Need behavioral health assessment and follow-up documentation templates.”
Inadequate payment	13 (25%)	“Reimbursement for Medicare services require LMSW. Must open to other licensure LLP, LPC, LMFT etc.” “Lack of knowledgeable billers on behavioral health issues.” “[Do not receive] fair payment for behavioral health providers [or] recognition of this role. As a result, there is a situation of behavioral and medical providers coming and leaving the organization.”
Individual patient barriers	13 (25%)	“Health insurance, stigma, denial.” “Demand for services is high, but no show appointment rates continue to be high compared with other specialties.” “We see 500+ refugee patients every year. Competency (cultural competency or cultural curiosity) takes lots of time and energy and digging. Behavioral health is usually a foreign concept. Somatic complaints are frequent - makes integration very challenging.”

Limited Health Center Resources

Eighteen CHCs (34%) indicated that limited resources were a barrier to integration. They reported that the number of BH clinicians was insufficient to meet demand, and they lacked funding and space to hire additional clinicians. CHCs also reported that referral options for patients requiring specialty mental health treatment were overwhelmed or non-existent in their communities.

Challenges Changing Systems and Practices

Seventeen CHCs (32%) described challenges changing systems and practices to support integration. Some had difficulty getting buy-in from leadership, clinicians, and staff, and this problem was exacerbated because turnover was high. Respondents desired better communication between primary care and BH clinicians, and better coordination with community BH clinicians. They expressed a need for more integrated health information technology, particularly BH documentation templates.

Another barrier was inadequate understanding by primary care and BH clinicians of how to optimally practice in an integrated clinic. Respondents commented that primary care clinicians were sometimes resistant to changing their practices or unaware that BH services could be effective for improving health behaviors. One respondent wrote that it was difficult “getting clinicians to fully use BH to also look at barriers to health not just mental illness.”

Similarly, respondents stated that some BH clinicians lacked training in integrated care or found it difficult to adhere to brief interventions while serving a high-need patient population. For example, one reported challenges “maintaining brief solutions-focused therapy when many patients present with chronic, long-standing mental health issues.”

Inadequate Payment

Thirteen CHCs (25%) wrote about barriers related to payment. They indicated that reimbursement was inadequate and that certain clinicians were reimbursed at lower rates or not permitted to bill. They struggled with reimbursement for services necessary for providing integrated care, including brief interventions without diagnostic assessment and telepsychiatry services. As one respondent wrote, “CMS is still developing CPT codes and payment systems to support integrated care. Our care is in front of being reimbursed for it.” BH carve-outs limited reimbursement for some CHCs.

Individual Patient Barriers

Thirteen CHCs (25%) identified patient barriers that impeded integration. They commented that lack of awareness, stigma, language and cultural barriers, transportation issues, and cost prevented patients from engaging in treatment. Time was also a limitation for patients, as one respondent described: “So many important requests are given to each patient at each visit. Investing their time is often prohibitive – they have already been at our clinic a long time.”

Discussion

In this study, we found that Midwestern CHCs have implemented many aspects of integrated BH, including co-located BH and primary care services, shared scheduling and EHR systems, warm hand-offs, brief interventions and short-term therapy, and systematic depression and substance use disorder screening. CHCs that were reported by their BH director to be in the maintenance stage of implementing BH integration were more likely to have a psychologist on staff, a system for tracking BH referrals, a BH registry, and PCMH designation. CHCs serving a higher proportion of Black/African American patients were less likely to be in the maintenance stage. The most common barriers to integration were difficulty recruiting and retaining BH clinicians and inadequate reimbursement.

Interestingly, CHCs with a psychologist on staff were much more likely to have an established integrated BH program than CHCs without a psychologist. There are several possible explanations. Psychologists might have particular training or experience that prepares them to work in an integrated setting. They might be attracted to working at CHCs with demonstrated commitment to BH integration. CHCs with more financial resources might be able to invest in both developing integrated BH and hiring psychologists, whose salaries are typically higher than other BH clinicians. Psychologists might be more likely to occupy leadership roles from which they can drive integration efforts. While many have discussed the importance of equipping providers to work in integrated settings, there is limited research on the impact of BH clinicians’ professional background on integration (Vogel et al., 2016).

CHCs in the maintenance stage were more likely to be taking a proactive, population health approach to managing BH concerns by tracking referrals and maintaining a registry. Similarly, a prior study identified systematic tracking of patients as a key hurdle during early implementation of integrated BH (Davis et al., 2013). CHCs have used these population health strategies to improve care for other chronic conditions like diabetes (Pollard et al., 2009). Applying similar methods for managing mental and physical health conditions might serve as an indicator of integration. Being in the maintenance stage was also associated with PCMH designation, a model which emphasizes comprehensive, team-based, coordinated care (Working Party Group on Integrated Behavioral Healthcare, 2014). PCMH standards also include competencies related to BH screening, care management, and integration (National Committee for Quality Assurance, 2020). Encouragingly, depression screening, substance use screening, and BH registries were much more widely implemented in our study than a prior survey of CHCs in 2010 (Lardiere et al., 2011; NORC, 2011).

Integrated BH is a broad term encompassing various models of care, which are often combined in real-world implementation (Blasi et al., 2018; Vogel et al., 2016). We found that many CHCs had adopted aspects of primary care behavioral health (PCBH) (Robinson & Reiter, 2016). In PCBH, primary care and BH clinicians work together in the same space and regularly collaborate. To meet the needs of the general population and serve a large volume of patients, BH clinicians are available for warm hand-offs and offer brief therapy (Reiter, Dobmeyer, & Hunter, 2018; Sandoval, Bell, Khatri, & Robinson, 2018). The collaborative care model (CoCM), on the other hand, involves systematic management of a specific population (e.g., patients with depression) using a registry. A care manager provides on-going follow-up and facilitates communication between primary care clinicians and a psychiatric consultant (Sanchez, 2017; Unützer, Schoenbaum, Druss, & Katon, 2006). Registries and psychiatrists, key elements of CoCM, were less common at the CHCs in our study.

CHCs have expanded their BH services in recent years, but the number of patients has grown faster than the number of BH staff (Shin, Sharac, & Mauery, 2013). In 2010, 83% of CHCs provided some BH services but only 33% provided BH services at all of their sites (Lardiere et al., 2011). In our study, this had grown to 97% and 61%, respectively. Difficulty recruiting and retaining BH clinicians, especially psychiatrists, was a barrier in 2010 and remained a major challenge for CHCs in our study (NORC, 2011). While the percentage of CHCs with social workers on staff was higher in our study than in 2010 (77% vs. 59%), the percentage with psychiatrists, psychologists, therapists, counselors, and nurse specialists did not increase. The percentage serving as training sites for BH clinicians also remained the same.

We found that CHCs serving a higher proportion of Black/African American patients were less likely to be in the maintenance stage of BH integration implementation. This finding aligns with a previous study which reported that neighborhoods with more Black/African American residents were less likely to have BH and primary care clinicians working in close proximity to each other (Wielen et al., 2015). Lower adoption of BH integration is likely due to the shortage of BH providers in areas with higher concentrations of racial and ethnic minorities (Cook, Doksum, Chen, Carle, & Alegría, 2013).

Recent years have seen some progress to address concerns related to payment and access. New billing codes were introduced for CoCM and health behavior assessment and intervention (American Psychological Association, 2019; Center for Medicare and Medicaid Services, 2019; Chapman, Chung, & Pincus, 2017). The majority of states now allow same-day billing for primary care and BH visits, which was one of the most significant barriers reported by CHCs in 2010. The only state in the MWCN region at the time of our survey that did not allow same-day billing was Iowa (Ertle & McKinney, 2015; Roby & Jones, 2016). Still, some respondents reported financial barriers to same-day visits, suggesting they may have been unaware of policy changes or the changes might not have removed all barriers. In Minnesota, for example, CHCs had to elect an alternative payment method in order to bill for same-day visits (Minnesota Department of Human Services, 2017).

This study has several limitations. Responding CHCs may have been more invested in BH integration than non-respondents. Stage of integration implementation was self-reported, and stage of implementation may not correspond to level of integration (i.e., the degree to which services are truly integrated). Causal associations cannot be determined due to the cross-sectional design. Results may not be generalizable to CHCs outside the Midwest or to other primary care settings given differences in patient populations, policies, and funding. Compared to other primary care settings, CHCs may be particularly driven to integrate BH due to the high burden of BH conditions and lack of other resources for uninsured or underinsured patients.

A major implication of our study is that additional improvements are needed to grow and sustain integrated BH in CHCs. Increasing reimbursement rates and expanding billing privileges to all types of licensed clinicians would help to address staffing issues (Behavioral Health + Economics Network, 2018). Establishing CHCs as training sites, additional financial incentives to work in underserved areas, licensure reciprocity, and efforts to increase workforce diversity might help to mitigate the uneven distribution of BH clinicians (Frogner et al., 2020; Holden et al., 2014; Kepley & Streeter, 2018; Page, Buche, Beck, & Bergman, 2017; Sanchez, Ybarra, Chapa, & Martinez, 2016; Watanabe-Galloway, Madison, Watkins, Nguyen, & Chen, 2015). Though not common among our respondents in 2016, telehealth usage has grown in recent years (Mandelbaum, 2020; National Association of Community Health Centers, 2020) and could improve access to BH interventions and psychiatric consultation if technological, workflow, and policy barriers are addressed (Fortney et al., 2019; Mace, Boccanelli, & Dormond, 2018; Mongelli, Georgakopoulos, & Pato, 2020; Raney, Bergman, Torous, & Hasselberg, 2017). Opportunities for training and technical assistance could facilitate efforts to implement integrated care (Hall et al., 2015), particularly for models like the CoCM that have a robust evidence base but have been less widely adopted in the CHC setting.

CHCs continue to be leaders in the movement toward integrated BH and their experiences can guide future work. Policies and interventions should support primary care clinics in instituting collaborative, multi-disciplinary, team-based care; implementing population health strategies for identifying, treating, and tracking patients with BH concerns; establishing a well-trained workforce prepared to work in integrated settings; and making high quality, culturally-appropriate integrated care accessible to all patients.

Public Significance Statement.

Community health centers have implemented many aspects of integrated behavioral health, including co-located behavioral health and primary care services, shared scheduling and EHR systems, warm hand-offs from primary care to behavioral health clinicians, brief interventions and short-term therapy, and systematic screening for depression and substance use disorder. Major barriers to integration reported by health centers are recruiting and retaining behavioral health clinicians and inadequate reimbursement.

Appendix 1. Survey

Appendix 2. Response Rate by State

Survey Response by State

State	CHCs Eligible	CHCs Responded	Response Rate
Illinois	21	16	76%
Indiana	13	7	54%
Iowa	6	4	67%
Kansas	5	4	80%
Michigan	24	15	63%
Minnesota	11	5	45%
Missouri	11	8	73%
Nebraska	4	0	0%
Ohio	24	14	58%
Wisconsin	10	5	50%

CHC=community health center. One CHC had locations in both Minnesota and Wisconsin and is included in the counts and response rate for both states.

Appendix 3. Bivariate Associations with Maintenance Stage of Integration

Health Center Characteristics by Stage of Behavioral Health Integration^*

	Preparation or Action (N=45)	Maintenance (N=32)	OR (95 % CI)†	p-value
Patient population, mean (SD)
Black/African American	28% (28%)	23% (20%)	0.92 (0.85–1.00)‡	0.048
Hispanic/Latino	18% (22%)	16% (20%)	0.94 (0.72–1.21)‡	0.61
Uninsured	20% (15%)	18% (9%)	0.91 (0.59–1.38)‡	0.65
Medicaid	54% (18%)	50% (14%)	0.88 (0.79–0.98)‡	0.02
Clinic sites, mean (SD)	6 (4)	5 (4)	0.96 (0.88–1.05)§	0.35
Location, N (%)
Rural	15 (33%)	11 (34%)	ref
Urban	19 (42%)	15 (47%)	1.07 (0.54–2.15)	0.83
Both	11 (24%)	6 (19%)	0.74 (0.29–1.88)	0.53
Mental Health Professional Shortage Area, N (%)	29 (64%)	27 (84%)	2.98 (1.48–6.01)∥	0.002
Patient Centered Medical Home at all clinic sites, N (%)	21 (49%)	22 (59%)	2.30 (1.10–4.81)¶	0.03

^*Stage of integration was determined based on the health center leader’s response to the following survey question: “Overall, how would you describe the status of your health center’s efforts to integrate physical and behavioral health care?” Those that selected “We are in the planning stage” were in Preparation. Those that selected “We have implemented some strategies, but have more work to do” were in Action. Those that selected “Our implementation is complete, and we are focused on maintenance and quality improvement” were in Maintenance.

^†Odds of being in maintenance stage of integration (vs. preparation or action stage)

^‡Per 10% increase in percentage of patients.

^§Per 1 site increase

^∥Compared to health centers not located in Mental Health Professional Shortage Area

^¶Compared to health centers with Patient Centered Medical Home designation at some or no clinic sites

Characteristics of Health Centers’ Behavioral Health Programs by Stage of Integration^*

	Preparation or Action (N=45)	Maintenance (N=32)	OR (95 % CI)†	p-value
BH services at all sites	23 (51%)	23 (72%)	2.44 (0.97–6.13)	0.06
Appointment types
Same day	30 (67%)	27 (84%)	2.70 (1.15–6.32)	0.02
Scheduled	41 (91%)	31 (97%)	3.02 (0.48–18.99)	0.24
Wait list	13 (29%)	10 (31%)	1.12 (0.33–3.78)	0.86
Walk in	27 (60%)	17 (53%)	0.76 (0.34–1.68)	0.49
Systematic depression screening at all sites	39 (87%)	30 (94%)	2.31 (0.59–9.07)	0.23
BH registry	20 (44%)	18 (62%)	2.05 (1.15–3.64)	0.02
Referral arrangement with external BH organization	24 (53%)	8 (25%)	0.32 (0.13–0.78)	0.01
Track BH referrals	33 (75%)	28 (90%)	3.11 (1.09–8.85)	0.03
Champion for BH integration	35 (80%)	30 (94%)	3.86 (1.01–14.72)	0.048
Co-located BH and primary care in same offices	32 (71%)	30 (94%)	6.09 (1.05–35.26)	0.04
Warm hand-offs	35 (78%)	28 (88%)	2.00 (1.22–3.29)	0.01
PCPs and BHPs can both access BH treatment plan	39 (93%)	28 (89%)	0.54 (0.17–1.73)	0.30
BHPs in clinic everyday	33 (73%)	28 (88%)	2.55 (0.94–6.89)	0.07
BHPs on staff
Psychiatrist	15 (33%)	14 (44%)	1.56 (0.49–4.94)	0.45
Psychologist	7 (16%)	12 (38%)	3.26 (1.74–6.09)	<0.001
Counselor or therapist	12 (27%)	16 (50%)	2.75 (0.94–8.07)	0.07
Social worker	33 (73%)	26 (81%)	1.58 (0.70–3.54)	0.27
BH training required for
PCPs	15 (33%)	15 (47%)	1.76 (0.66–4.74)	0.26
Clinical staff	11 (24%)	14 (44%)	2.40 (0.75–7.66)	0.14
Non-clinical staff	5 (11%)	10 (31%)	3.64 (0.82–16.18)	0.09
BH training required at least annually	9 (20%)	9 (28%)	1.57 (0.65–3.79)	0.32
Received ACA behavioral health integration grant	13 (29%)	13 (41%)	1.68 (0.46–6.21)	0.43

BH=behavioral health, PCP=primary care provider, BHP=behavioral health provider, ACA=Affordable Care Act. Values are N (%).

^*Stage of integration was determined based on the health center leader’s response to the following survey question: “Overall, how would you describe the status of your health center’s efforts to integrate physical and behavioral health care?” Those that selected “We are in the planning stage” were in Preparation. Those that selected “We have implemented some strategies, but have more work to do” were in Action. Those that selected “Our implementation is complete, and we are focused on maintenance and quality improvement” were in Maintenance.

^†Odds of being in maintenance stage of integration (vs. preparation or action stage), compared to health centers without each behavioral health program characteristic

Barriers to Behavioral Health Integration at Health Centers by Stage of Integration^*

	Preparation or Action (N=45)	Maintenance (N=32)	OR (95 % CI)†	p-value
Reimbursement rates from payers adequately cover the cost of providing BH services, N (%) disagree or strongly disagree	24 (56%)	14 (44%)	0.62 (0.26–1.46)	0.27
Reimbursement rates from payers adequately cover the cost of integrating BH services and medical care, N (%) disagree or strongly disagree	23 (53%)	15 (48%)	0.82 (0.41–1.61)	0.56
Financial barriers prevent patients from being seen by primary care and BH providers on the same day, N (%) agree or strongly agree	20 (44%)	7 (23%)	0.36 (0.12–1.09)	0.07
Certification or licensure requirements prevent trained providers from providing BH services, N (%) agree or strongly agree	9 (20%)	5 (16%)	0.74 (0.25–2.22)	0.59
We have difficulty recruiting or retaining BH counselors, N (%) agree or strongly agree	23 (51%)	12 (38%)	0.57 (0.27–1.24)	0.16
We have difficulty recruiting or retaining psychiatrists, N (%) agree or strongly agree	35 (80%)	20 (69%)	0.57 (0.29–1.11)	0.10

BH=behavioral health

^*Stage of integration was determined based on the health center leader’s response to the following survey question: “Overall, how would you describe the status of your health center’s efforts to integrate physical and behavioral health care?” Those that selected “We are in the planning stage” were in Preparation. Those that selected “We have implemented some strategies, but have more work to do” were in Action. Those that selected “Our implementation is complete, and we are focused on maintenance and quality improvement” were in Maintenance.

^†Odds of being in maintenance stage of integration (vs. preparation or action stage), compared to health centers that did not report each barrier