Care Coordination and Population Health Management Strategies and Challenges in a Behavioral Health Home Model

Abstract

Objective:

Behavioral health home (BHH) models have been developed to integrate physical and mental health care and address medical comorbidities for individuals with serious mental illnesses (SMIs). Previous studies identified population health management capacity and coordination with primary care providers (PCPs) as key barriers to BHH implementation. This study examines the BHH leaders’ perceptions of and organizational capacity to conduct these functions within the community mental health programs implementing BHHs in Maryland.

Methods:

Interviews and surveys were conducted with 72 implementation leaders and 627 front-line staff from 46 of the 48 Maryland BHH programs. In-depth coding of the population health management and primary care coordination themes identified sub-themes related to these topics.

Results:

BHH staff described cultures supportive of evidence-based practices, but limited ability to effectively perform population health management or primary care coordination. Tension between population health management and direct, clinical care, lack of experience, and state regulations for service delivery were identified as key challenges for population health management. Engaging PCPs was the primary barrier to care coordination. Health information technology (HIT) and staffing were barriers to both functions.

Conclusions:

BHHs face a number of barriers to effective implementation of core program elements. To improve programs’ ability to conduct effective population health management and care coordination and meaningfully impact health outcomes for individuals with SMI, multiple strategies are needed, including formalized protocols, training for staff, changes to financing mechanisms, and HIT improvements.

Introduction

To address well-documented premature mortality due to physical health conditions^1–5 and sub-optimal access to and quality of care for those conditions^{6 7} among people with serious mental illness (SMI), there has been a recent proliferation of “behavioral health home” (BHH) models, which integrate primary care services into specialty mental health settings.^{8 9} Early research evaluating BHHs has shown some promising results, including improvements in provision of guideline concordant care for diabetes and hypertension¹⁰ and small improvements in cholesterol and hypertension.^{11 12}

BHHs, typically led by a nurse care manager or coordinator(s), are responsible for three domains of care: (1) delivery of basic preventive health services; (2) population health management for physical health conditions; and (3) coordination of physical health services with primary care providers (PCPs).^{11 13 14} To implement a BHH, a financing mechanism must be created whereby specialty mental health programs can bill for services within these domains.^{14 15}

Studies of the implementation of BHHs in Iowa, Maryland, Missouri, and New York have identified population health management capacity and coordination with PCPs as two main challenges to implementation.^{12 13 16 17} Population health management addresses needs along a continuum of health states for a specified population using continuous health monitoring and targeted interventions (e.g., using reports in electronic health records to track Hemoglobin A1c and providing outreach to patients with poor diabetes control).¹⁸ Care coordination strategies organize patient care activities across multiple providers.¹⁹ Understanding how to improve these processes may help BHHs meet their full potential and meaningfully impact health outcomes for individuals with SMI.

In a prior study, we used mixed-methods to examine overall implementation strategies, barriers, and facilitators among community mental health programs implementing Medicaid BHHs in Maryland; that study describes Maryland’s BHH program in detail.¹³ Briefly, Maryland BHHs are based in psychiatric rehabilitation programs (PRPs). PRPs interested in becoming Medicaid BHHs must apply to and be certified by the state’s Department of Health. During our November 2015-December 2016 study period, there were 255 PRPs in Maryland, 48 of which were active BHH sites. PRPs receive a $102.86 per-member per-month payment contingent on the delivery of two health home services to each consumer per month.

We found that, as in other states, population health management and care coordination were key barriers to BHH implementation.^{12 13 16 17} In the present study, we conduct an in-depth examination of population health management capacity and coordination with PCPs in Maryland’s program. No prior studies have analyzed population health management and care coordination barriers, facilitators, and strategies in the context of integrated care for people with SMI. Our specific aims are: (1) to describe the organizational capacity to conduct population health management and care coordination within the community mental health programs implementing BHHs in Maryland and (2) to explore Maryland BHH leaders’ perceptions of the implementation of population health management and primary care coordination in their programs.

Methods

Study Population & Design

BHH leaders and PRP staff were recruited from each of the 48 active Maryland Medicaid BHHs between November 2015 and December 2016. At each site, the nurse care manager, PRP director, and all front-line staff were eligible to participate. All study participants completed an informed consent process. BHH leaders received a $50 gift card and other staff received a $20 gift card for participation. This study was approved by the [blinded for review] Institutional Review Board.

Data Collection & Measures

The nurse care manager at each BHH was asked to complete a survey assessing the organizational structure of and services delivered by that site. The survey included 110 items and took approximately 40 minutes to complete. We report the results of 11 items measuring key structural components of population health management and primary care coordination.

Front-line PRP staff, including case managers, counselors, and peer leaders, were asked to complete a separate survey during regularly scheduled staff meetings measuring their perceptions of their program’s organizational capacity and implementation climate. We report the results of 15 items adapted from Texas Christian University Organizational Readiness to Change (ORC-D4)²⁰ and implementation climate scales.²¹

The nurse care manager and PRP directors at each site were invited to participate in semi-structured, 30–45 minute interviews conducted by a trained research staff member examining perceptions of BHH implementation.

Analysis

Descriptive statistics from the surveys were analyzed using Stata Version 14. Interviews were coded using a hybrid inductive/deductive approach, with two coders iteratively coding transcripts to ensure reliability.¹³ We performed in-depth coding using NVivo Version 11 of the population health management and primary care coordination themes identified in the original study.¹³

Results

Data spanned 46 of the 48 (96%) BHHs. Forty-one nurse care managers representing 46 BHHs, 31 PRP directors representing 37 BHHs, and 627 PRP staff representing 38 BHHs completed 72 interviews and 673 total surveys. We do not have access to staffing information at sites that declined to participate; survey response rates were therefore calculated within participating sites only. For the nurse care manager survey, completed by nurses representing 46 sites as noted above, the response rate was 100%. The response rate for the staff survey was 83% overall and >50% (range: 54–100%) across the 38 participating sites. Demographic characteristics of study participants are reported in the original study.¹³

BHHs lacked key structural elements needed to conduct population health management and primary care coordination (Table 1). For example, only 24% of BHHs had a formal partnership with an offsite PCP and only 63% had registries of all consumers’ physical health needs and services.

Table 1.

Structural characteristics of Maryland’s Medicaid behavioral health home program (N=46 Health Home Sites as described by 41 nurse care managers¹)

Formal partnerships between behavioral health home programs and primary care providers:	% of Sites
Co-located primary care provider (physician, nurse practitioner, physician’s assistant)	15%
Formal partnership2 with an offsite primary care clinic or practice	24%
Shared electronic medical record access:
Shared electronic medical record access between behavioral health home and primary care provider(s)	35%
Communication between behavioral health homes and primary care providers:
Health home providers consistently send/receive notification of changes in participants’ status to/from   primary care providers	46%
Frequency of regularly scheduled in-person, virtual, or telephone meetings between health home staff   and primary care providers:
Daily	0%
Weekly	17%
2–3 times per month	33%
Once per month	43%
Less than once per month	13%
No regularly scheduled meetings	43%
Monitoring of consumer health needs:
Monitor the status of the entire health home population on key indicators (e.g. blood pressure, tobacco use) to identify and prioritize population-wide needs and trends?	93%
Have an electronic registry with information about all consumers’ needs and services received in the following categories:
Physical health needs and services	63%
Mental health needs and services	52%

¹Some nurse care managers worked at more than one behavioral health home site

²Formal partnerships were defined as involving a MOU or contract

PRP staff reported organizational climates supportive of implementation of evidence-based practices (Table 2). For example, 71% of staff reported that effective use of evidence-based practices was one of the organization’s main goals. However, responses related to organizational capacity, educational support and incentives indicated a limited ability for the organizations to implement these practices, e.g., 73% percent of respondents reported that frequent staff turnover was a problem.

Table 2.

Front-line community mental health program providers’ perceptions of their organization’s capacity to implement behavioral health homes (N=627 PRP Staff)

Organizational capacity:	% Agree
Frequent staff turnover is a problem for this organization	73%
This organization has enough staff to meet current client needs	26%
Staff at this organization are able to spend enough time with clients	50%
The heavy workload at this program reduces program effectiveness	60%
Compensation and benefits at this organization are adequate	34%
Focus on evidence-based practice:
One of this organization’s main goals is to use evidence-based practices effectively	71%
People in this organization think that the implementation of evidence-based practices is important	68%
Using evidence-based practices is a top priority in this organization	60%
Educational support for evidence-based practice:
Staff training and continuing education are priorities at this organization	59%
This organization provides conferences, workshops, or seminars focused on evidence-based practices	49%
This organization provides evidence-based practice trainings or in-services	54%
This organization provides evidence-based practice training materials, journals, etc.	44%
The budget here allows staff to attend professional conferences each year	28%
Incentives for evidence-based practice:
The better you are at using evidence-based practices, the more likely you are to get a bonus or raise	12%
This organization provides financial incentives for the use of evidence-based practices	13%

In the leadership interviews, three overarching themes related to both primary care coordination and population health management emerged: health information technology (HIT), staffing ratios, and staff recruitment and turnover; see Table 3 for detailed discussion of these themes. Two strategies for population health management and three related challenges emerged from the interviews (Table 4). Strategies included tracking health indicators to look at trends over time and prioritization of identifying high-risk consumers. Challenges included tension between direct clinical care and population health management, e.g., decisions about allocating limited nurse care manager time to seeing patients versus performing population health management tasks like tracking health indicators, lack of BHH provider experience conducting population health management, and the requirement that each client receive two BHH services per month.

Table 3.

Cross-cutting population health management and primary care coordination themes (N=72 leaders)

Key Themes and Illustrative Quotes
Cross-cutting challenges to population health management and primary care coordination
Key Theme: Health IT (mentioned by 51 leaders):
“If services are done through [affiliated primary care clinic], I can see radiology, microbiology. But not all of our clients use that clinic, and that becomes much more problematic in getting the data I need, like trying to figure out when someone had a colonoscopy, because our patient population are not good historians…so that really hampers the data.”
“So population health management, we haven’t really gotten a groove on that because of our lack of electronic medical record. It’s really hard to do that because you’re doing all of that manually, and who’s got time for that? Nobody.”
Key Theme: Staffing ratio (mentioned by 27 leaders):
“I think the client to staff ratio is a little high. It’s a little overwhelming. [We] Shouldn’t have like 250 clients to 1 person. I think It could make you feel like you’re not really doing anything and you’re just always drowning…”
I think it’s a lot of work. I think [nurse care manager] is often overwhelmed on the amount of work that she needs to do. There is not enough staffing resources, the Health Home staff.
Key Theme: Staff recruitment and turnover (mentioned by 30 leaders):
“And staffing. Finding the right staff, keeping the right staff. It seems like by the time you have them trained, they’re gone, and you’ve got to look for someone new.”
“We were looking for someone who we felt was competent enough to do the job and who would accept the pay because that’s the biggest thing. This department does not operate for profit. It’s a very tight budget. And we only have so much money that we can spend on a salary. So getting someone who is highly qualified with years and years of experience, probably not going to want or accept what we are able to pay.”
Sub-Theme: Nurse role (mentioned by 12 leaders):
“And here was another thing. I was finding with the nurses that I was interviewing, that is we can sit here all day long but they do not want to simply do population health management. I think it’s a rare nurse that I’ve seen that want to just do that. They want to bandage things and poke at things”
“You have to be careful how you word the job description in the ad for the nurse because you don’t want somebody who is a nurse and is expecting to see patients every day and do nursing stuff. You need a nurse who is interested in teaching, and doing oversight, and managing care. And that’s a different kind of person.”
“[Nurse care manager] actually started part-time until she graduated, and sat, and got her license, and then became our initial full-time, nurse care manager, until she decided that she actually missed the true, hands-on nursing component of nursing.”

Table 4.

Behavioral health home leaders’ perceptions of population health management in Maryland’s Medicaid Health Home Program for people with SMI (N=72 leaders)

Key Themes and Illustrative Quotes
Population health management strategies:
Key Theme: Tracking (mentioned by 35 leaders)
“We track that, so we track not only ER usage but we track what the reasons were. Are there any trends there? Is there any one particular health issue that really stands out that more people are going for? Are they going more for somatic stuff versus psychiatric stuff? We take a look at both of those components.”
I just pick a health indicator for the month and do a report of, “Okay, who’s been flagged as needing care? Who’s been compliant [with their care plan]?” Then, if they’re not compliant I follow up with their rehab specialist. January, or maybe it was December, was those that might have asthma or COPD - had they been prescribed an inhaler or used an inhaler? If they were flagged as having asthma or COPD, I would notify the rehab specialist, “Hey, do you see them using an inhaler? Do you know if they’ve been prescribed this inhaler? When they fill the prescription, do they actually use it?” So I get a lot of that response from the staff. I’ve requested medical records from primary care providers in the past, and I just do not have the luck of having much returned.
Key Theme: Prioritization (mentioned by 15 leaders)
“Well, I’ve kind of been grouping people into two categories, and there’s one that-- a higher acuity group of people who have very complex needs, or they’re severely medically ill, or they’re older and they’re going to need placement in some type of higher level of care, and then there’s those that maybe have a chronic condition like that but they’re managing it very well, or it’s not at a point where it’s that severe, so I’d consider them more of a preventative or maintenance type of group that we still follow up with.”
“Within the Health Home, it’s really population care management, where you’re looking at, ‘These seem to be the issues amongst this specific population.’ And the population that we’re looking at are the clients that are in our program. So, among the clients in our program, if there’s a high level of diabetes, there’s a high level of asthma, our nurses should be looking at that phenomenon and coming up with things that we can do to get better at reducing that statistically across the board.”
Population health management challenges:
Key Theme: Tension between direct clinical care and population health management (mentioned by 27 leaders):
“It was a bumpy few months… [the nurse care manager] was doing a lot of direct care, versus case management. And I think that’s why it wasn’t working and it was very stressful to her.”
“I mean we really are looking at the care coordination model, which sounded like what it was supposed to be and not the hands on. ‘Oh, the nurse is on site, I need a band-aid.’ No, the PRP staff can help you with finding a band aid and addressing that.”
“So the nurse care manager is, I feel, like a school nurse. Anytime somebody is throwing up it’s ‘Go get [the nurse], or anytime somebody has a cut, go see [the nurse] and get a Band-Aid when there’s First Aid kits everywhere. It’s just like I’m not able to do the more meaningful population health management that I would like to do because I can’t get away from that school nurse role.
“Do we really need an RN that just sits and punches in numbers? Because it’s all you can do when you’ve got a caseload of 250. An actual RN is much more beneficial to moving the quality numbers and the health of our clients, if they actually have time to spend with the clients and the staff.”
Key Theme: Lack of health home provider experience conducting population health management (mentioned by 11 leaders):
“You have to run these population management reports. I don’t see-- I guess it benefits the client if you take that report, you look at it, and you’re like, “We have all these clients with obesity. Let me run a group on that,” or, “Let me ensure the counselors are trained on that.” But you have to make sure that gets done. And if it doesn’t, it’s just like a report sitting there. We’re just billing for it, and it’s not really doing anything to benefit the client.”
“While the nurses that we have were great trained RNs - they were great medical model trained RNs - that doesn’t necessarily equate to good trained staff for community work.”
“And so, it was hard for us, I would say, to wrap our heads around the population health management piece and it seemed-- it still seems to me to this day that there’s a lot of - I’m not saying it’s push - a lot of-- the motto is a little bit more hands-off than probably what we’re doing.”
Key Theme: Requirement that each client receive two health home services per month (mentioned by 7 leaders):
“Now, if someone comes forward, and they already have two services, you’re certainly not going to turn them away, and that’s the problem. They could already be fulfilled with their services, but they come in, and they have a raging fever or infection, and you’re going to spend the time with them. So it’s a juggle. It is a juggle and a struggle because I want to get those people that have the empty spaces on my tracking sheet, but I also can’t turn my back on somebody.”

For primary care coordination, two strategies and one challenge emerged from the interviews (Table 5). Strategies included formal partnerships and informal relationships with external PCPs. Interviewees mentioned several strategies for establishing and maintaining informal partnerships, such as sending out introductory letters. The key challenge to primary care coordination was difficulty engaging PCPs.

Table 5.

Behavioral health home leaders’ perceptions of care coordination with primary care providers in Maryland’s Medicaid Health Home Program for people with SMI (N=72 leaders)

Key Themes and Illustrative Quotes
Primary care coordination strategies:
Key Theme: Formal partnerships with primary care providers (mentioned by 33 leaders):
“It’s been easier for us to access specialty practices because of the affiliation with the primary clinic that’s part of [health system].
It gives them [clients with SMI] more opening to specialty practices which might not take somebody except they’re part of the [name of health system] so they will take these clients coming from this direction.”
Key Theme: Informal relationships with external primary care providers (mentioned by 43 leaders):
“We have nurtured over the last two and a half, three years a very good working relationship with the community health center. And their mission really has a lot of overlap with our mission.”
Sub-Themes: Strategies to foster informal relationships
Relationship building (mentioned by 22 leaders)
“I am [the] squeaky wheel. I call and call and call, and I am really sugar sweet…and now, I’ve forged so many relationships. I have relationships with the hospital and all the social workers. They know my name. They call me now.”
Directing consumers with SMI to specific primary care providers (mentioned by 10 leaders)
“Some doctors, they’re not the ones you want to send your people to…they’re not going to understand. They’re not going to Get why this person is so disheveled. That’s going to be all they can focus on, so we try to work with providers that get it.”
Standard letters and forms (mentioned by 7 leaders)
“I did send out a letter introducing who I was and what the program is about. I did not get any response… probably the office staff gets it. Is it even getting to the primary care provider? Is it sitting on somebody’s desk? These are answers I don’t know.”
Accompanying consumers to primary care visits (mentioned by 5 leaders)
“All of us [health home providers] would probably say we just don’t feel like they [clients with SMI] get looked at the same way that you or I would. They might get passed off a little bit, so you can’t pass somebody off when you got [health home staff member] right there in the [PCP’s] office.”
Primary care coordination challenges:
Key Theme: Challenges engaging with primary care providers (mentioned by 56 leaders):
“All of our clients have-- I mean every client has a different primary care doctor…it’s really inconsistent, so being able to have those strong relationships with those outside providers has really been a challenge, because everyone’s totally different.”
“A couple of challenges. One has been, a lot of [primary care] physicians don’t want to hear from you.”
Sub-Theme: Lack of clear incentive for primary care providers to engage in care coordination (mentioned by 11 leaders):
”From my experience I [health home leader] just need to meet them [primary care providers], contact them, establish rapport in some way. I don’t know what I could do for them, yet, that would help them help me…it’s one of the big challenges because they’re just sort of outside of the organization and so don’t have the same team-based incentives as us [health home team].”
Sub-Theme: Negative attitudes about consumers with SMI (mentioned by 21 leaders):
”A lot of times, our folks are discounted by medical professionals, because they’re crazy. It’s a joke. Not my words, but…my nurse will say she told client-related information to a doctor, and nothing really was addressed.”

Discussion

Results suggest that while programs that opted to become Maryland BHHs had cultures supportive of the implementation of evidence-based population health management and care coordination processes, organizational and structural barriers limit programs’ ability to effectively perform these tasks. Similar barriers, especially challenges engaging external providers and lack of HIT infrastructure, have also been identified by BHH programs in other states.^{11 12 16 17 22} However, there is considerable variation in the structure of BHH programs across the U.S., and Maryland is unique in implementation of a BHH program in the PRP, as opposed to mental health outpatient clinic, setting. Thus, our discussion of results and recommendations for improving BHH population health management and care coordination should be interpreted as derived from research on Maryland’s BHH and not necessarily considered generalizable to other BHH programs.

Maryland BHH leaders reported using two main population health management strategies: tracking and prioritization. Providers lacked prior experience conducting population health management and there were beliefs among some that direct clinical care for consumers with SMI was more important than population health management tasks. A related challenge was Maryland’s requirement that Medicaid BHH programs deliver two services per month to each participant in order to receive the per-member-per-month reimbursement, regardless of health needs; some leaders perceived this requirement as detrimental to their ability to focus services on the highest-need consumers.

Most BHHs did not have formal partnerships with PCPs in the community, and those that did still reported challenges with engaging external providers as not all consumers used the contracted or onsite provider. Leaders perceived challenges in engaging external PCPs due to lack of incentives for PCPs to take the time to engage in care coordination or negative provider attitudes toward consumers with SMI.

Our findings suggest that to improve Maryland BHHs’ ability to conduct effective population health management and care coordination, multiple strategies are needed. Creation of standard guidelines and protocols for population health management and primary care coordination within BHH programs could improve implementation in Maryland and elsewhere. Such guidance could be created by professional groups like the American Psychological Association or embedded in BHH accreditation standards.^{23 24}

In addition, BHH staff need training in population health management and care coordination. Offering continuing education units or certifications may help ensure receipt of the training and provide opportunities for staff development, though funding will be needed to develop and disseminate such training; use of online training platforms may enhance feasibility. Improvements to HIT systems are also needed to ease communication and data management for both population health management and care coordination. Ideally, electronic reminders and clinical decision support tools imbedded in the workflow should be used to facilitate uptake of the population health management and care coordination tasks put forth in the guidelines and covered in the trainings discussed above.²⁵ Given the low-resource nature of many of the community mental health programs implementing BHHs in the U.S.,²⁶ financial incentives to spur HIT improvements are needed. The Medicare and Medicaid Promoting Interoperability Program²⁷ could serve as a model, though it is important to note that incentives alone may not be enough to increase uptake without interoperable platforms for sharing data with external providers.²⁸

Financing mechanisms have the potential to help address the BHH population health management and care coordination challenges identified in our study. Reimbursement requirements, like Maryland’s requirement that all BHHs deliver two services per month to each participant in order to receive the per-member-per-month health home reimbursement, should be considered carefully. Study participants viewed Maryland’s two-service requirement as a barrier to effective population health management, which emphasizes prioritizing service delivery to high-need consumers.¹⁸

Financing mechanisms also have implications for BHHs’ ability to coordinate care with external providers. In Maryland’s Medicaid BHH model, the entire $102.86 per-member-per-month reimbursement goes to the community mental health programs, giving external PCPs no direct financial incentive to engage in care coordination activities. An alternative model used in Vermont’s Medicaid health home program for people with opioid use disorders is the hub-and-spoke model. In this payment model, reimbursement is split between the hub (specialty substance use treatment programs implementing health homes) and spokes (team of collaborating PCPs)²⁹ providing financial incentive for external provider participation in the health home.

Another challenge with the BHH program’s financing mechanism is that community mental health programs are not accountable for achieving quality-of-care or consumer health outcomes; financial reimbursement is not tied to quality or outcome metrics. This challenge could be addressed by integrating BHHs with payment models that incentivize high-quality care, including global budgets³⁰ and accountable care organizations.³¹

This study has limitations. The study was conducted only within Medicaid BHHs in Maryland and all data was self-reported. The measures used may have been influenced by social desirability bias or staff’s desire to portray the BHH positively. To minimize this, results of survey measures were aggregated at the site-level and the informed consent process included language that employment would not be affected by study participation.

Previous studies identified population health management and primary care coordination as two main barriers to BHH implementation.^{12 13 16 17} Within Medicaid BHHs in Maryland, obstacles including HIT, staffing, and engaging PCPs need to be overcome to implement effective population health management and care coordination.