Background

California is one of the many states that utilize a “carve-out” to provide mental health services for publicly insured individuals in need of specialty mental health care. Mental health services that require a mental health specialist are “carved out” of the Medicaid managed care system, while general mental health needs that can be met within primary care remain a part of Medicaid managed care.1 The County Mental Health Department effectively becomes the Mental Health Plan for Medicaid participants with moderate to severe mental health conditions.

For Medicaid-insured children, these county mental health services are often provided by a community mental health clinic that contracts with the County Department of Mental Health (DMH). Thus, children who need specialty mental health (MH) care will be referred by their primary care provider (PCP) to a DMH or DMH-contracted mental health clinic. Although the system was designed to achieve better MH access for Medicaid participants, this separation of services and plans also creates several potential barriers to access and quality care for children in low-income communities, including a lengthy and complex referral process, and poor communication and coordination between PCPs and specialty MH care providers.2

Studies suggest that interventions aimed at creating collaborative systems of care for PCPs and specialty MH providers can improve MH outcomes among children and adolescents.3 These models play a critical role in improving MH access and utilization among children who need specialty MH care;4,5,6,7,–8 however, models for collaborative care, co-location, and integrated care may be logistically and financially challenging under a carve-out system, when payers for primary care and specialty MH care are completely separate and under separate plans.

In lieu of integration, PCPs may develop strategies to communicate and coordinate with community MH clinics to facilitate referral and follow-up and support families through specialty MH care. It is not clear how avenues for communication and coordination between community primary care and community specialty MH care in this carve-out system for publicly insured children affect access (e.g., via the referral process) to MH care services for children. It is also important to identify target areas for interventions aiming to improve coordination of referrals from primary care to MH care across these organizations.

The study objective was to explore the function of communication between pediatric PCPs and specialty MH providers across the primary care to specialty MH referral and care processes. The study aim was to describe primary care and specialty MH providers’ views on the referral and care process, and the communication and coordination challenges faced at each of these phases. The focus was on children receiving primary care at six clinical sites of a multi-site federally qualified health center (FQHC) and two community MH clinics that serve families in and around the FQHC’s community.

Methods

A convenience sample of clinicians, clinical staff, and executive and administrative leadership from six clinics of a multi-site FQHC and two local MH clinics serving low-income families in the San Fernando and Santa Clarita Valleys in Los Angeles, CA was utilized for this study. This multi-site FQHC and the two MH clinics were the community partners for a larger community-partnered study to improve the mental health referral process, and thus were the only clinics approached to participate in this qualitative study. In 2016, the FQHC clinics provided 112,101 primary health care visits for 37,367 children ages 0–18 annually; 98% of these patients had an annual household income under 200% of the federal poverty level, and 86% were Latino. These two MH clinics are contracted by the Los Angeles County Department of Mental Health (DMH) to provide mental health services to publicly insured children near the geographical area served by six clinical sites of the FQHC. The MH clinics provide individual, family, and group therapy utilizing a variety of evidence-based practices, psychiatry services, case management services, consultation services, and crisis intervention services. Of note, the FQHC and MH clinics regularly obtain patient authorization for sharing of patient information between providers at the respective organizations.

Potential participants (six PCPs, three MH providers, three FQHC administrators, and three MH program manager/directors) were identified by the medical and/or executive director of each organization and were invited to participate in a 45- to 60-min interview. We aimed to interview a variety of providers and staff at each clinical site, until reaching saturation, or the point at which no new themes emerge from interviews.9, 10 A letter of invitation and information sheet was given to invited participants; all invited agreed to participate. Potential participants received information about the study objective, time commitment, and eligibility criteria. All interviews were conducted by the study PI, a primary care clinician, with extensive experience in conducting qualitative studies (TC).11, 12 The interview guide was developed by the study team (TC, BB, LP, ML), led by a primary care pediatrician and researcher (TC), and a psychiatrist and anthropologist (EB). The interview guide was pilot tested with three primary care clinicians; adjustments to the protocol were made based on these practice interviews with clinicians not associated with the study clinical sites. The study team met weekly during interviewing to review and reflect on interview content. This study was approved by the University of California, Los Angeles Office for Protection of Research Subjects. Sixteen individuals were invited to participate, agree, and complete an interview.

Individual semi-structured interviews were conducted in English with 10 PCPs and staff at six clinic sites of the FQHC; participants included five pediatricians, one physician assistant, one referral coordinator, the pediatric medical director (also a PCP), the chief operating officer, and the director of nursing. Each of the six participating clinic sites of the FQHC employs three to six providers who work from 40 to 100% of a full-time position. One full-time provider from each clinic site was interviewed. This represents 16 to 33% of the potential full-time providers, depending on clinic site. The majority of the providers interviewed were women who have been working at the FQHC for more than 5 years and are representative of the larger population of providers. Three additional individual and small group interviews were conducted at the two local MH clinics. To represent a range of providers and staff at the MH clinics, two psychiatrists were interviewed (a staff psychiatrist and a medical director/psychiatrist) and a program manager together (one interview) and a therapist in an individual interview at one MH clinic. At the second MH clinic, the clinic manager and director were interviewed together in one interview and a therapist in an individual interview. All interviews were conducted between November 2013 and February 2014.

The interview protocol included the following topics: (1) current structure and processes for MH specialty referrals, (2) communication and coordination between primary care and MH referral sources, and (3) challenges within the current referral and care process for children referred to specialty MH. These topics were discussed from the perspective of both primary care and specialty MH providers and staff.

All sessions were digitally recorded, transcribed, and imported into qualitative data management software (Atlas.ti 7). All transcriptions were independently verified for accuracy of translation and transcription.

The codebook was created and revised in an iterative process. First, members of the research team (TC, LP, and EB) read a subset of interviews and discussed text examples relevant to the study objectives. Then, two experienced qualitative coders independently coded three randomly selected transcripts across stakeholder groups to develop a set of codes that marked the communication or coordination challenges encountered and solutions devised. Then, because communication or coordination challenges appeared to be distinct depending upon the phase of treatment, codes were developed to mark where in the treatment or referral process a challenge or solution had emerged. Codes were refined to differentiate phases of the treatment process. Next, within each phase, coding focused on comprehensive descriptions of types of communication challenges. In iterative discussions, the coders and research team made additional changes to the codebook, listing each code, a detailed definition, inclusion and exclusion criteria, and an example of the code from the data. Because we sought to see that two coders understood interview content related to phases and challenges in similar ways, we used percent agreement to test inter-rater reliability. In coding for phases in 10% of the data in three iterative phases, with review of discrepancies between each, two independent coders achieved agreement 74%, 84%, and 89% of the time. In coding for challenges, two coders achieved 84% agreement in a first round of coding 10% of the data; and after discussion of discrepancies, the two coders independently coded another 15% of the transcripts with agreement of 92%. One coder completed coding of all transcripts.

Four members of the research team reviewed all coded data using a constant comparative approach to discuss codes for each phase of the treatment and referral process and for each participant group (MH, primary care). Based on these discussions, two coders worked together to group codes within each phase into themes that described common communication challenges.13,14,15,–16 The coders refined emerging themes by comparing one interview to another and comparing MH with primary care to search for confirming and disconfirming cases. Themes were then compared across phases to finalize analyses, which were then reviewed by the full research team for comprehensiveness.

Results

Interview participants identified the referral and care process as a series of distinct phases (see Appendix), each under the primary responsibility of either primary care or MH. As described below, PCPs and MH providers had different reasons to desire or to seek communication and coordination at each phase of the referral and care process. Below, we describe each of the phases, as well as barriers and solutions to these communication and coordination gaps identified by participants.

Pre-referral Phase

This phase begins when the PCP considers a MH concern before an actual referral has been made. A PCP participant described, “after gathering information from the parents and seeing what kind of interaction the child may have through the exam, then...I talk through with the parent what I think might be going on and we develop a plan together….”

During this phase, participants reported that PCP-MH provider communication and coordination was minimal. Participants described the lack of referral options, the provider time required in patient visits to fully understand the extent or even existence of MH problems, and the barriers faced by parents to access services as key issues. Some providers described challenges of knowing whether a referral would be necessary, and seldom could identify a MH specialist to help with this decision.

Referral Phase

This phase begins after the decision has been made to make a MH referral and before the first visit to the MH clinic. At this phase, the most salient challenge identified was from the PCP’s perspective in trying to determine which referral sources were available and appropriate for the child. PCPs had to determine whether to access sources for mental health or behavioral/developmental sub-specialties and account for other child and family factors, such as available transportation and eligibility for services. Depending upon the nature of the problem, PCPs could not always turn to a single option or follow a clear algorithm to decide where to refer. Sometimes, PCPs were unaware of available resources. One PCP participant described finding out about a nearby mental health clinic: “The truth is, I didn’t really know about them [the MH clinic] and I don’t even know if many of my coworkers know about them...I found out about them because my son plays baseball with this other woman’s son and [she works there].”

During this phase, communication between PCPs and specialty providers was minimal. PCPs expressed frustration about the difficulty of sharing information between organizations and particularly the lack of feedback from the MH clinics they referred to. A PCP participant reported: “…many times it’s like I only know that they’re seeing my patient for MH services because yearly, the psychiatrist needs to have the PCP do a history and physical and do labs.” PCPs expressed interest in more communication with MH clinic during this phase. One PCP participant said: “A warm hand-off would be nice…if we could even, do a short phone call or something with someone to initiate [care].”

MH providers also expressed the need for more communication with PCPs. A therapist described a hypothetical situation in which easy information exchange would improve services: “I’ll get a phone call from one of the doctors and, they’ll [say] there’s a case that I really want to expedite quickly because perhaps, you know, the client is having major depression and Mom is really concerned. And so, at that point, I gather the name, we try to expedite the services.” A psychiatrist describes communication challenges and a possible way to improve communication “[We really want] to collaborate more with the primary care doctors and really, you know, for a variety of reasons that’s been really challenging…A quick [electronic] portal where we could give information but we could also get information. That would be wonderful.”

Post-referral Phase

This is the phase that occurs after the referral has been made and before the child has been seen by the specialty provider. The responsibility shifts toward MH providers and away from the PCP. The limited number of MH providers for publicly insured children and the long waitlist that families encounter to access services add to the challenges that PCPs face when they want their patients to receive specialty services in a timely manner. One PCP participant said: “It’s very hard. At one point [the MH clinic] was completely closed… ‘No, we’re not taking any new patients.’ And when parents called these certain places, they’ll always say there’s a waitlist or we’re full”.

The MH referral is a lengthy process that involves several steps before a patient can receive services. After the PCPs complete a referral form to the MH clinic, the MH clinic contacts the parent, usually by phone, to ask questions about whether or not the child is eligible to receive services. If the child is eligible for services, an intake interview is scheduled, usually in person, to help to determine the type of behavioral health program or therapy that would be indicated for the child. One solution that interview participants proposed to these challenges at this phase was to guide parents through the referral process; one pediatrician described “…the warm handoff is critical.”

MH providers expressed the need for a more collaborative approach to referrals with PCPs. A therapist describes this need as “just being able to communicate with the PCPs and collaborating with them on information [about] the client.” A MH program manager says “The [MH] doctor is concerned about their medical history and wanting to make sure they’ve had a physical within the past year and just finding out if there are any medical concerns.”

At this time in the post-referral phase, the MH provider takes a more active role in the child’s care. PCPs describe that families expect for them to have answers about the process during this phase. However, PCPs also report that they usually receive little to no information from the MH clinic and consequently are unable to answer family’s questions. One PCP participant expresses frustration that little information comes back to primary care about the status of the referral.

“Sometimes we get reports back, for example a one-page form that checks off what happened--if the patient is eligible for services, [or] the patient is not eligible, or [the MH clinic] attempted to contact this patient and couldn’t get a hold of them. So those are usually the three options. Very rarely do we receive any detailed reports back on the patient. I think, in the last five years, I’ve maybe seen two sent directly to the [primary care] health center.”

Pre-medication Phase

During this phase, the child and/or family receives therapy services, behavioral interventions, or other non-medical interventions. The PCPs report that they continue to be unable to address families’ concerns and questions about the MH treatment plan and process. A PCP participant said,

“When patients come back to us for a follow-up, we have difficulty receiving information back...we’re often relying on the parent to give us a report of what happened. And, sometimes, you’re playing telephone – the information doesn’t get back exactly the way it was communicated.”

Another PCP participant suggests that “a phone call would be nice [with the MH clinic] so that the long-term plan can be discussed.”

Medication Phase

After an initial trial of non-medical treatment, the patient may receive an internal referral to the MH clinic’s psychiatry department. PCPs report that many families express their preferences about medication treatment to them, and yet, the PCPs are unable to relay these preferences to the MH clinic staff or providers. At this phase, PCPs describe communication with the MH clinic as very sparse. One PCP participant says

“Once in a while, we’ll get a report back if the parent brings it to us…. many times, I only know that they’re seeing [a MH provider] because their name is on the medication bottle … But that’s kind of a very sketchy two-way communication, you know, ‘oh, they’re on a new medication, okay. Now I know.’ But there’s not very much communication that comes back.”

Despite this lack of communication, PCPs still field questions and concerns from families. As one PCP participant says

“I feel so bad sometimes… [Once] I actually tried to call one of the [psychiatrists] because the mom was really upset that he prescribed medication for her child and it was a stimulant. And she’s adamant that her child does not have attention deficit or hyperactivity. She’s like, ‘he’s on the wrong medication.’ I don’t even know.”

Stable Phase

In this phase, the referral and treatment process is complete and the patient is stable to be transferred back to primary care. At this phase, the lack of communication and coordination between PCP and MH providers generated considerable tension. MH providers during the stable phase are interested in transferring primary responsibility for the patient back to the PCP. As one MH provider says “If a kid has been on the same medication for a year and has been stable and has had no real problems over the course of the year, they’re really good to go.” However, PCPs reported not feeling comfortable taking these patients back for primary care management of their mental health condition due to a lack of comfort with psychotropic medications, and a lack of reliable communication channels with the MH clinic in case additional concerns arise about the child’s mental health care.

A MH specialty provider commented “Trying to get them back to primary care to track them after we’ve taken them as far as we can take them [has been] rough…almost non-existent. [PCPs] just didn’t feel comfortable taking on psychotropic meds. Period.” Another MH provider says “at the end of the day when it is sorted and [the patient is] doing really well… We don’t know how to transfer them out because there’s nobody on the receiving end who feels comfortable managing their medications.”

A primary care participant reported recently sharing an email from a MH clinic director at a PCP meeting:

“[The MH clinic director] was asking if they could send patients who were stable on medication therapy and no longer needed non-medical therapy. If they could send them back to primary care and have the same medications be prescribed by PCPs. As soon as I read the email out loud in our provider meeting I could see many providers shaking their heads no, like, vocally saying ‘no, I’m not comfortable doing it.”

Another PCP participant reported, “I personally just do not feel comfortable at all [prescribing psychotropic medications]. I never have prescribed psychotropic meds and I feel like I would need much more training if I’m going to start doing that.”

Both MH and PCP interview participants described frustration with this impasse for different reasons, yet both groups also reported that a lack of a collaborative relationship between the primary care and MH clinics was the source of the problem. A MH provider reported, “it puts us in a bind in the sense that to collaborate and take [referrals from PCPs], we need to have a back and forth flow. Because if we keep taking them we just don’t have the resources to take new [patients] and keep the old ones.” The PCPs described considerable concerns that they would not have MH support in accepting transfer patients. As one pediatrician says, “I feel like once the [MH provider’s] name is not attached to them anymore… I have this fear that we’re going to be trying to get in touch with them and then -- I mean, we really don’t know what to do, you know?”

Discussion

This qualitative, exploratory study demonstrates that both PCPs and specialty MH providers viewed primary care-mental health clinic communication and coordination as a key element of the MH referral and care process. While communication gaps are well-recognized, few studies have examined which gaps are most problematic for which stakeholders, and whether perceptions of those gaps overlap. This is important because different stakeholders may perceive different solutions, and this can stymie cooperation to address them. PCPs perceived challenges in all phases of the referral process while MH providers mainly perceived problems in the stable (final) phase.

In this qualitative analysis of challenges to communication and coordination between primary care and MH clinics, we identified a fragmented referral and care process. The lack of communication and coordination in any one phase seemed to contribute to the challenges in other phases as well. For example, transfer of a patient is difficult when PCPs have not been involved for the entire process and are not aware of patient’s progress or treatment. Initial referral is also difficult when there is little communication or feedback on previous referrals made. Interventions to improve communication and coordination at any one phase have the potential to improve the process at all phases. Other studies of primary care to specialty referral systems have reported similar findings, in terms of challenges of primary care-specialty communication within the referral process.17 A systematic review identified the following elements that need to occur in a primary care to specialty referral process: referral decision making, tracking the referral, specialty access, and information transfer to specialty provider and then back to primary care.17 The participants identified challenges in each of these steps of the referral process, but specific to their own processes.

These data can provide insights about the promise of strategies that can be used by FQHCs to improve the mental health referral process. For example, late-phase challenges, like those of the stable phase, which were of high concern to MH providers and may be addressed with investment in communication and coordination in the earlier phases. Strategies such as e-referrals or e-consults and phone consultations18,19,20,–21 may be adapted to this referral and care system to increase communication between primary care and specialty care providers early in the referral process and potentially lead to improvements for late phase challenges.

Although the number of states with carved-out system for mental health care have been recently on the decline (17 states in 2013, to 11 states in 2017), it is critical that work toward improving coordination within this system continues.22

The limitations of this study include a smaller sample of MH providers compared to PCPs, which may have led to an underrepresentation of the challenges faced by the former. Moreover, data come from a limited number of community clinics in two urban regions of a major metropolitan area that primarily provide care to Medicaid-insured children; thus, these findings may not generalize to children in rural areas, other clinic types, or to privately insured children. Additionally, we focused on barriers and challenges to communication and coordination at each phase, and did not specifically collect data on facilitators, including more structural and systematic facilitators of communication. There are multiple ways to integrate primary care and specialty MH care services, these include (1) coordinated care (the exchange of information on a routine basis when patients are in treatment in both settings), (2) co-located care (medical and MH services in same location), and (3) integrated care (having only one treatment plan rather than two from medical care and mental health care).23 This study did not assess participants’ perspectives on co-location or true integrated care but instead focused on how to optimize coordination across organizations during the referral and care processes.

To assess inter-rater agreement, we used percent agreement, as opposed to Cohen’s kappa coefficient; percent agreement is a robust and well-accepted measure of inter-rater reliability in many circumstances when kappa is not possible. Since we coded phases, the data did not present a defined set of text segments to be rated and the number of text segments for each phase was small.

Despite these limitations, the data provide rich information about challenges in key phases of the referral and care process.

In conclusion, the interview participants identified distinct phases within the specialty mental health referral and care process for children at a multi-site FQHC. Primary care and mental health clinic providers and staff identified different phases for priorities in improving communication and collaboration among their organizations. Interventions should focus on primary care-specialty mental health care communication and coordination through all phases of the referral and care process, with the goal of improving access at two key points: from primary care to the initial specialty mental health visit, and transfer of stable patients back to primary care for ongoing medication management.

Implications for Behavioral Health

PCPs and MH providers recognize the need for improved collaboration, but are motivated to improve collaboration at different phases of the MH referral and care process. This study suggests that strategies are needed to help PCPs and MH providers collaborate across the MH referral and care process to improve patient outcomes and develop shared perspectives on the clinical care of children with behavioral health needs. These findings will be used to develop an intervention that can potentially address the challenges identified at each phase of the referral and treatment process.