Implementation of an Adapted Collaborative Care Model

Our CoCM partnership began in 2018, with a goal to serve all Kosraeans with behavioral health concerns. Overall, 7.7% of Kosraeans report that their mental health is “not good” (4). Historically, however, treatment was sought only for manic or psychotic symptoms. Individuals receive care from a KCHC primary care provider (PCP), together with a nurse and community health workers from Kosrae’s Behavioral Health and Wellness Center. The KCHC CoCM team included PCPs, a care manager with a nursing background, additional nurses, community health workers, and clinic administrators. UH partners included a Micronesian program manager with expertise in Micronesian mental health, three psychiatrists with experience in CoCM or Micronesian mental health, and administrators.

Early interactions within the partnership focused on building trust among partners while developing shared goals. Over the 3-year partnership, many e-mail communications, several videoconferences, three visits to Kosrae by UH consultants, and a visit to Hawai‘i by KCHC providers occurred. The first Kosrae visit entailed an informal needs assessment to identify key stakeholders and organizations, behavioral health goals, infrastructure capacity, and available human resources. The next two visits involved training sessions for KCHC and other Kosraean medical staff on how to implement CoCM and workflows, brief behavioral interventions, and core mental health topics. The World Health Organization’s Mental Health Gap Action Programme (mhGAP) training was delivered during the third visit (5). Throughout the collaboration, regular videoconferencing meetings were held. Topics for these meetings included administrative issues, case reviews, training sessions, CoCM development, and workflow planning. On the basis of the needs assessment and ongoing discussions, a modified CoCM was developed and implemented at the end of the first year of the partnership.

A registry was created for KCHC clinics to track screening results for depression and anxiety (i.e., with the 4- and 9-item Patient Health Questionnaires and the 7-item Generalized Anxiety Disorder questionnaire), alcohol (i.e., Alcohol Use Disorders Identification Test), and drug and tobacco use (i.e., a one-question screen that assesses use). This registry also enabled tracking of active CoCM patients. During a 1-year period (March 2020–February 2021), the largest clinic screened 234 unique patients, and 24 patients screened positive in the screenings—20 for tobacco use, two for drug use, three for alcohol use, and none for depression or anxiety. On average, the CoCM registry had six active cases, detected by screening or clinical interview. Despite the low rate of positive screening results, the screening was still felt to be important, because the use of screening tools is best practice and fulfilled funding requirements.

The screening process also helped to normalize mental health concerns, primarily through the care manager, who made the screening procedure educational and conversational. She orally translated the questionnaires into Kosraean while also giving context for the questions, such as when asking about mood: “I want to ask about how you have been feeling lately. It’s normal to feel sad sometimes, but if you are feeling sad often, and it is getting in the way of your everyday activities, then we want to know about that, so we can give you treatment and help you to feel better. Over the past two weeks, how have you been doing?”

Weekly videoconferences included case consultations among the psychiatrist, care manager, and PCP; workflow discussions and modifications; and provider education. In 2020, an average of one new case per month was discussed, with symptoms including depression, anxiety, psychosis, mania, suicidality, tobacco use, and alcohol misuse. Substance use was typically detected through the screenings, whereas other conditions were detected via clinical interviews. Patients would typically be followed up until symptoms improved or the patient was lost to follow-up.

The KCHC team members decided which cases to discuss, an approach that was different from protocols used by many CoCMs, where registries guide case discussions to ensure no cases are overlooked. Because the active case panel was typically small, the team never established a regular registry review. The case consultations focused on diagnostic questioning, pharmacologic and nonpharmacologic treatments, potential adverse effects of medications, brief counseling techniques, psychoeducation for the patient and family, and measures to ensure safety.

Ongoing education via videoconferences was woven into case reviews and during times when no clinical discussion of cases was required. The educational videoconferences were attended by a psychiatrist and a core CoCM team consisting of a primary care physician and the care manager. The training sessions helped to reinforce behavioral health concepts. The psychiatrist used a mix of impromptu teaching and prepared materials for case presentation. By coupling the teaching topic with the active case review, these sessions became more relevant and practical. Teaching efficacy may have been compromised by the psychiatrist’s limited ability to prepare materials beforehand, because some topics were chosen on the basis of the week’s cases.

Additional monthly videoconference check-ins offered an opportunity to discuss what was working well and what adjustments to the program were needed to facilitate accountable care. These meetings solidified a culture of continuous quality improvement, with decision making and program changes primarily taking place through group discussions, which were felt to be the most culturally appropriate vehicle for such efforts.

Unique logistic and cultural considerations surfaced in this international partnership. Frequent communication via e-mail and videoconferencing was key, but this mode of communication required coordination among partners within the constraints of a 21-hour time difference. KCHC needed to install videoconferencing equipment and train staff on its use. This effort successfully met the technological requirements for the meetings, although some meetings were still missed or delayed because of poor Internet connectivity or other challenges.

Cultural differences affected both meetings and e-mail communication. Expectations in a U.S.-based health care system include that planned meetings start at the agreed-upon time; in Kosrae, it may take some time for all participants to gather for a meeting, and it is expected that the meeting will start only after all attendees have arrived. These different expectations led both partners to make adjustments: KCHC attempted to be ready at the planned start time, and UH adjusted for the possibility that meetings start and end later than scheduled.

Stigma of mental illness within the community, and to a lesser degree within the medical team, was a formidable barrier. Some Kosraeans consider only severe psychosis to be a mental illness, with individuals who experience psychosis being seen as different from the rest of the community. Recognizing conditions like depressive and anxiety disorders as common medical conditions entailed a framework shift. This perspective is necessary for Kosraeans to feel comfortable disclosing details about their mental health with the general medical team. The consultants and KCHC staff had several discussions about stigma. Health care providers became accustomed to speaking more openly about mental illness among themselves. Informal opportunities for local health care providers to lead the perspective shift in daily interactions within the community also arose. Language was adjusted to normalize mental health struggles, by using terms like “behavioral health” instead of “mental illness” and “sadness” instead of “depressed mood.”

Within KCHC’s CoCM, care providers responded differently to taking on behavioral health responsibilities: the team’s main PCP, who was the medical director, acclimated relatively easily to his role, whereas other PCPs were more hesitant to participate in CoCM care. All attended training sessions, but the medical director was the main physician participant in delivering CoCM care. This experience indicated the degree to which the mindsets of providers can influence CoCM provider participation. Before CoCM, the medical director had performed behavioral health care in which other providers had little experience. With more exposure and time, the providers could begin to weave this care into their daily practice.

The need for additional behavioral health training also arose around the care manager’s utilization of brief therapeutic techniques. The convening of the smaller active case panel provided time to discuss how the care manager was employing techniques like motivational interviewing and behavioral activation. The psychiatric consultants ultimately felt that teaching behavioral skills like these was feasible in the context of this CoCM. However, they also felt that more complex psychotherapies, such as cognitive-behavioral therapy, may require formal behavioral health training.

CoCM was implemented in an LMIC, the remote Pacific Island state of Kosrae, to enable provision of sustainable and readily available behavioral health care. Anticipating the expiration of funding, Kosraean health care workers needed to quickly and competently address mental illness in their community. Thus, the CoCM implementation focused simultaneously on creating an integrated system of care while planning for limited or no access to a psychiatric consultant in the future.

This project laid the foundation among the local workforce to better handle the community’s mental health needs by employing main CoCM elements, such as regular case reviews, together with additional training sessions. The Kosrae health care team was receptive to learning the CoCM process; however, it takes time to fully embrace a new model of care. In our observations, shifts in roles and improved competence and confidence, especially of the care manager and the program’s main PCP, occurred through the longitudinal process of skill and relationship building within the CoCM team.

Stigma of mental illness may be one of the larger, longer-lasting barriers to mental health care in Kosrae. Open discussion about mental health among Kosraean medical professionals was a newer practice, and acceptance of individuals with mental disorders may be slow in Kosrae. Stigma will likely continue to affect both staffs’ readiness to implement the CoCM program and patients’ willingness to disclose behavioral health symptoms. Open communication about mental health both within the CoCM team and with patients, along with successful CoCM outcomes for patients, could help to decrease community stigma by demonstrating that mental illnesses are treatable (3).

A limitation to this program’s implementation was the lack of clinically useful measured outcomes. While the screening tools were being used, Kosraean team members suggested that Kosraeans did not want to answer formal mental health questionnaires and did not understand what the questions were asking, even when translated into the local language. This observation suggested a gap between the perceived utility of measurement-based behavioral health care in the experience of practitioners in the United States and other Western countries compared with experiences in other CoCM settings. It is unclear whether these measurement-based tools developed in Western countries will be as useful in models adapted to settings such as Kosrae. This question would be a helpful area for further study to give guidance to future CoCM programs in similar settings.

The CoCM adaptation described here offers an alternative to training-only models of behavioral health capacity building in resource-limited settings. Our team’s ongoing relationship has allowed CoCM refinement for more effective delivery of integrated behavioral health care. This program is also an example of CoCM delivery with international consultants. In settings that lack psychiatrists, international mental health experts may be the best option for regular clinical consultation and team-based care, if the technical capabilities exist and if specialists gain an appropriate cultural understanding of the population. The length of the relationship and frequency of interactions enabled the specialists to better understand Kosraean culture, which resulted in more appropriate recommendations and advice to the care team. Challenges to a long-term relationship can arise from several sources, however, including limited funding, changing personnel, and competing clinical priorities.

Overall, implementation of CoCM has been deemed successful in Kosrae, because the team’s partnership has been active for 3 years and the principles of CoCM are in current use. Lessons learned from this integrated behavioral health care implementation include that it is important to prepare for a long implementation period to allow for building of adequate rapport and addressing challenges. Additional lessons included that stigma should be acknowledged and addressed and that CoCM specifics can be adapted to an LMIC setting without compromising the model’s core principles.

The authors thank Caitlin Engelhard, M.D., Ph.D., Deborah Goebert, Dr.P.H., Earl Hishinuma, Ph.D., and Tulpe Timothy, A.D.N., for their contributions to this study and column.