Posttraumatic stress disorder (PTSD) and bipolar disorder are prevalent within primary care but are often undetected or undertreated. Within primary care clinics, 9%–23% of patients screen positive for PTSD (
1–
6). However, without systematic screening, only 11%–18% of primary care patients with PTSD are given a diagnosis (
7,
8). Up to 10% of primary care patients screen positive for bipolar disorder (
9). Undiagnosed bipolar disorder can result in prescription of antidepressants without a mood stabilizer, contrary to clinical guidelines (
10). PTSD and bipolar disorder often co-occur with general medical conditions that are treated in primary care (
11,
12) but may be difficult to treat when mental health symptoms are poorly managed.
Psychiatric disorders are more prevalent among Medicaid enrollees and those in poverty (
13), and access to mental health services for these populations may be more limited. Federally Qualified Health Centers (FQHCs) provide services to Medicaid enrollees and underserved populations yet may not provide sufficient access to psychiatrists or licensed clinical psychologists (
14) and struggle with providing patients access to community-based specialty mental health services (
15–
17).
Virtual psychiatric consultation can mitigate access challenges and support clinical teams and their patients with PTSD and with bipolar disorder. We studied telepsychiatry collaborative care (TCC), delivered as one arm of a pragmatic clinical trial to patients with PTSD or bipolar disorder (
18). Collaborative care addresses patients’ mental health needs in primary care with support from a care manager (CM) and psychiatric consultant (
19). Collaborative care has been extensively studied, has been found to be effective for treating depression and anxiety in primary care (
20–
23), and has had varied success in treating PTSD through primary care, with greater success when models included brief CM-delivered psychotherapy (
24,
25). Collaborative care models for bipolar disorder have been tested in specialty care settings in randomized controlled trials (RCTs) (
26–
28). However, although observational studies of collaborative care for bipolar disorder have occurred in primary care settings (
17,
29–
31), no previous RCTs have been conducted. In our RCT, TCC included care coordination and brief psychotherapy performed by the CM and a diagnostic and treatment planning visit with the consulting telepsychiatrist. TCC is effective for managing PTSD and bipolar disorder in primary care, increases access to high-quality mental health care that patients and their primary care clinicians appreciate, and could be a beneficial alternative to specialty care referrals (
18,
31). Further study is needed on how primary care clinicians experience TCC to better inform future implementations of the model and to assist clinic leaders and primary care clinicians with integrating general medical and mental health care for patients with serious mental illness. Findings could also identify important adaptations to TCC to make it more appealing to primary care providers.
Methods
Setting
The Study to Promote Innovation in Rural Integrated Telepsychiatry (SPIRIT) trial was conducted in 24 community health centers (CHCs) associated with 12 FQHCs in underserved areas of Arkansas, Michigan, and Washington between 2016 and 2020 (
32–
34) (
Table 1). FQHCs were recruited if they did not have psychiatrists or licensed clinical psychologists onsite. Patients were recruited if they screened positive for bipolar disorder or PTSD, excluding patients to whom psychiatric clinicians had already prescribed psychotropic medications. Patients were randomly assigned to receive TCC (N=508) or referral for direct treatment by a telepsychiatrist or telepsychologist (N=496) for up to 12 months. Our analysis in this article focused on clinicians’ experiences with the TCC arm only.
Model
Patients in TCC saw a CHC-based CM, usually a licensed clinical social worker or registered nurse, who coordinated care with the patient, primary care clinician, and consulting telepsychiatrist. The consulting telepsychiatrist conducted an initial videoconferencing visit with patients to establish a diagnosis and treatment plan, because these steps were considered more nuanced for this patient population than for patients with less serious psychiatric disorders (e.g., depression or anxiety). The consulting telepsychiatrist communicated the diagnosis and treatment plan, typically through the electronic health record (EHR) or CM, to the primary care clinicians, who maintained patient oversight, including prescribing medications. The CM met with patients regularly in person and by telephone to monitor treatment response, promote treatment adherence, and deliver brief behavioral activation psychotherapy (
35). Behavioral activation has been used in collaborative care since its inception (
23) and has been studied in primary care for PTSD (
25) but not for bipolar disorder. The clinician and telepsychiatrist consultant could communicate via telephone, EHR, or the CM. The CM and telepsychiatrist consultant regularly reviewed a Web-based care registry during their virtual case consultations to discuss patient progress. TCC differed from the telereferral study arm, in which the telepsychiatrist prescribed medications and the telepsychologist delivered psychotherapy—cognitive-behavioral therapy (CBT) (
36,
37) or cognitive processing therapy (CPT) (
38).
Sample
In total, 127 primary care clinicians participated in the SPIRIT trial. Clinicians designated as clinician champions and those with more than nine participating patients were included in our sample (N=37) to ensure that participating clinicians were familiar with the intervention. Champions acted as advocates for both TCC and the telereferral arm at their CHC. We selected clinicians to ensure heterogeneity of state, CHC, and academic degree by using an iterative sampling approach. We selected a small number of clinicians for interviews, analyzed the data, and used insights to inform subsequent sampling decisions. This process continued until we reached saturation, which occurred after 22 interviews. Seven of the 29 clinicians contacted either did not respond or declined to participate.
Data Collection
We conducted semistructured interviews during the last year of the intervention (October 2019–June 2020) to ensure clinicians had adequate experience with TCC to reflect on. We used an interview guide to conduct interviews and asked clinicians to describe their role in TCC, experiences communicating and working with the TCC team, and the extent to which TCC changed how they managed patients’ mental health care. Interviews lasted 30–60 minutes. Interviews were conducted by using videoconferencing software or a telephone and were audio-recorded with permission, professionally transcribed, deidentified, reviewed for accuracy, and entered into Atlas.ti, version 8, a qualitative software package. All procedures were approved by the University of Washington, University of Arkansas for Medical Sciences, and University of Michigan institutional review boards.
Analysis
A diverse team of researchers with qualitative expertise and backgrounds in primary care, integrated care, and health services research (T.J.H., J.D.H., L.I.S., L.H.T., M.N.D., D.J.C.) analyzed the interview data, initially as a group. Team members then independently analyzed an interview to discuss at group meetings. Key passages were described and labeled, leading to an initial set of codes. This process continued until the group developed a code list with clear definitions that members applied consistently. As the code list was updated, we ensured that changes were applied to all transcripts. The team transitioned to analyzing the remaining data, with at least two people analyzing the remaining transcripts, and discussing discrepancies with each other and the larger group. Preliminary findings were discussed and shared with the larger study team and the study’s community advisory board, comprising mental health advocates and individuals with lived experience with PTSD, bipolar disorder, or both.
Results
Clinicians who participated in the study (race-ethnicity data were not available) had one of four different medical degrees and came from 12 CHCs across the three study states (
Table 2). Both clinician champions (N=6) and other clinicians (N=16) were interviewed.
Findings were focused on four key areas. First, clinicians—especially early-career clinicians—appreciated support from the telepsychiatrist consultant both in medication management and in promotion of ongoing learning and skill development. Second, clinicians identified the CM role as critical to team communication and communication with the patient. Third, clinicians valued TCC because it included and supported them in improving treatment of patients’ mental health conditions, opening opportunities for clinicians to also focus on co-occurring general medical conditions. Fourth, clinicians did not find the model burdensome in large part because of the CM role, which reduced clinicians’ administrative burden while increasing patient engagement. In addition, TCC team members treated patients with previously unmanaged bipolar disorder, which also contributed to a reduction in burden for providers.
Support With Medication Management
Primary care clinicians were responsible for medication management and found that the medication selection and dosage support from the telepsychiatrist consultant expanded their medication management confidence, comfort, and skills. Clinicians leveraged guidance from the telepsychiatrist consultant when prescribing recommended medications, managing laboratory work, and sharing possible side effects with patients. Clinicians reported feeling comfortable following recommendations, which were detailed and well laid out.
I would be doing the medication monitoring and initiated the medication on the recommendation of the psychiatrist. They did really lay out the recommendations beautifully, though. . . . It was, “We’ll start them here. We’ll move up to here. They have these side effects. We can decrease here. Check these labs.” (clinician 3, Michigan)
Clinicians appreciated the telepsychiatrist consultant’s availability to them and their patients, although such consultations were rare. For patients with complex cases, some clinicians would have preferred a referral option so that the telepsychiatrist could prescribe medications directly to those patients.
Opportunities for Continued Learning
Working with the telepsychiatrist consultant (e.g., reviewing the telepsychiatrist consultant’s notes) improved clinicians’ ability to recognize, diagnose, and treat PTSD and bipolar disorder and to monitor patients’ engagement in treatment. “I learned how underdiagnosed bipolar disorder is. And I’m starting to now see that in several more patients where I probably would’ve never even thought to look for it” (clinician 11, Michigan).
Early-career clinicians in particular appreciated the learning opportunity this partnership offered. Continued learning areas were also highlighted as opportunities for improvement for future implementations; these areas included making introductions with the telepsychiatrist consultant earlier to improve comfort with reaching out for support and offering clinic-level education opportunities.
Essential Role of the CM
Clinicians reported that CMs were critical to supporting communication within the team, coordinating care, monitoring patient progress, and communicating regularly with patients, which improved clinical care. Although clinicians reviewed the telepsychiatrist consultant’s notes in the EHR, they had minimal direct synchronous interaction with the telepsychiatrist consultant and largely relied on CMs for this communication. CMs ensured that information on diagnosis, medication management, and side effects was clearly communicated to the clinician and patient while also relaying treatment progress to the telepsychiatrist consultant.
[The CM] probably had the bigger role in this because she would have follow-up with the patients and then report back to the psychiatrist. . . . They seemed to have more of a collaborative role even than I. I was sort of getting, “Here are the recommendations,” . . . and then I would just kind of act on that. (clinician 22, Washington)
Clinician communication with CMs varied in intensity, frequency, and mode, including in-person check-ins, EHR communication, and brief huddles. Closer physical proximity between the CM and clinicians led to greater ease of communication.
CMs communicated regularly with patients through telephone and CHC visits. They helped coordinate visits, offered behavioral activation, and mitigated social and economic barriers to care.
If she [the CM] wasn’t there, I don’t know how successful this would have been, just because she was reminding patients of their appointments. If they needed to fill a gas card or a bus ticket, she’d make sure they [had] one. She’d also ask them how the medication is going because many times, patients, if they don’t like how they feel initially, they’ll just stop and they won’t contact you. (clinician 3, Michigan)
Patient Engagement and Oversight
Primary care clinicians valued being involved in patients’ mental health care and receiving the additional support TCC afforded. Increased patient engagement and improved management of mental health conditions opened opportunities for clinicians to address patients’ medical concerns. For example, bringing patients in for regular psychiatric medication refills could improve engagement in medical care and open opportunities for clinicians to address chronic medical concerns longitudinally and continuously: “I thought [TCC] helped [patients’] overall care because, quite often, they had other chronic issues that we were dealing with. Trying to bring everything together and get the full picture, I got a lot more information and could try to help them” (clinician 15, Washington).
Clinicians perceived that with TCC, patients could avoid long waits for an initial psychiatrist visit and experience shorter wait times for medication dose changes compared with what patients typically experienced when working with community psychiatrists. With TCC, the CM could engage patients on the day a need was identified and schedule with the telepsychiatrist consultant soon after.
I think just being able to get patients in front of services at the time that they need [them] has really made more people utilize and access [them], because a lot of people, if you maybe schedule them an appointment [in] a week or 2 weeks, maybe won’t come back. Life will get busy or they’ll decide, “Hey, maybe I’m doing fine.” (clinician 6, Washington)
Low Perceived Burden
The TCC model was not perceived as burdensome, in part because of the small caseload of patients with PTSD or bipolar disorder among most clinicians and because of the supportive and coordinating role of the CM. By assuming responsibility for managing communication and tasks related to care coordination, the CM was pivotal in ensuring that the additional work of managing psychiatric needs was not burdensome to the clinician. Clinicians commented about their burden being relieved because of collaboration with the CM. If the CM could not manage the administrative work for medication management, on the other hand, clinicians were burdened by these administrative tasks. In addition, TCC reduced patient care burden because clinicians were working and visiting with patients whose bipolar disorder was better managed: “If you’ve got a patient [who] has unmanaged bipolar [disorder], they can take up a lot of your time. And I saw that a lot of these patients did really well and were able to have really productive medical visits” (clinician 14, Arkansas).
Discussion
Primary care clinicians in underserved areas valued TCC and offered positive feedback about their experiences treating patients with PTSD or bipolar disorder. Clinicians gained confidence in their roles of identifying these conditions and prescribing medications with support from the telepsychiatrist consultant’s detailed prescribing recommendations and other communications. Such support offered the opportunity for continued learning among the clinicians. CMs supported clinicians in medication management and patient engagement while also offering behavioral activation to patients. Clinicians valued the oversight TCC offered them in managing patients’ general medical and mental health conditions simultaneously, with psychiatric support from the telepsychiatrist consultant. Clinicians also generally did not perceive the model to be burdensome, in large part because of CM support. These findings suggest that there are multiple facilitators of widespread implementation.
Our findings align well with the literature, although studies of collaborative care differ in how much burden clinicians experience. The views of primary care clinicians differ on how much work is involved in collaborative care; some reported added burden, whereas others did not or said it reduced their workload (
39). Several studies found that colocation of the CM and clinicians was a facilitator of successful implementation of collaborative care because it increased the regularity of contact between the providers (
39). Close proximity was also highlighted by clinicians in the SPIRIT trial. Collaborative care has not been studied extensively in the care for patients with PTSD or bipolar disorder, so our findings add to a growing body of literature (
17,
26–
31). Primary care patients who have depression, anxiety, insomnia, or other problems may also have PTSD (
24) or bipolar disorder (
17,
29–
31). Models of collaborative care for PTSD in primary care are evolving, with some offering psychotherapy within the model and some facilitating outside referral to psychotherapy (
25). TCC included CM-delivered behavioral activation, which may add to the effectiveness of the model, thereby contributing to a lower perceived burden among clinicians. Importantly, clinicians in the SPIRIT trial were willing to manage patients with these complex mental health needs, likely because of the support from a broader care team (
40,
42). Finally, one other recent study has highlighted the benefit of the collaborative care model in providing continuing education for rural providers (
41).
TCC increased the accessibility of evidence-based care, was as effective as more resource-intensive telereferral to specialty mental health services (
18), and was well received by patients and clinicians (
31). CMs’ outreach helped engage patients in care more quickly than specialty mental health care in the community, which often has long wait times. In a study that used a “mystery-shopper” method in an urban area, whereby research assistants called the offices of 150 randomly selected local psychiatrists and asked for the earliest appointment date for a family member, only 7% (N=11) of psychiatrists were able to make an appointment within 2 weeks, and only 77% (N=115) could be reached at all (
43). Primary care providers are therefore forced to treat patients with bipolar disorder because of limited access to specialty care as well as stigma about seeking mental health treatment (
44).
There were limitations to using TCC, however. Some clinicians felt more comfortable having a psychiatrist manage medications directly for some of their patients with complex cases. Primary care nurse practitioners in Arkansas were not legally able to prescribe psychotropic medications, which may pose a TCC implementation barrier in some clinics. Also, some patients might benefit more from intensive psychotherapies, such as CBT or CPT (
31,
36–
38,
45). Therefore, referral to local specialty mental health providers or telereferral (e.g., with academic medical centers) may be an important complement to TCC in order to address these patients’ needs and thereby improve implementation of TCC in primary care. Offering TCC in primary care can be better for those with social anxiety or pressing social needs (
31) and may be preferred when stigma, urgent comorbid conditions, and poor access to specialty mental health care are barriers to care (
25,
43). Clinics may adapt TCC by offering a stepped care approach (
46,
47) that refers to specialists those patients whom the clinicians are not comfortable managing or who do not respond well to TCC. Clinician involvement when implementing TCC can ensure that patient and clinician needs are the foci when developing workflows and resources to support clinicians.
This study had several limitations. We waited until the final year of the study to conduct clinician interviews to ensure that clinicians had adequate experience with TCC. This delay may have contributed to recall bias for a few providers, weakening potential reflections on how TCC was experienced. The delay in recruiting resulted in turnover of some clinicians in these rural clinics. We sampled clinicians with more experience with the program to better understand their experiences with care and thus were unable to examine experiences of clinicians with fewer patients. However, we expect that these clinician experiences are representative of experiences of those with fewer patients in the study because all patients were recruited on the basis of positive screens. Patients were not recruited via clinician referral. We also were unable to examine experiences of those who refused an interview or did not respond to outreach (N=7 of 29 contacted), who may have had different experiences that may be important to explore.