Abstract
Background: Primary care practices in underserved and/or rural areas have limited access to mental health specialty resources for their patients. Telemedicine can help address this issue, but little is known about how patients and clinicians experience telemental health care.
Methods: This pragmatic randomized effectiveness trial compared telepsychiatry collaborative care, where telepsychiatrists provided consultation to primary care teams, to a referral approach, where telepsychiatrists and telepsychologists assumed responsibility for treatment. Twelve Federally Qualified Health Centers in rural and/or underserved areas in 3 states participated.
Results: Patients and clinicians reported that both interventions alleviated barriers to accessing mental health care, provided quality treatment, and offered improvements over usual care. Telepsychiatry collaborative care was identified as better for patients with difficulty developing trust with new providers. This approach also required more primary care involvement than referral care, creating more opportunities for clinician learning related to mental health diagnosis and treatment. The referral approach was identified as better suited for patients with higher complexity or desiring specific psychotherapies.
Conclusions: Both approaches addressed patient needs and provided access to specialty mental health care. Each approach better aligned with different patients' needs, suggesting that having both approaches available to practices is optimal for supporting patient-centered care.
- Medically Underserved Area
- Medication Therapy Management
- Mental Health
- Patient-Centered Care
- Primary Health Care
- Psychiatry
- Psychotherapy
- Qualitative Research
- Referral and Consultation
- Rural Health
- Telemedicine
Introduction
Patients with complex mental health conditions like posttraumatic stress disorder (PTSD) and bipolar disorder often rely on care provided in primary care settings. The majority of patients with these mental illnesses do not receive care in specialty mental health settings.1,2 This is especially true in underserved areas where there are few mental health clinicians available. Mental health conditions are highly prevalent in primary care.3⇓⇓⇓–7 According to a systematic review,8 PTSD prevalence in primary care approximated depression, and another study found that bipolar disorder occurs in 0.5%–4.3% of primary care patients.3
Diagnosing and treating these conditions is complex, and the first-line treatment for PTSD includes trauma-focused psychotherapies9 and mood stabilizers for bipolar disorder.10,11 Yet, professionals familiar with these treatments are in short supply in rural or underserved communities.12 As a result, primary care clinicians are increasingly caring for patients with these disorders without psychiatrist or psychologist support. Primary care clinicians have reported feeling obligated but unprepared to diagnose and treat patients with PTSD and bipolar disorder.13 Consequently, only 10% of patients with these diagnoses receive adequate care in primary care settings.1,14
A promising approach for addressing this gap is to make use of telemedicine to facilitate necessary specialty consultation or management. While previous literature describes the effectiveness of collaborative care, including patient and clinician experiences,15⇓⇓⇓⇓⇓⇓–22 and there is a growing literature on telepsychiatry/telepsychology,23⇓–25 telepsychiatry collaborative care has not been extensively examined qualitatively. The Study to Promote Innovation in Rural Integrated Telepsychiatry (SPIRIT) is a pragmatic randomized comparative effectiveness trial that compared 2 approaches for treating patients with PTSD and/or bipolar disorder: telepsychiatry collaborative care (TCC)26⇓–28 and telepsychiatry/telepsychology enhanced referral (TER).29 Telepsychiatry and telepsychology use telecommunications technology to deliver psychotherapy and pharmacotherapy services to patients remotely.30 TCC used collaborative care, which made a consulting telepsychiatrist and on-site care manager (CM) available to primary care clinicians and their patients.18,31,32 TER did not include CMs, and telepsychiatrists and telepsychologists were directly available to patients. SPIRIT found that following treatment, patients with PTSD and/or bipolar disorder experienced substantial symptom improvements in both interventions.33
Given that both approaches were determined equally effective, this study identifies patient and clinician experiences and preferences regarding each approach. We interviewed clinicians and patients to identify attributes of these approaches that could help primary care practices decide which option is best for their practice and their patients.
Methods
Setting
Between November 2016 and June 2020, the SPIRIT trial was conducted in 24 practices that were part of 12 Federally Qualified Health Centers (FQHCs) serving underserved and/or rural patients in Arkansas, Michigan, and Washington.34 We recruited FQHCs without psychiatrists or licensed clinical psychologists on-site and patients who screened positive for bipolar disorder and/or PTSD who had not been prescribed psychotropic medications by a psychiatric clinician. Eligible patients were randomized to receive the TCC intervention (n = 508) or TER (n = 496), and participating primary care clinicians had patients in both arms.35 Telepsychiatrists and telepsychologists were credentialed and privileged to practice as FQHC providers and charted in their electronic health record.36 Study participants received up to 12 months of treatment and received interactive-video encounters at the practice.
Telepsychiatry Collaborative Care
This arm had 3 components: the telepsychiatrist who met with patients via interactive video to develop a diagnosis and treatment plan; the primary care clinician who prescribed medication suggested by the psychiatrist; and a CM (usually a registered nurse or licensed clinical social worker), who coordinated care, promoted treatment adherence, monitored treatment response, and delivered behavioral activation psychotherapy.37 In this approach, the primary care clinician retained patient care responsibility and the telepsychiatrist was a consultant.
Telepsychiatry/Telepsychology Enhanced Referral
This arm had 3 components: referral to a telepsychiatrist for initial diagnosis and treatment plan via interactive video, medication management televideo visits as needed, and an optional referral for psychotherapy based on patient interest, readiness, and need. Therapy was provided by a telepsychologist and included cognitive behavioral therapy (CBT)38,39 or cognitive processing therapy (CPT),40 both of which included standardized psychotherapy homework. Unlike the TCC arm, the telepsychiatrist delivered direct treatment to patients and monitored symptoms at the beginning of each interactive video session.
Sample
Patients participating in SPIRIT were purposively selected for this study to ensure variation by state, FQHC, treatment arm, diagnosis, and level of engagement (ie, engagers were defined as having 2 or more appointments with a telemedicine provider in the first 6 months of the intervention). A total of 153 patients were selected, and 49 were interviewed. We purposively selected nonengagers to participate because we wanted to understand factors that influence engagement. However, during interviews we learned they did not participate because of unrelated life circumstances (eg, moving to another state) or miscommunication issues where patients reported never receiving an invitation for services. Since nonengagers did not have experiences to recount, results were primarily from the perspective of engaged patients. Primary care clinicians who saw 9 or more SPIRIT patients and had patients in both arms were selected to ensure they had substantial experiences to recount. Clinicians were purposively selected to ensure variation on state, FQHC, medical degree, and the extent to which the clinician was familiar with the study. Thirty clinicians were selected, and 21 were interviewed.
Procedure
Researchers followed an interview guide (see Appendix) to conduct semistructured interviews. Patients were asked about their treatment experiences, access to mental health care, and satisfaction with care. Clinicians were asked about their experiences with each study arm and to compare the 2 approaches regarding patient benefit, care system burden, clinician benefit, and workflow impact. We approached these data inductively because there were few conceptual models appropriate for guiding a study of the integration of telemedicine in primary care. Interviews were conducted remotely by video or phone; averaged 60 minutes in length; were audio recorded, transcribed, and deidentified; and were entered into Atlas.ti for data management and analysis.41 This protocol was approved by the University of Washington, University of Michigan, and University of Arkansas for Medical Sciences institutional review boards.
Analysis
We used an iterative process of analyzing and interpreting data to develop findings.42 A 6-member interdisciplinary team with qualitative methods, primary care/behavioral health integration, and clinical expertise conducted inductive analysis concurrently with data collection. As transcripts were available, we analyzed data as a group. We discussed passages and developed an initial set of inductively derived codes that denoted the meaning of a segment of text. The team selected and labeled text with codes related to each treatment group, participating health professional roles, and key components of the intervention according to patient and clinician experiences. We continued this process as more data were collected. When definitions were clear, and we applied codes similarly during analysis meetings, a smaller team coded the remaining data. We continued to hold meetings to ensure reliability, discuss analytic questions, and identify emerging findings. We used an iterative process of moving between sampling, data collection, and analysis, which allowed us to use early findings to refine recruitment targets, as needed, and monitor for saturation.
To compare and contrast patient and clinician experiences within and across treatment arms, we filtered the data by intervention arm and developed data matrices43 to facilitate further comparison and discussion of emerging findings related to access to mental health care, perceptions about care quality and satisfaction, intervention suitability, and clinician involvement and skill development. Clinicians' reported preferences and experiences with the interventions did not vary by these characteristics, and patient experiences did not vary by diagnosis, state, or practice location. Preliminary findings were discussed and shared with the study's advisory boards, composed of clinicians, administrators, mental health advocates, and individuals with lived experience with PTSD and/or bipolar disorder. Board members found the findings aligned with their experience and identified areas where more details were needed.
Results
We completed 49 patient interviews; 29 patients received TCC and 20 received TER. Patient participants were distributed across states and screened positive for PTSD and/or bipolar disorder. Twenty-one clinicians completed an interview and varied by state, type of medical degree, and champion status (see Table 1).
Participant Summary Table
Table 2 summarizes clinician and patient experiences with TCC and TER. Five areas of comparison emerged: access to mental health care, care quality and satisfaction, alignment with patient readiness and suitability, clinician involvement and responsibility, and clinician skill development. Below, we describe patient and clinician experiences in more detail.
Study Findings by Telemedicine Approach
Both Interventions Increased Access to Mental Health Care
Clinicians and patients in both groups reported that the interventions increased access to specialty mental health care and alleviated barriers, regardless of rurality or patients' ability to pay for care. Clinicians considered both approaches as viable options for addressing patients' unmet mental health needs and, if given the opportunity, would eagerly adopt either approach.
I was very pleased with both arms. If one or the other service was offered to our community, boy, I would jump at either of them.…As far as we were concerned, it was access to psychiatric care, and that made us really excited about both arms. (Michigan, Clinician #1)
Both approaches were also considered convenient, reduced the stigma associated with visiting a specialty mental health clinic, and increased patient comfort by meeting with the psychiatrist in the familiar setting of their primary care practice.
[Patients] really like [coming] here and see[ing] us, or being able to come here to see the psychiatrist over telehealth.…They want to do everything here if they can. They're more open to get the help that they need if they find out it's here. (Arkansas, Clinician #9)
Both Interventions Provided High-Quality Care and Patient Satisfaction
Patients and clinicians were satisfied with the quality of care in both approaches. Patients described symptom improvements (eg, fewer medication side effects and fewer mania, depression, and PTSD symptoms) and satisfying relationships with new care team members (eg, telepsychiatrist, telepsychologist, CM). Receiving care from telepsychiatrists that were connected to respected institutions influenced patients' perception about the quality of their care.
Another reason I agreed to join it is because [university] was involved.…So that was a big part, too. I don't know if that sounds stupid, but I trust [university clinicians]. (Michigan, Patient #10, TCC Arm)
This was echoed by clinicians who recognized the differences between the treatment approaches and reported that both approaches were effective at treating patients and improving outcomes.
Each Approach Offered Differing Benefits Based on Patient Need
TCC and TER varied regarding their abilities to address specific patient needs. Patients reported that TCC offered benefits for patients with social anxiety and reluctance to work with new clinicians, as care was received from familiar care teams.
I like how it was all in one place. I have some social phobia in having to go to this place and that place. It's very stressful for me. And being able to stick to the place that I know, it's really a comfort…I like the people that I have been working with, and I don't want to have to deal with starting new. (Washington, Patient #25, TCC Arm)
Clinicians also reported that TCC worked especially well for these patients. Patients in this treatment group continued to receive most of their care from their primary care clinician, building on a previously established relationship, and considered the on-site CMs part of their primary care team.
Patients randomized to TCC may not have realized the CM was engaging them in therapy. They described receiving informal and practical advice from CMs but did not identify it as behavioral activation, a brief psychotherapy approach. Patients reported enjoying and benefiting from talking to CMs; they equated it to talking to a friend. Patients shared that CMs called them if they missed appointments, reminded them about their upcoming visits, and were readily available by phone. Clinicians reported additional benefits to the TCC model and CM role:
It gives the patient a sense that somebody at the office is coordinating their care…It lessens the workload on the patient if they feel like they have someone locally who's right there that can help them with anything they need. (Michigan, Clinician #10)
Benefits of TCC, according to clinicians, included providing care coordination and barrier mitigation and addressed a broad range of patient needs.
TER also offered unique benefits. Patients who may have benefited from CBT or CPT did not receive that type of treatment with TCC. Clinicians reported that TER was better suited for patients who desired or required CBT or CPT from a trained psychologist. Thus, there were patients whose needs were not fully met with TCC:
I have not found the quality of care that I feel is necessary to help me get past these problems. I liked [CM] as a person—super-nice person. And I know she means well…It was like I was going out to lunch with a friend and we're just shooting the breeze…[I never saw] anybody that would really give me—challenge me to work out my problems. (Washington, Patient #32, TCC Arm)
Clinicians speculated that these patients may have been better served if they had access to both treatment modalities:
Both [approaches] can be appropriate. There were certainly some patients, they kept trying to come back [to see me]. If we had the option, they would've been better for [TCC], and then others that wanted the more classic CBT therapy, would've maybe done better with [TER]…So I think they're just two different methods. They can be used both effectively within the clinic and to the same effect, especially if you're able to pick and choose and help guide patients to someone that might work better for them. (Washington, Clinician #6)
Clinicians saw the benefit of access to both approaches, which would allow them to align the treatment modality to patient needs, complexity, and readiness for therapy.
Telepsychiatry Collaborative Care Required Greater Clinician Involvement and Responsibility
Clinicians reported that TCC required more of their involvement than TER (eg, they saw patients more frequently and managed a fuller spectrum of their care needs). Clinicians reported having less contact with patients randomized to the TER arm, which operated like a traditional referral with reliance on the specialist to manage treatment. Clinicians varied in which arm they preferred. Some preferred TCC because it fostered greater involvement in care and facilitated better care continuity for patients.
We were in a better position for continuity with [TCC]. We were more involved in those cases I thought as family practice, so we were just more on top of it than the referral arm. (Arkansas, Clinician #5)
Others preferred TER for the ability to delegate medication management to a specialist and felt it was less time intensive, allowing clinicians to focus on patients' other needs. Clinicians did not describe additional work burden from either approach. While TCC had the potential to be more burdensome, CMs mitigated burden by monitoring patient symptoms and side effects and communicating and coordinating with the telepsychiatrist.
Telepsychiatry Collaborative Care Facilitated Clinician Skill Development
Clinicians reported that TCC created more opportunities for skill development than TER. Collaborating with the consulting psychiatrist was particularly helpful for complex patients, especially those unresponsive to treatment. Clinicians reported this collaboration fostered greater awareness, proficiency, and nuance in their ability to diagnose bipolar disorder and PTSD and to conduct a differential diagnosis between bipolar disorder and depression.
Since working with [telepsychiatrist], my prescribing practices now mimic hers. There is a level of comfort now, where I would be comfortable starting some medications for PTSD and bipolar while waiting to try to transition and get further help. Even just to simply try to recognize and diagnose is much more comfortable now. (Washington, Clinician #4)
Collaboration with the consulting psychiatrist was also described as enhancing clinicians' ability to treat these conditions (eg, improved awareness of medication side effects and their management, greater comfort prescribing specific medications they previously avoided, and confidence using new medication combinations and adjunctive medication therapies), all which were absent from TER.
Discussion
The clinicians and patients that participated in SPIRIT agreed that TCC and TER were effective at increasing access and providing high-quality care to patients in rural and/or underserved areas. However, these 2 approaches differed in the following ways: TCC was better suited to patients with social anxiety and reduced the need to start anew, while TER was better suited for complex patients and/or those that were in need of CBT or CPT. Clinicians recognized the benefits of both approaches for different types of patients. TCC required greater primary care clinician involvement, workload, and responsibility compared with TER, but it was not perceived as burdensome because of the CM role and because it facilitated clinician skill development (eg, diagnostic accuracy). While it was possible to develop enhanced skills from reviewing the psychiatry notes for patients randomized to TER, this required clinician motivation to study the telepsychiatry notes. In contrast, consulting with the TCC telepsychiatrist made skill and knowledge acquisition comparatively effortless.
Given the unique benefits of each approach, it seems likely that primary care practices would benefit from having access to both approaches, particularly if clinicians understood when to appropriately select and how best to align each approach with patients' needs, readiness, and complexity. TCC offers some benefits that could address a larger population of patients with mental health needs, while preserving psychiatrist time for patients who need and would benefit from direct care the most.33
Our results share some similarities to other work that highlights the importance the CM plays in facilitating patient engagement,44⇓⇓–47 as well as the development opportunities that collaborative care offers to clinicians wishing to expand their mental health medication management skill sets.48⇓–50 Despite its evidence base, there have been challenges to adopting and implementing collaborative care when treating mental illness outside of study trials, particularly for garnering clinician and staff buy-in, and securing the resources (time and money) needed to support implementation.51⇓–53 Although less studied, similar challenges to implementing collaborative care are likely to manifest when implementing a referral approach. These qualitative findings may mitigate many concerns clinicians might have about deploying these models in their practice.
The evidence base for collaborative care has focused on lower acuity mental health conditions, namely depression and anxiety.48,49,51,52 Fewer studies have focused on its effectiveness for patients with complex psychiatric disorders.16,17,28 While collaborative care trials have studied veterans experiencing PTSD,54⇓–56 few have focused on civilian populations in rural settings. Unlike these studies, SPIRIT clinicians compared 2 mental health interventions that differed from their usual procedures. The emphasis on the comparison offers valuable insight on benefits of each approach for certain patients, as well as the best way to support primary care clinicians treating acute mental health needs (eg, including CMs, levels of diagnostic support).
To make these important services available to practices and their patients, policy changes are needed to ensure the ability to pay for CMs and appropriate infrastructure to bill for telepsychiatrist time, as well as communicate, share, and exchange information with clinicians at external institutions.13 In addition, some states restrict nurse practitioner prescribing practices, which particularly impacts the feasibility of implementing a collaborative care approach.36,57 Future work could focus on scaling these interventions to a larger population through demonstration work or hybrid implementation trial. While our work did not focus on understanding implementation barriers to either model, surfacing these issues and ways to address them is another important area for future research.
The COVID-19 pandemic increased the need for mental health care but also expanded access to mental health specialists through telemedicine, opened billing codes, and facilitated investing in infrastructure needed to deliver telemedicine.25,26 The policy changes that expanded the use of virtual modalities to deliver mental health care could address the growing need for these services, but it is unclear whether these policies will endure.
Our study suggests that if these policy changes were continued and if resources to support telemedicine did exist, this would improve access to mental health care for areas where specialists are scarce. The patients and clinicians in the regions we studied appreciated the partnerships with specialists from the local state medical school and liked receiving care from them.
Our study has some limitations. Selection bias is possible as we experienced difficulty recruiting patients, particularly nonengaged patients. Most nonengagers in our sample did not participate at all; therefore, we missed the opportunity to study reasons patients may stop engaging in these care models. In addition, interviewed clinicians saw relatively small numbers of patients, so it is unclear how their experience would scale to a full patient panel. Recruitment was based on screening rather than primary care clinician referral, so it's unlikely that clinician buy-in influenced the number of SPIRIT patients on their panels. In addition, only primary care clinicians were interviewed, although telepsychiatrists, CMs, and telepsychologists were also involved in delivering the interventions and including their perspectives is an important area of research.23
Conclusion
TCC and TER were viewed favorably by primary care clinicians and their patients. This was encouraging as most primary care clinicians are not comfortable managing complex mental health disorders.13,58 Since each approach is best aligned with different patients' needs, our study suggests that having both approaches available to practices is optimal for supporting patient-centered care, and clinicians will need to understand which approach is best aligned with each patient's needs. To make these services broadly available to practices and their patients, local, state, and federal leaders will need to adapt a complex set of policies that obstruct the ability to deliver telemedicine.
Acknowledgments
We would like to thank the patients and staff from the participating FQHCs, our community and policy advisory board members, and Dr. Karen Drummond and Ms. Roopradha Datta for their contributions to data collection and analysis.
Appendix
Notes
This article was externally peer reviewed.
Funding: This work was supported by a grant from the Patient Centered Outcomes Research Institute (PCORI-PCS-1406-19295).
Conflict of interest: We do not have any conflicts of interest to declare.
To see this article online, please go to: http://jabfm.org/content/35/3/465.full.
- Received for publication September 15, 2021.
- Revision received January 10, 2022.
- Revision received January 19, 2022.
- Accepted for publication February 8, 2022.
References
In this issue
Jump to section
Related Articles
Cited By...
- Model of integrated mental health video consultations for people with depression or anxiety in primary care (PROVIDE-C): assessor masked, multicentre, randomised controlled trial
- Understanding the uptake of virtual care for first and return outpatient appointments in child and adolescent mental health services: a mixed-methods study
- The "Telehealth Divide"--Who Are the Underserved, and What Care Is Improved?