Collaborative care clinician perceptions of computerized cognitive behavioral therapy for depression in primary care

Abstract

In Veterans Health Administration’s (VA) Primary Care–Mental Health Integration (PC-MHI) models, primary care providers, care managers, and mental health clinicians collaboratively provide depression care. Primary care patients, however, still lack timely, sufficient access to psychotherapy treatment. Adapting PC-MHI collaborative care to improve uptake of evidence-based computerized cognitive behavioral therapy (cCBT) may be a potential solution. Understanding primary care-based mental health clinician perspectives is crucial for facilitating adoption of cCBT as part of collaborative depression care. We examined PC-MHI mental health clinicians’ perspectives on adapting collaborative care models to support cCBT for VA primary care patients. We conducted 16 semi-structured interviews with PC-MHI nurse care managers, licensed social workers, psychologists, and psychiatrists in one VA health-care system. Interviews were audio-recorded, transcribed, coded using the constant comparative method, and analyzed for overarching themes. Although cCBT awareness and knowledge were not widespread, participants were highly accepting of enhancing PC-MHI models with cCBT for depression treatment. Participants supported cCBT delivery by a PC-MHI care manager or clinician and saw it as an additional tool to engage patients, particularly younger Veterans, in mental health treatment. They commented that current VA PC-MHI models did not facilitate, and had barriers to, use of online and mobile treatments. If effectively implemented, however, respondents thought it had potential to increase the number of patients they could treat. There is widespread interest in modernizing health systems. VA PC-MHI mental health clinicians appear open to adapting collaborative care to increase uptake of cCBT to improve psychotherapy access.

INTRODUCTION

Recognizing that one in five U.S. Veterans have depression [1], the Veterans Health Administration (VA) instituted Primary Care–Mental Health Integration (PC-MHI) collaborative care models to improve access to depression treatment in primary care clinics nationally [2]. In PC-MHI collaborative care models, embedded mental health clinicians and primary care providers (PCPs) jointly treat VA primary care patients with mild-to-moderate depression, facilitated by nurse care managers who provide care coordination for this patient registry [3]. PC-MHI more commonly enables clinicians to support PCPs in prescribing medication, but collaborative care models struggle with providing patients with timely, sufficient access to psychotherapy [4–6]. Although VA primary care patients with depression may prefer psychotherapy to medication [7], substantial caseloads limit clinician capacity to provide effective face-to-face individual or group psychotherapy [6]. It is unclear whether PC-MHI clinicians would perceive computerized cognitive behavioral therapy (cCBT) to be a viable, effective solution to increasing access to psychotherapy for VA primary care patients.

Though not routinely used in U.S. primary care practice, cCBT can be accessible 24/7 via the Internet on desktop computers and mobile devices and has been shown to effectively treat depression in 40 trials [8–10]. When delivered with modest asynchronous clinician support (approximately 2 hours total), cCBT has fewer dropouts [11], achieves greater effectiveness [9,12], and is noninferior to face-to-face psychotherapy [12–14]. cCBT is currently recognized as an alternative treatment option in major depressive disorder clinical practice guidelines at the VA [15] and is offered free to Veterans. Yet, VA’s sponsored cCBT for depression [16] has not been paired with systematic VA clinical support, reaches only a thousand Veterans monthly nationwide, and is completed by approximately 12 percent of users. In addition, VA primary care patients are notably older and have more chronic health conditions than average cCBT users [10]. Adapting PC-MHI’s collaborative care model to incorporate cCBT with VA clinician support (hereafter referred to as “cCBT-enhanced collaborative care”) has the potential to improve widespread access to psychotherapy and, in turn, quality of depression care.

Although patients have generally found cCBT to be acceptable [17,18], we know little about U.S. mental health clinicians’ perceptions of cCBT-enhanced collaborative care. Uptake of evidence-based cCBT is relatively low [19]; one factor identified by two UK studies (i.e., a survey of 329 mental health clinicians [20], a qualitative study of nine clinicians [21]) has been clinician misconceptions about cCBT efficacy relative to face-to-face therapy. Other concerns have been elicited from mental health clinicians, including confidentiality of patient information and liability issues [22]. In an Australian survey of 142 mental health clinicians, greater knowledge about cCBT’s efficacy, on the other hand, can promote more positive attitudes toward cCBT [23]. Although most cCBT studies were conducted within research trials, one UK study surveyed 33 practicing therapists in one inner-city clinic that offered cCBT through National Health Service’s Improving Access to Psychological Therapies program and similarly found that misconceptions regarding cCBT’s evidence base were prevalent, possibly contributing to low uptake [24].

Even less is known about cCBT use and mental health clinicians’ perceptions of its utility in primary care settings. In general, few cCBT studies have been conducted among primary care patients [25], who are not as easily recruited for Internet-based mental health treatments [26,27]. Swedish researchers conducted 11 interviews with primary care-based mental health clinicians and found overall positive experiences and attitudes toward cCBT. They also noted practical implementation barriers, including overall poor integration of mental health services into primary care settings and low patient uptake of cCBT [28].

Because cCBT-enhanced collaborative care holds promise for improving access to psychotherapy in the primary care setting but has been difficult to fully implement, we propose studying mental health clinician’s attitudes toward pairing of VA’s free-to-Veterans cCBT program (Vets Prevail) [16] with systematic depression care manager support to improve its uptake among untreated Veterans with depression in primary care. This study is among the first in the U.S. to examine primary care-based mental health clinicians’ thoughts and concerns about adapting PC-MHI collaborative care to support cCBT use among Veterans with depression, in preparation for implementation and evaluation of this care model.

METHODS

Setting

VA Greater Los Angeles (GLA) Healthcare System, consisting of two ambulatory care centers, a tertiary care facility, and eight community-based outpatient clinics, is the largest healthcare system within the VA. Most GLA clinics have access to PC-MHI care [29]—more than 1 in 10 GLA primary care patients receive these services. VA GLA cares for 1.4 million Veterans residing throughout five counties: Los Angeles, Ventura, Kern, Santa Barbara, and San Luis Obispo. Women, younger Veterans (i.e., Operation Enduring Freedom or Operation Iraqi Freedom [OEF or OIF]), and rurally located Veterans together constitute a minority (approximately one-tenth) of VA primary care patients in GLA. More than half of GLA primary care patients have a service-connected disability. Chronic physical health conditions are common among GLA primary care patients, particularly hypertension (38 percent), obesity (33 percent), and diabetes (24 percent). Rates of mental illness diagnoses, such as posttraumatic stress disorder (PTSD), are notably higher among GLA primary care patients than national averages. Finally, rates of homeless patients served in primary care are over ten times higher in GLA compared to national VA averages (Table 1).

Table 1 |.

Veterans Health Administration (VA) primary care patient characteristics

	VA Greater LA (n = 73,099) (%)	All VAs (n = 5,558,419) (%)
Women	8	9
Rural	11	33
OEF or OIF veterans	11	11
Service-connected veterans	58	58
Homeless primary care	5	0.3
Hypertension	38	46
Obesity	33	37
Diabetes	24	27
Posttraumatic stress disorder	20	15

Data obtained from VA Support Service Center (VSSC) reflecting the month of January 2019. Abbreviations: OEF or OIF = Operation Enduring Freedom or Operation Iraqi Freedom.

Study design

We used a case study approach [30] to study cCBT-enhanced collaborative care treatment for depression as perceived by PC-MHI care managers and clinicians within VA GLA.

Participants and recruitment

We conducted qualitative semi-structured interviews with GLA PC-MHI clinicians to explore their perspectives on adapting collaborative care to support cCBT use in VA primary care patients, from August to September 2018. Participants represented all professions among GLA primary care-based mental health clinicians, including PC-MHI nurse care managers, licensed social workers, psychologists, and psychiatrists. GLA PC-MHI leaders were asked to recommend PC-MHI clinicians and subject matter experts to contact for interviewing. A member of the project team then contacted the suggested clinicians via VA e-mail to introduce the project, determine eligibility, and schedule participation if appropriate. We achieved 57 percent voluntary participation via e-mail recruitment of 28 GLA PC-MHI clinicians and subject matter experts identified by PC-MHI leadership.

Data collection

The study team developed a semi-structured interview guide with probes, which was refined by clinical and research experts before use. Domains of inquiry included current care for Veterans with depression seen in PC-MHI clinics, knowledge of and experiences with online or mobile treatments for depression, and perspectives on adapting collaborative care to support cCBT use, including concerns, target patient population, and suggestions for improvement. The study team sought reactions to cCBT generally and did not elicit feedback for any specific cCBT program.

The second and lead authors (K.D., L.L.) conducted all interviews either over the telephone or in-person at a day and time most convenient to the participant. This evaluation was conducted as a quality improvement (non-research) effort at the VA, as determined by the VA GLA institutional review board. Participants were read an informed consent script before the start of the interview but were not required to sign consent forms.

Data analysis

Interviews were audio-recorded and transcribed verbatim for analysis. We imported de-identified transcripts into qualitative analysis software (ATLAS, ti v.8; Scientific Software Development GmbH, Berlin) for targeted coding of data specific to adapting collaborative care to support cCBT use. We further refined the preliminary codebook after coding a subset of data. Transcripts were coded by the second author (K.D.) using the constant comparative method [31] and checked by the first and senior authors (L.L., A.H.); discrepancies were resolved through discussion and consensus. Coded segments were then analyzed to identify emergent themes.

RESULTS

A total of 16 interviews were conducted and each lasted an average of 47 min. All participants either provided care in or supervised PC-MHI clinics. They reported training and working in the VA for an average of 5.4 years. Participant characteristics are summarized in Table 2.

Table 2 |.

Veterans Health Administration (VA) Greater Los Angeles (GLA) Primary Care–Mental Health Integration (PC-MHI) clinician participant characteristics

	Number of clinicians, n = 16
Roles
Nurse care managers	2
Licensed social workers	2
Psychologists	7
Psychiatrists	5
Gender
Women	11
Men	5
Years working in VA
0–1	1
2–4	4
5–6	6
7+	5
Patients with depression
<50%	6
50%–74% (“majority”)	4
75+%	6
Use of any online or mobile treatments
Yes	10
No	6

The following sections present qualitative findings on the context of current depression care at GLA, perspectives on virtual mental health treatments, and perceptions of current depression care and the potential role of cCBT-enhanced collaborative care. Quotes have been lightly edited for clarity and readability.

Context of current depression care at VA GLA

Clinicians noted that their caseloads consist of a sociodemographically diverse group of patients, with the most common mental health issues being depression, anxiety, PTSD, and substance use issues. Most participants (n = 11) estimated that depression was present in at least half of their patients, oftentimes presenting as a comorbidity. As a social worker noted, “It’s very rare that I see somebody that doesn’t have some depression along with PTSD and anxiety.” Given these prevalence estimates, most participants felt that depression, and by extension suicide risk, were significant clinical priorities.

Owing to persistent staffing shortages (and high patient demand) in the specialty-based mental health clinics, several participants noted that their respective PC-MHI clinics have been left to provide more intensive services than originally intended. For example, one nurse care manager noted:

“Usually with PC-MHI we’re only supposed to be seeing mild to moderate depression, but a lot of the patients here, they have a long history of depression or [major depressive disorder], they have PHQ-9 scores way above the threshold or have had some history of suicide attempts or suicidality […] Those are patients we would send to the Mental Health Clinic. But because of the limited access to the specialty Mental Health Clinic right now, we’re taking in a lot of those kind of patients.”

Participants described a variety of traditional treatment options offered for depression through PC-MHI clinics, including medications, individual psychotherapy, and group therapy. They also identified numerous barriers to engaging Veterans in these treatments. Both patient and clinic barriers were noted and related to individual, structural, and cultural issues. Primary barriers to treatment engagement included the inherent characteristics of depression itself (e.g., lack of motivation), Veterans’ competing work schedules and obligations, the stigmatization of mental health issues and treatment, and structural challenges in accessing care (e.g., inflexible clinic hours, staffing shortages).

Participants also identified meaningful facilitators to treatment engagement, including the flexibility and cohesion characteristic of the PC-MHI model, open access and same-day scheduling, and having a good therapeutic relationship with a clinician knowledgeable in depression and its treatment. PC-MHI being embedded within primary care clinics was seen to ameliorate the stigma associated with mental health care. For example:

“The nature of the program being open access and same day access with Veterans, I think it does facilitate them feeling comfortable coming in because then they don’t have the stigma of seeing a mental health professional because they see us as embedded in their Primary Care clinic, so I think coming in they don’t feel the stigma of having to go up to Mental Health.” –Social worker

Perspectives on online and mobile mental health treatments, including cCBT

Nearly two-thirds of participants reported that they have referred their patients to use online or mobile mental health treatments for depression or other behavioral health issues (e.g., PTSD Coach) but noted many obstacles, predominantly on the clinic level. Although they noted that some patients are not facile with technology, they acknowledged that VA GLA and current PC-MHI models do not facilitate the use of these alternative care modalities for patients who are interested. Reasons ranged from lack of infrastructure (“We don’t have WiFi, so they can’t download it here in session” –Psychologist) to poor integration into clinical work:

“The apps that are out there now, there are apps where a patient could use them to track their symptoms but that data doesn’t go anywhere, it just sits on their phone, and so it really depends on them coming back and saying, ‘Okay here’s my app and this is what I’ve done’ […] Long story short, I think the apps could be useful, [but] they’re just starting to come out now and the stuff that already exists really doesn’t add a lot of functionality or efficiency, or anything that you really can’t do with just pen and paper.” –Psychiatrist

No participants were aware of the existence of the VA’s free-to-Veterans cCBT program [16], but they generally reacted positively to the potential incorporation of cCBT in their clinical practice (e.g., “So, when is this happening?” –Psychiatrist; “I love this idea and I hope Veterans would love it too.” –Nurse care manager). Most expressed a willingness to refer to cCBT when appropriate for certain patients.

“It would be another tool, I mean in PCMHI, all we’re looking for is more tools so it’s what’s helpful for the Veteran. I’m not wedded to any particular theoretical perspective, [just] what can be helpful. The fact of the matter is people have different life experiences and they’re going to respond to different things and not everybody is going to respond to the same thing. I think with computer-based intervention, for some people that might be really great, so absolutely, I would use it a lot, especially when we have low [staffing].”–Psychologist

Perspectives on cCBT-enhanced collaborative care model

Target population

Overall, participants felt that a subset of patients would be receptive to and appropriate for cCBT-enhanced collaborative care, particularly younger Veterans who are comfortable with and interested in technology (such as OEF/OIF Veterans). Other potentially appropriate groups would be patients with mild-to-moderate depression or illness of shorter duration, those who want to “dip their toes” in mental health treatment, those who prefer remote treatment to face-to-face (e.g., to avoid mental health-related stigma), and patients whose schedules or family obligations make it difficult to attend appointments during clinic hours.

“I think especially with the younger Veterans, they could prefer this type of interaction as opposed to meeting with someone face-to-face, because that’s a little intimidating.” –Psychologist

“I’m sure it would work for some patients. I think it would be hard for me to pick this often over an in-person visit with a psychologist, but again for that patient who wants medication, really just doesn’t want to talk to a psychologist, maybe it’s too much in terms of their work schedule, maybe they’re just not open to talk therapy, I think that’s the patient I’d be most likely to offer this to.” –Psychiatrist

This care delivery model was not perceived as appropriate for patients with severe or recurrent depression, or for patients who are socially isolated, desire human interaction, are not technologically savvy or interested, lack access to a computer or telephone, lack stable housing or financial resources, or are suspicious and distrustful of “the system” (e.g., concerned about privacy, use of data, and potential surveillance).

“If they don’t have stable housing … or maybe there are some other concerns that just make it hard for them to stay on top of things.”–Psychologist

Clinical support for cCBT

Participants were consistent in their view that patients should be followed and supported by either a PC-MHI clinician or care manager while embarking on a cCBT program. As a psychologist noted, “A lot of people have trouble engaging with this stuff, even though they may be interested in it. That’s why I feel like having them checking in with the provider is helpful.” –Psychologist. Some emphasized that depression is often accompanied by social isolation, so these patients should not be using online or mobile technologies as treatment solely on their own. They differed on who might be the most appropriate person to take on this role, but most frequently suggested either a social worker, nurse care manager, or psychologist.

Q: “Who would be the most appropriate person to support the cCBT users?” A: “If we had any [staff member] in the world, it would be a provider. So a social worker or psychologist, psychiatrist. In the real world, whoever we got.” –Psychologist

Regardless of educational background, participants emphasized that this person should be knowledgeable and well trained in CBT to provide sufficient patient support. Although many clinicians desired the option to follow their own patients through the program, psychiatrists universally preferred deferring to an assigned care manager.

Q: “So who would be that ideal care manager then?” A: “Somebody that I really trusted and I trusted their clinical judgment as well as their conscientiousness, that I would be reassured that they were really doing what the model was. If not, then I would prefer to do it myself.” –Psychologist

Most participants regarded cCBT as another tool in their therapeutic toolbox and believed it could function both as a standalone treatment and as an adjunct to more traditional treatments such as psychotherapy or medication. Some regarded it as a first-line intervention to pique patients’ interest in mental health treatment and identify whether more intensive treatment is warranted. Another participant suggested it as an option to provide extra support for interested patients who had completed their course of brief in-person treatment and could then apply CBT concepts more concretely to their online experience. Participants generally expressed flexibility in how they would use cCBT depending on patient preferences. For example:

“I would love for it to be able to be fluid, that it is set up to be standalone and we could just refer people and say, ‘Hey look, this is another option for you if this feels better for you give it a try.’ And then if they are willing to engage in treatment or kind of have the fluidity to come back into treatment after engaging in some of that and wanting to apply the skills, we could then interface with it and have access to the data and use that clinically with the patient. I think that would be beneficial.” –Psychologist

Feasibility and readiness to adopt

On the whole, participants felt that cCBT could be easily integrated into their existing PC-MHI clinical structures and that implementation was feasible. As one psychiatrist noted, clear instructions would be essential to the implementation process: “I think if it was clearly laid out how the program would work, what the referral process was, [and] the care management nurses are trained in it, then we could roll it out.” –Psychiatrist Several remarked that the flexibility inherent to the PC-MHI model would facilitate adoption:

“I think as a whole PC-MHI is a very limber model and if you work in PC-MHI, to be successful, you yourself have to be limber. So, if there’s going to be an area where you’re going to bring something like this in, I think PC-MHI is a perfect place.”–Psychologist

When asked about clinic readiness to adopt, participants’ responses were more varied. Most felt that clinicians are theoretically receptive to the use of an additional tool and willing to adopt it. However, the process could be hindered by clinicians’ inconsistent knowledge of or comfort level using online and mobile technologies, lack of time for learning, or hesitation adding to existing clinical burdens. For example, contracted clinicians (i.e., non-VA employees) were perceived to be less willing to adopt a new technology: “I will be brutally honest—[we’re] not ready … Because it’s a contract clinic, it would be like, ‘Oh yeah, one more thing…whatever’. They’re not going to buy into it” –Social worker. A psychiatrist cautioned that the larger VA context of mental health care might eventually impede adoption:

“I think it’s a great idea with a lot of potential, but [it’s important] to acknowledge at the same time that in the background of an intervention…is a mental health system that’s not quite optimal yet, so when the background is a maybe-not-quite optimal mental health system, to bring in something new, it could have its challenges.” –Psychiatrist

Suggestions for improvement

Participants offered a number of suggestions to improve any VA-sponsored cCBT program. Most underscored the need for a mental health-trained clinician or care manager to follow patients who use cCBT, which is not standard practice (if even used at all) in the VA. Additional support, such as necessary training in CBT, should be provided to this person as needed. Other suggestions were directed at the online cCBT delivery platform, most of which have already been incorporated into the extant cCBT program. These include ability to electronically monitor mental health assessments over time, offering incentives or contests, personalizing the homepage (e.g., including Veteran testimonials, success stories, and program benefits), incorporating videos and interactive tools within the program, and sending automatic reminders. They emphasized that patients need clear instructions, a user-friendly format, and reliable access to support and troubleshooting.

Participants provided additional recommendations to increase clinic readiness for adoption and implementation. Initial framing of the cCBT program was seen as important; primary care or nursing leadership buy-in is necessary to emphasize its importance and to reduce perceptions of another arbitrary task: “What I’m finding is I think it needs to come from [above], whoever is in charge of Primary Care saying, ‘Thou shalt do this. This will be done.’ Rather than leaving it up to us, people on the ground” –Social worker. For patients, cCBT should be framed as an additional option available to them, to avoid any perception that the VA is defaulting to online and mobile tools for budgetary reasons or minimizing the importance of Veterans’ mental health care. As noted by a psychiatrist:

“[There are more severe patients who may] perceive this as validating their opinion that the VA just doesn’t want to allocate resources to them, or [doesn’t] care about them, or for whatever reason won’t have them see a provider. I think there are a lot of attitudes like the VA is not providing the resources that they need. A lot of Veterans feel like they deserve individual treatment, individual psychotherapy weekly at the time of their choosing forever and ever, right? This is kind of the want that they have. So, I think we have to be very careful in how we frame this treatment, that it doesn’t seem like we don’t care or we’re not feeling that they deserve an individual provider. How do we do that? […] I think if it was an option offered, that I like that option, then that would be great. Rather than, ‘This is what you should do,’ or ‘This is all we have.’ Then that would be frustrating for them.” –Psychiatrist

Several participants also referenced the need for buy-in and involvement from frontline PCPs to facilitate patients’ engagement in the program:

“I think the key is actually the primary care physician. I think ownership of the care by the [primary care] team and in particular by the primary care provider who has referred them, I think that’s really crucial. That’s one of the areas we’ve struggled the most in PC-MHI, which is buy-in from our primary care docs, and we get it. Primary care at the VA, there is never enough time and too many things to cover.” –Psychiatrist

Clinicians suggested offering training to clinicians initially, as well as the provision of clear instructions or a manual on how clinicians can use the treatment, how patients can meaningfully interact with it, and which patients are most appropriate for referral. As one participant suggested, “Maybe when it gets rolled out, have providers that are geared for it to be trained in it, and how to present it and mainly how to use it in the [therapeutic] process.”–Social worker. Finally, participants repeatedly underscored the importance of evaluating the effectiveness of the program to provide guidance for future use.

DISCUSSION

We found a high level of enthusiasm among study participants for PC-MHI collaborative care enhanced by clinician supported delivery of a cCBT program for depression. Given the opportunity to react to the proposed care model, GLA collaborative care clinician impressions were generally affirmative. Several justifications for positive remarks were offered, such as increasing PC-MHI care access for Veterans desiring more convenient care. In particular, this enhanced care model may potentially engage half of OEF/OIF Veterans, who report untreated mental health needs and would be open to receiving treatment via the Internet [32]. Participants overwhelmingly reported being short-staffed and felt that cCBT could increase the number of Veterans they would be able to treat in clinic. Although primary care patients may prefer [7], or even need,[33] psychotherapy (in conjunction with medication), health systems remain challenged by poor mental health clinician access [34], driven in large part by chronic workforce shortages in the VA [35]. Collaborative care clinicians felt cCBT was appropriate for inclusion in the PC-MHI model and that the proposed program could be feasibly adopted and implemented in VA primary care clinics to close the psychotherapy access gap. Finally, they noted that PCP buy-in for cCBT treatment is important and, thus, a necessary area of future investigation.

Participants were clear in their views that patients need clinical follow-up and support while using cCBT for depression, which is not routinely available in the VA or in other health systems in the U.S., to the authors’ knowledge. Depression was brought up as a condition in which affected patients are not motivated in self-care [36], which would, in turn, make it difficult for patients to engage in cCBT. Most participants were strongly in favor of a care manager providing clinical support for cCBT users. Clinician-supported delivery of cCBT within the PC-MHI model seemed to alleviate concerns about poor patient engagement with cCBT technology and resulting ineffectiveness from low patient uptake of treatment. Participants acknowledged that current PC-MHI models do not facilitate alternative care modalities. Many of them shared personal experiences about their attempts to engage their patients in VA mobile apps and online self-help, which have been met with barriers similar to those documented in poor cCBT uptake [19]. As such, the VA and other health systems will need additional research to guide them on which primary care patients to target for cCBT and how to engage users with appropriate levels of clinical support.

In this US-based study, few participants knew anything about, let alone ever used, cCBT for depression in their clinical practice. Interviewees also confirmed that no GLA clinicians are using cCBT for depression and stated that they were not aware of any cCBT program being integrated into real-world (non-research) clinical care in the VA or elsewhere. cCBT knowledge among our study clinicians was low to non-existent, similar to observations among mental health clinicians in international locales where cCBT is more established [20,21]. For this reason, implementation will be challenging within our US-based primary care clinics, even while participants express enthusiasm for the proposed care model. Clinician buy-in on cCBT may also be affected by the few studies conducted specifically with primary care patients, who are older and more medically complex than average cCBT users [10,25,28]. Clinician misperceptions about cCBT efficacy and inferiority to face-to-face psychology can limit cCBT uptake [23,24] and will also need to be addressed to effectively engage clinicians in supporting cCBT use. Additional research is needed to tailor cCBT to be an effective, easy-to-use tool in clinical practice to treat primary care patients with depression. These obstacles must be anticipated and addressed in partnership with service leadership and frontline clinicians before cCBT roll-out in VA primary care clinics.

A case study approach provided us in-depth information regarding implementation of our proposed cCBT-enhanced collaborative care model from the perspective of VA GLA PC-MHI care managers and clinicians but is affected by limitations. First, our clinician interview sample was small and from a single site. Yet, it consisted of the majority of local VA PC-MHI clinicians and, as such, will inform program implementation and aid in future stakeholder engagement. Second, we conducted interviews only with PC-MHI mental health clinicians and, therefore, lack PCP and patient perspectives. These perspectives will be needed in future research to paint a comprehensive picture to prepare for program implementation. Finally, the case study is bounded by the context of present-day VA GLA and its PC-MHI care model; however, speculations about cCBT offered by primary care-based mental health clinicians here may inform other VAs that are similarly mandated to implement PC-MHI and for other health systems that use collaborative care to treat depression among their primary care patients.

Although cCBT awareness and knowledge were not widespread, PC-MHI clinicians were highly accepting of adapting collaborative care models to increase uptake of cCBT for depression in this study. Participants favored cCBT delivery supported by a PC-MHI care manager or clinician to overcome concerns that patients with depression would not engage in cCBT treatment or that the program would not meet patients’ interpersonal needs. Clinicians felt that younger Veterans would prefer the convenience afforded by cCBT and saw it as an additional tool to engage patients in mental health treatment. Participants acknowledged that current VA clinics and PC-MHI models do not facilitate the use of online or mobile treatment modalities but that cCBT, if effectively implemented, had potential to increase the number of patients they would be able to treat in clinic. Given continued calls to modernize health systems like the VA, primary care-based mental health clinicians seemed agreeable to adapting collaborative care to improve uptake of cCBT for depression to increase provider capacity and ultimately patient access to effective psychotherapy.

Implications.

Practice:

Primary care-based mental health clinicians are amenable to adapting collaborative care models to improve uptake of computerized cognitive behavioral therapy (cCBT) for depression to increase patient access to effective psychotherapies.

Policy:

To improve access and outcomes among primary care patients with depression, the Veterans Health Administration may consider enabling effective use of cCBT by adapting its existing Primary Care–Mental Health Integration programs.

Research:

Implementation research is needed to facilitate the use of online and mobile depression treatments, including cCBT, in primary care clinics to close the access gap in psychotherapy.