Interviews identified implementation barriers and facilitators for CBT-I use in primary care. These barriers and facilitators fell under three broad CFIR constructs: intervention characteristics, inner setting, and outer setting. In addition, interviews revealed the importance of the implementation process of engaging, including active CBT-I champions and strong support from opinion leaders. These barriers and facilitators are described in detail below and outlined in Fig. 1.
Intervention characteristics
Evidence strength and quality
All primary care physicians interviewed were familiar with CBT-I as an effective treatment for insomnia offered within VA, but most were unfamiliar with specific guidelines or primary evidence. Knowledge of CBT-I was mainly derived from local champions and on-the-job training rather than systematic education. As one physician said:
I have been educated in research talks from an expert in the area. I’ve forgotten the specifics of studies that show effectiveness, but I believe it is effective based on what I have heard.
Some physicians utilized CBT-I as a first-line treatment and had a “low-threshold” for making referrals, whereas others preferred to start with sleep hygiene advice and medication, making a CBT-I referral later if insomnia persisted. One PCMHI psychologist suggested integrating CBT-I education into resident lectures about general primary care problems, like pain, anxiety, and depression to increase the use of CBT-I as a first-line treatment.
Primary care physicians noted that positive feedback from patients is influential in increasing CBT-I referrals and “getting doctors on board.” Reviews from patients were mixed, with some enthusiastically reporting improved sleep after decades of insomnia and others reporting no treatment effect. One physician used positive testimonials from past patients to sell the treatment to new patients. Physicians acknowledged that negative reviews may reflect a lack of engagement rather than a failure to benefit from CBT-I.
Adaptability
All providers appreciated the potential scalability and convenience of CBT-I for primary care settings, including brief versions of the treatment and telehealth delivery. Self-management CBT-I, in which patients independently complete an online web-course or paper-based treatment manual, received mixed reviews. PCMHI psychologists were the most positive about self-management and felt that this fit within their model of care and increased access, especially if it was combined with provider check-ins. Primary care physicians and CBT-I coordinators were concerned about patient follow-through and comorbid sleep disorders, such as sleep apnea, that may not be diagnosed and treated with a self-management approach. One physician preferred in-person CBT-I to self-management because it provides a “foot in the door” for patients that need additional mental health treatment:
I like sending them out to PCMHI because I secretly hope they will address some of their other underlying mental health issues so sometimes I use CBT-I as a foot in the door. I’ve had a couple patients where sleep is a major issue, but so are these other things that they are not ready to address yet, so I hope that once they start talking, they’ll keep talking.
In general, primary care and CBT-I providers were more likely to consider self-management for patients without comorbidities and as a last resort for those who were unable or unwilling to work with a CBT-I provider. Primary care physicians also felt they would not have enough information or time to provide patients with self-management resources and training.
Inner setting
Networks and communication
Strong connections between primary care clinics, PCMHI teams, and sleep medicine clinics increased utilization of CBT-I. Within VA, CBT-I training is provided through the VA Office of Mental Health and Suicide Prevention (OMHSP) and many CBT-I providers are located within mental health clinics. Highly successful CBT-I programs were often closely integrated with both mental health and sleep medicine clinics and functioned as a sleep team or community, with shared consult and referral systems. As one CBT-I coordinator said:
A patient is typically sent the sleep disorders clinic. […] The consult is reviewed by one of our sleep fellow and there is an internal consult that goes to the behavioral sleep medicine clinic and they get CBT-I.
These sleep teams provided ongoing services and education within primary care clinics. In some cases, individual relationships increased CBT-I use. For example, primary care residents who were exposed to CBT-I champions during their training often took full time staff positions and went on to utilize CBT-I.
Compatibility and relative priority
Primary care physicians did not proactively assess for insomnia and sleep loss was not considered as urgent as other health concerns. As one physician said:
If someone convinced me that there was a strong causal pathway between inadequate sleep and serious morbidity and mortality, then I would be thinking ‘wow we need to move on this asymptomatic killer of people.’ It is hard for me to think of insomnia as one of those asymptomatic killers we are supposed to screen for.
Widespread screening for insomnia was not seen as compatible with primary care appointments given limited time, competing demands of other medical conditions, and patient costs (e.g., patient fatigue, damage to rapport if patients get annoyed with screening). There was also an assumption that mental health providers were assessing and treating insomnia. Physicians suggested that lack of screening may also be related to uncertainty about how to treat insomnia, a preference for not prescribing sleeping medication and a general reluctance to bring up sleep because of a preference for not prescribing medication. The general consensus among physicians was that sleep took a “backseat priority” in complex patients and was not prioritized by patients or providers. As one physician said:
Sleep is definitely a backseat priority. It’s important and it affects everything but patients are coming in with blood pressure at 170 over 100 and their A1C is 12 and they are sobbing because of their…..whatever. And patients don’t always recognize how much sleep affects them and you have to get patients to buy in especially if it involves behavioral change. And they just want a pill, they want me to prescribe Zolpidem because they perceive that does help and they don’t want to do the hard work.
Both PCMHI psychologists and CBT-I coordinators expressed the need for more provider education about the importance of sleep, but also acknowledged that providers are stuck between wanting to decrease sleeping medication and their patients not wanting to engage in CBT-I.
If patients brought up insomnia during appointments, primary care physicians commonly referred them for further evaluation and treatment. Physicians were comfortable providing a basic overview of CBT-I to their patients but did not have the knowledge or time to describe the treatment in-depth. Several physicians wanted written guidance on how to describe CBT-I to patients. CBT-I coordinators pointed out that comprehensive sleep evaluation and treatment discussion is part of the CBT-I intake; however, it may be helpful to have primary care providers set the stage:
[Patients] have no clue why a behavioral intervention is related to insomnia at all. The backdrop is you treat insomnia with sleep medications so what in the world could I do sitting in a room talking with people, how would that help my sleep. They can’t even wrap their head around it. So the primary care provider has to be the salesperson for that intervention. A good chunk of people don’t show up or engage because they don’t see how it applies to them.
In contrast to primary care physicians, PCMHI psychologists viewed insomnia screening and treatment as highly compatible with their model of care. They routinely screened for insomnia during functional assessments and prioritized sleep treatment in accordance with patients’ treatment goals. PCMHI psychologists found CBT-I easy to deliver in a PCMHI setting, especially in briefer formats.
Available resources and access to information and knowledge
Most primary care physicians were satisfied with CBT-I resources in their facility. They reported having ready access to CBT-I providers through electronic medical record consults and warm-hand offs to PCMHI teams (i.e., patient has same day appointment with PCMHI team). A variety of consults were used across and within facilities, including general sleep medicine, behavioral sleep medicine, mental health, and PCMHI, and this led in some cases to confusion about how to refer to CBT-I. Several primary care physicians expressed the desire for a more centralized resource to learn about CBT-I and make referrals, something easy and “within their pattern of practice” such as established order sets within the electronic medical record. This was particularly useful after-hours or on weekends when warm-hand offs were more difficult. For example, one primary care physician said:
There’s no central repository of information, no place you can go for a reference.[…] Sometimes you just want to find something out without going to track someone down.
In contrast to primary care physicians, PCMHI psychologists and CBT-I coordinators expressed a need for more CBT-I providers within primary care, especially in off-site clinics. Demand often overwhelmed supply, with up to 6 months wait for primary care patients to see a CBT-I provider. One PCMHI provider suggested the need for local training due to the limited number of slots within the national training program. Telehealth was suggested as a potential way of increasing access within PCMHI settings.
Outer setting
Patient needs and resources
There was a consensus that CBT-I provides a valuable nonpharmacological treatment option for patients struggling with insomnia. There were contrasting views among primary care physicians about patient willingness to consider CBT-I. One physician noted that patients were generally receptive since “not many patients just want medications.” Conversely, another estimated that up to half of his patients were not interested in CBT-I since they “just want trazodone and to take care of it all in primary care.” One CBT-I coordinator emphasized the need to educate both physicians and patients about the availability of nonpharmacological options like CBT-I:
My experience is providers are really uncomfortable with meds but they perceive that is what their patients want. […] They have this tension. We asked patients why they didn’t talk to their doctor and they said, ‘all my doctor has to give me are pills and why would I bring it up if I didn’t want one.’ And we talked to providers who said, ‘my patients come asking for pills and I don’t know what else to do.’
Nearly all physicians downplayed the mental health aspects of CBT-I when making referrals, avoiding the terms “psychologist” and “therapy.” They felt patients were more receptive when CBT-I was described as a “sleep intervention” delivered by “sleep experts.” Several physicians commented that the behavior change is more difficult than taking sleeping medication and that some patients do not understand how sleep affects health. They recommended patient education to promote wide-spread uptake of CBT-I.
External policies and incentives
One primary care physician pointed out that it was difficult to focus on increasing the use of behavioral sleep treatments as an alternative to sleeping medication because there was no measurable data to show improvement in practice, unlike other deprescribing initiatives:
Sleep doesn’t have measurable data where you can say we reduced our benzo users from 20% to 10%. You can’t say I fixed insomnia [treatment] by 10%. A lot of things we follow have measurable action items.
A CBT-I coordinator suggested that efforts to increase the uptake of CBT-I be linked to deprescribing initiatives, with CBT-I offered as an alternative to initiating or continuing sedative hypnotic prescriptions, including benzodiazepines. Several primary care physicians reported that they are already taking this approach. As one said:
I’m in a long-term project with my folks who have been on sleeping pills to change the way we are managing their sleep.
Similarly, CBT-I coordinators are advertising CBT-I as a safer and more effective alternative to medications to increase referrals:
It helps that the DoD/VA guidelines came out. […] I’ve been waving that like a big flag and using that to highlight that CBT-I is best practice. No one is saying that medication is better.
Importance of engaging
Providers identified two key facilitators that they believed contributed to successful implementation and widespread dissemination of CBT-I: local champions and leadership support. Local champions represented the most consistent and effective method of increasing insomnia awareness and CBT-I referrals within primary care. Primary care physicians who routinely utilized CBT-I were made aware of resources through formal talks and informal conversations with local champions. As a result of these contact, they were more likely to be listening for sleep problems among their patients and have a “lower bar” for CBT-I referrals.
CBT-I coordinators at well-established CBT-I programs suggested that leadership support was crucial in growing their services. They described how CBT-I was “warmly embraced” with “a lot of enthusiasm” by leadership in sleep medicine, primary care, and mental health. Leadership support took the form of educational initiatives, rolling out CBT-I clinics and protected time for CBT-I providers. Supportive leaders valued CBT-I as an essential service and these attitudes had a trickle-down effect to primary care clinics, increasing CBT-I referrals. One CBT-I coordinator described the positive reception for CBT-I programming at a primary care meeting:
Primary care providers here overall prefer not to write prescriptions for sleep. […] Someone said ‘we hear you are developing a non-medication treatment program for insomnia’ and he said ‘yes we are’ and he actually got a round of applause because they were so happy to have an option that wasn’t sleeping medication.
It was noted that receptiveness to CBT-I varies substantially among sites and that lack of leadership support has a chilling effect on referrals and provider willingness to engage.