Katie Doyle, M.S.N., R.N., vice president of operations at Chicago-based Northwestern Medicine, recently spoke at a meeting put on by the National Committee for Quality Assurance and West Health about the integration of behavioral health into primary care. Following the meeting, Doyle spoke with Healthcare Innovation about Northwestern Medicine's multi-year journey implementing the Collaborative Care Model at scale.

The Collaborative Care Model is an evidence-based approach to identifying and treating patients with behavioral health conditions such as anxiety and depression in primary care settings.

Healthcare Innovation: Could you start by describing your role at Northwestern Medicine? 

Doyle: Northwestern is an academic medical center in Chicago, and I support the primary care portfolio. We have done a lot of good work investing in a service line approach and an infrastructure that provides good leadership for strategies to go down to the practice level, and also for us to bubble up what's impacting our physicians. My responsibility is to facilitate that strategy. 

HCI: How did the work on the Collaborative Care Model get started? 

Doyle: We started by measuring depression screening about 10 years ago. Everyone knows it's important, but our physicians pushed back about the measure. They were concerned about screening patients when they didn't know how to support patients and offer needed follow-up. It took us a few years to make screening part of our workflow. We share performance from our quality team monthly, and we create targets yearly. We started getting to this place where depression screening was truly enculturated into our practice for our patients during their annual visit. 

Our Collaborative Behavioral Health Program (CBHP) had been around at Northwestern for about eight years in small pockets across the system, and we really felt that for population health and primary care transformation work, it is imperative that we have somewhere for patients to go when our physicians identify that they have a need they can't handle.

That's where we started to partner with Department of Psychiatry. West Health came in to help us, and we've been able to scale access to CBHP to all 70 clinics and 500 physicians in the organization. And it really started with that depression screening measure from CMS. If we didn't have the awareness of how well or poorly we were doing and understand the barriers that our physicians were feeling, we might not have made this change as quickly as we did. 

HCI: What are some things you had to work through with the primary care clinicians? 

Doyle: One hard part has been helping our physicians understand that just because you can prescribe something doesn't mean that the patient's actually better. Our physicians would say “I prescribed the patient Lexapro or an SSRI. I told them to call me back if they don't feel better.” Well, patients may not call back. But also, how do they know what better feels like? Just because you prescribe something doesn't mean that the intervention actually equals improvement in outcomes. 

With the CBHP program, we've been able to highlight the frequency of visits and care, which I think most of our physicians are wowed about. They say, “I can't believe my patient gets six touch points with an LCSW, where they're talking about their needs, and that LCSW is working with a psychiatrist adjusting medications.” So the frequency of care has been a huge win for our PCPs to engage. The fact that they can see the data has been very impactful.

Our goal is to get at least 30% of the patients eligible in our service line to get into CPHP. We're about halfway there, but I think we can get there.

HCI: Do you have enough behavioral health resources inside of Northwestern Medicine to handle that many patients or do you have to find external partners, too?

Doyle: We actually have both. We have used the internal Department of Psychiatry at Northwestern to be the owner and builder of the Collaborative Behavioral Health Program. They are supporting the psychiatrists and the LCSWs and the model to meet the quality expectations and the demands of our primary care docs. We also have about 80,000 pediatric lives in our service line, so we're using a vendor called Concert Health to do the same work, but for our pediatric patients. That work is virtual. 

On the adult side, it's just nice to have everything in the EMR, and it's easier for everyone to be connecting. What's hard for our primary care physicians — and I think in any screening and intervention that requires follow-up in primary care — is that there isn't an incentive or a payment model that actually appreciates all those touches. In reality, there's a lot of work that happens. The physician does the screening with the patient. They then discuss what the outcome is. Then there's the collaboration for referral. Then there are all these touch points. If the psychiatrist makes a medication change, that messaging goes back to the physician. The physician has to then order it, and if the patient has a side effect, that message doesn't go to the psychiatrist; it goes to the internal medicine physician, who has to then educate the patient about the side effects. And every medication change is an intervention that the physician has to do for primary care, which is good, but it's just work that's not really appreciated, because it’s administrative back and forth that requires a lot of time. We use the collaborative care codes, and they're helpful, but it doesn't necessarily compensate the physician enough, at least in our model. We are trying to figure out how to help our physician stay incentivized and find ways for them to see that we appreciate the work they're doing.

I think the thing that's probably the most different across health systems is how people compensate physicians.

HCI: After a program like this has been deployed for a few years, are there attempts to figure out if it's having a positive effect on the health system’s financials? If you have these patients doing better with their behavioral health needs, is it actually beneficial in terms of health system utilization as well as to the patients?

Doyle: We think of this as part of our population health strategy, and we also see this as a retention strategy for primary care physicians as we build clinical models. 

The only bummer about value-based care is that you don't really see the win as soon as you want to in order to create momentum. But we believe in the next three to five years we're going to see it impact total cost of care for our value-based care patients. We believe we'll see the improvement in chronic disease metrics like A1C and hypertension, because we know there are co-morbidities with patients who have complicated chronic disease and depression or anxiety. There is literature that supports this, and if we do this right, we're going to see that impact. We have about another 18 months with West Health to fully operationalize the scaling of the program, and then we'll take some time to review the impact on total cost of care.

HCI: What does West Health bring to this? What kinds of things do they help with?

Doyle: They’ve helped us in a couple ways. They have brought in the Meadows Mental Health Policy Institute, which is a clinical expert in this space nationally. Meadows has been able to help us optimize the program. They have helped our leaders in psychiatry with the work plans. It's very complicated to do program design. It takes listening to people, understanding what their concerns are, and data sharing. West Health has been able to help us think through that work plan so it can truly be scaled, which has been very helpful. And the investment they gave us helped us accelerate it. 

HCI: I read that one of the goals of this NCQA meeting you attended was to agree on a core set of measures that can help guide how integrated care is delivered and paid for consistently at scale. Was there some agreement or work toward that at the meeting?

Doyle: I don't know if we actually found a core set of measures. I think what was definitely heard was that in order for this work to be successful we've got to ensure that there is a reimbursement model that supports the work and that is not transactional, but longitudinal. In the fee-for-service model you can definitely have a blood pressure that gets under control with one medication, and that's a win. But the patients who are truly complicated or have many barriers to their health, like social drivers of health, they need much more time. They need to trust the healthcare system. They need to build relationships with their primary care physician. They need help with the cost that comes to them for whatever the intervention is. With the measure set, we’re not yet there in terms of helping primary care teams appreciate the long-term commitment needed for patients to get better and have really good, tangible outcomes.