Steve Levine, MD: Dr Nordberg, thinking about your health system as an integrated primary care system, you’re often having patients transition from primary management by primary care to specialty care with psychiatrists. How do you identify the patients in need of alternative therapies for TRD [treatment-resistant depression] outside of their oral medications?

Samuel Nordberg, PhD: There are a variety of different hallmark markers that tend to help them rise to the surface in an integrated primary care environment. One of them Dr Rosenzweig referenced in talking about bogging things down. One of the things that often happens is you may be making a number of referrals and treatment recommendations for patients with TRD, and they just keep coming right back into the primary care system over and over again because those treatments are failing them. They’re maybe not even engaging in those treatments because they don’t have the wherewithal to do so.

One of the things we see is churn. We thought we had a disposition and a treatment plan for that patient, and then they’re right back with us or right back into the emergency department [ED]. Other key indicators are use of crisis services or family outreaching for support and help because they’re at their wit’s end. With TRD, you often see patients who lack the motivation and wherewithal to advocate on their own behalf. You may also see it on behalf of the patient from family members or friends. One of the things that we see is a lot of churn, a lot of noise, and a lot of phone calls and conversations. We see a lot of frustration on the part of the primary care physician who’s trying to coordinate all of this work but meeting with real difficulty. That’s 1 indicator that we see that you can track in the health record and develop algorithms to surface patients where there are enough things like ED visits and phone calls. We do that at Reliant Medical Group.

At Reliant, we have the advantage of having consulting psychiatrists who support primary care physicians [PCPs] in prescribing new medications, changing dosage, etc. One of the things we also have is we get eyes on those patients, where this is maybe the third, fourth, or fifth medication change. At that point, we’ll pull them in with us and do a complete robust differential diagnosis and try to get to the root cause of what’s going on. That gives us an opportunity to take and triage those patients into another level of care.

The last thing that we also use is patient-reported outcomes. We’re using the PHQ-9 [Patient Health Questionnaire-9] for depression, but shortly we’ll be boosting that with a bunch of other more computer-adaptive measures that will hopefully allow us to get at a bit more of the nuance and complexity that’s going to allow us to picture that better. You can use data over time. You alluded earlier to the complexity around differentiating between 1 episode vs multiple strings of chained-together episodes. Those longitudinal data can also be very helpful in surfacing that information.

Those are some of the key ways that we start to identify treatment-resistant depression. You can also do a claims lookback, which can sometimes be helpful but suffers from the fact that our claims are lagged by 3 or 4 months. The better indicators are much more real-time and exist either in the health record or in the real face valid treatment experience that our PCPs are having with their patients.

Steve Levine, MD: You take a very comprehensive approach, and you also measure outcomes. Good for you. As you know, many don’t. It sounds like within Reliant, many of your patients get a level of care that isn’t necessarily the standard across the mental healthcare landscape.

Samuel Nordberg, PhD: We’ll talk about whether they actually get that care in a little bit, but we have a better opportunity to identify them.

Steve Levine, MD: Thanks for making that important distinction.

Transcripts edited for clarity.