Megan Maroney, PharmD, BCPP: Dr Cannon, what are some of the limitations associated with implementing a TRD [treatment-resistant depression] definition in claims databases?

Eric Cannon, PharmD, FAMCP: I don’t know that we implement the definition in a claims database. We’re using it in terms of data analytics to determine the economics of things. What are the outcomes associated with it? So claims data [are] just that, and your results and [implementation] are only as good as the data you have. We’ll know that someone’s been on antidepressants for a period. Or we might see, as Dr Thase mentioned, they’re on an SSRI [selective serotonin reuptake inhibitor] and you added aripiprazole, but we don’t know for certain [whether] that person’s responding to treatment. As we look at data sets, being able to join that with a medical record or at least a data warehouse associated with the medical record is very helpful.

Megan Maroney, PharmD, BCPP: Dr Skolnik, how common is TRD? How prevalent is it? And how do the rates differ across different genders and ethnic groups?

Neil Skolnik, MD: I’m going to distinguish between [patients] who have a partial response, which we see a lot in primary care, and [patients] who have TRD, where there’s that drop off in the STAR*D trial [NCT00021528]. From what we heard from Dr Thase, 50% of [patients] responded to first-line treatment. Of those [who] were left, 40% of [patients] responded to second-line treatment. You get a number that’s consistent with other numbers out there, that about a quarter of [individuals] have TRD, meaning [they] don’t respond to 2 or more therapies with different medicines. There’s a whole category of partial response that is of equal or even greater prevalence than complete TRD. And this was talked about in the recent American College of Physicians guidelines on the [management] of depression. In those guidelines, they talked about up to 70% of [individuals] not having a full response. That seems a bit high, but recognize that at least an equal number of [patients] have a partial response as the 25% of [patients] who have treatment-resistant depression. I don’t know [whether] there [are] other data that panel members want to add.

Michael Edward Thase, MD: In a cure setting, maybe 10% of [individuals] don’t fill that first prescription, and another 33% don’t refill the second prescription. And the folks who don’t fill their prescriptions typically remain [with depression]. You can get that 70% unimproved amount if [individuals] aren’t in treatment. So if you think about the folks whom you intend to treat, we don’t get as much good as we wish to do because we can only treat those who will take the treatment.

Eric Cannon, PharmD, FAMCP: Neil or Michael, how many of these patients [who] are not responding in typical real-world practice are [not responding because of] an inadequate dose? Does that happen frequently?

Neil Skolnik, MD: There are different layers where there can be causes for an inadequate response. I don’t know the data on it, but I sense that pushing the dose is one of those things we’re used to doing in primary care. So I doubt it’s the fact that an inadequate dose is used as much as [what] Michael talked about: poor adherence. I think that our anecdotal experience is influenced by the amount of support we give. I know there [are] good data on the benefit of CBT [cognitive behavioral therapy] of interpersonal therapy. There [are] less solid data on what we can do in the office with our own type of interpersonal therapy. My hope, which can interfere with the assessment of the evidence, is that what we do mean[s] a lot. Things like enlisting social support for the patient, where there [are] good data on that, [do] not go beyond what we can do in primary care. I had a patient, a young adult, who was [experiencing] depression, with tough stuff going on at home. Her mom had metastatic cancer. The young woman came in. I prescribed an SSRI; it was evening hours. I brought her family in…she called, and they all came in at the end of evening hours. We sat together and talked about things. I saw her 2 weeks later and started the SSRI, and she was doing so much better. I [thought], “I know my stuff.” I started the right medicine. I’m doing a great job. Of course, she hadn’t started the medicine; it was the increased social support from her family, where [before] the focus had been on the mom who had a terrible illness. The family focus shifted to this young adult, and it made all the difference. I think we can do a lot in the office by enlisting social support: simple behavioral things, behavioral activation, and setting up a program for increased physical activity. All those things enter our therapeutic approach in a practical way. And things [such as] adherence are things that interfere in a practical way. Michael, your thoughts?

Michael Edward Thase, MD: I teach our students and house officers that about 75% of the good we do comes from the care we provide and the specific antidepressant choices we make account for the [remaining] 25%. So if you want to get better outcomes, take better care of your patients. To Eric’s comment, I think one of the beauties of the second generation of antidepressants was that their starting therapeutic dose was often the right dose and hit the target with 80% accuracy so [that] patients would take the medicine. There wasn’t much of a dose-response relationship for the SSRIs with maybe the exception of escitalopram.

About 75% of the good we do, which is what I teach our house staff and students, comes from the care we provide for our patients. And the icing on the cake is the specific choices of the medicines we make and how we manage those. One of the beauties of the newer generation of antidepressants was that they were generally in an effective dose range, even from the starting dose. This was especially true of the SSRIs; only escitalopram shows a dose-response relationship. It’s true of the newer generation. [In] serotonin-modulating medicines, only vortioxetine shows a dose-response relationship, and venlafaxine and mirtazapine showed a dose-response [relationship]. So [in] 4 of 12 newer antidepressants, you get some extra traction with dose adjustment. But generally, you start these medications, and if the patient will fill the prescription and adhere to treatment after 6 to 8 weeks, they’ve gotten what they’re going to get with that medication. And if they’re not tangibly better, it’s time to move on.

Eric Cannon, PharmD, FAMCP: Megan, you asked what was difficult about implementing TRD in claims data, and I think you just heard it. This notion of that social aspect of the actual care that’s delivered by a physician is untracked. It may be documented somewhere, but it’s not available for us to report on.

Megan Maroney, PharmD, BCPP: It’s one of the reasons we see such high placebo response rates in clinical trials. It’s hard to quantify that.

Michael Edward Thase, MD: The placebo response is a good measure of the patient’s inherent ability to get better with proper support and the therapeutic impact of the support that’s prescribed. Maybe the color of the placebo or the ingestion of the placebo has a little to do with it, but being cared for is what matters the most.

Transcript edited for clarity.