Deprescribing Super Special III: Constance Fung, Emily McDonald, Amy Linsky, and Michelle Odden

Eric 00:00

Welcome to the GeriPal Podcast. This is Eric Widera.

Alex 00:03

This is Alex Smith.

Eric 00:04

And Alex, we’ve got another deprescribing super special with us today.

Alex 00:07

We have a full house to talk about deprescribing. We’re still working on this issue. We have not solved it, apparently. Our first guest is Emily McDonald, who’s a physician in general medicine and epidemiologist and Associate professor at McGill. Emily, welcome to the GeriPal Podcast.

Emily 00:23

Thanks for having me.

Alex 00:24

Great. And our next guest is Amy Linsky, who’s a physician and researcher at the VA Boston Healthcare System and associate professor at Boston University University. Amy, welcome to GeriPal.

Amy 00:34

Excited to be here. Thanks.

Alex 00:36

And next we have Connie Fung, who I know well through the Beeson community and is a physician, researcher, professor of medicine at UCLA. And we were talking about the fire situation there. I’m glad that’s starting to resolve. Connie, welcome to GeriPal.

Connie 00:50

Thank you. Great to be here.

Alex 00:52

And our last guest is Michelle Odden, who we know well, used to be here, is now epidemiologist and research scientist at the Palo Alto VA and is associate professor in Stanford. Michelle, welcome to GeriPal.

Michelle 01:04

Thank you for having me.

Eric 01:06

All right, we got a lot to cover. Before we do, we always start off with song request. Amy, do you have the song request.

Amy 01:11

For Alex, how about Just Like a Pill by Pink?

Eric 01:14

Why did you pick this song?

Amy 01:16

I mean, in addition to the lyrics being pretty spot on, I kind of wanted to see Alex singing Pink.

Alex 01:23

Yeah. This is a good challenge. Love it. Here we go.

Alex 01:35

(singing)

Alex 02:33

That was perfect for this deprescribing super special. I had to make one little edit there. I changed a word. Some of you may have noticed that.

Emily 02:42

A new anthem for deprescribing.

Eric 02:44

That’s right. I didn’t notice. I’m not a Pink super fan.

Alex 02:47

So there was a word...”I tried to call a nurse again, but she’s being a little…” I changed it to “rich” [laughter].

Eric 02:56

Okay. This is like our, what, fourth deep prescribing super special? We haven’t yet figured it out. Why is it so hard? Well, I guess the question to all of you is, why is deprescribing so hard? It seems very easy to prescribe medicines, and we do that very well. Why is it so hard to deprescribe? In your thoughts, who’s up for taking that one?

Amy 03:21

At the risk of repeating what you’ve had said on your prior deprescribing ones, it’s hard to undo things. And the analogy that I always give with deprescribing is the idea of cleaning out your closet. Like, it sounds like a great, you know, rainy day activity. You take everything out, but then somehow your prom dress still ends up back in the closet because you might need it one day.

Eric 03:42

That’s the hard thing.

Alex 03:43

Right.

Eric 03:43

Like, I’ve seen a lot of, like, benzo deprescribing interventions, and while it may work during the intervention, six months, a year later, a lot of people are back on it.

Michelle 03:56

Yeah.

Emily 03:56

There are so many forces acting against us. Right. The system is not set up to deprescribe. It’s not easy to communicate changes once you’ve made them. The patients don’t always know what medications they’re on. I mean, you know, there’s tons of studies on why this is so hard and what the barriers are. We could have a whole episode just on, you know, why it’s so difficult.

Eric 04:20

Yeah. It does seem, though, there. There are barriers to taking your pills. You have to remember you got to pick it up at the pharmacy. Sometimes you have to pay for it, sometimes you have to pay a lot for it. There’s all of these also barriers that seem to be like, it’s very hard for me to remember to take any medicine, but when it comes to these medicines like benzos and gabapentin and PPIs, man, people just stick with them.

Emily 04:47

I mean, some of them are very sticky. Right. And we don’t have a lot to offer sometimes other than a medication. So, you know, we don’t have good access to physiotherapy. I’m from Canada. You know, it’s not necessarily covered or accessible for everybody. And so it is just so much easier to prescribe a pain medication than to get somebody into physiotherapy, just as an example.

Alex 05:09

Yeah, yeah, I’ll say. There are so many points at which deprescribing initiatives could have some Benefit, but because it’s so multifactorial, that’s why it’s hard to get people off medications. If you just focus on one point, then you’re probably not going to address everything. There’s no magic bullet failure point that we could address here. The other thing I’d just. In my perspective, the other thing I’ll say is what I’ve learned from these podcasts is that there’s this big sort of cultural momentum towards prescribing in Western medicine.

And there’s so much work on, you know, you’ve got to take these medicines, you got to prevent this treat, that there isn’t that same sort of clinical momentum towards deprescribing. And I think of the analogy of, like, cancer screening. You know, we got to prevent cancer. We got our pink ribbons. We’re marching like, that’s the focus. The stopping the cancer screening doesn’t have that same sort of momentum. I wonder if our guests feel that that’s maybe sort of an underlying issue here.

Emily 06:11

Yeah, we definitely have a pill for every ill. Go ahead, Connie.

Connie 06:15

Yeah, I was just going to say that I think that there’s certainly a lot of expectations, right, that these pills are going to do something. And so one of the things that we were thinking about is like, well, if we want to get people off of these pills, do you have to address those expectations? You know, there are a lot of placebo effects out there, and there’s a lot of expectancy that when you take this pill, it’s going to help you. And there have been a lot of deprescribing studies for benzodiazepines and Z drugs where you’re really targeting the pharmacological aspect, but not those expectancies. Right. And so that could be part of it.

Eric 06:53

So you do a randomized control trial. You have your placebo group and your control group and your intervention group. Let’s say it’s a Z drug. And you see at like, two weeks, both groups improved by 30% on some symptom score. For insomnia, we see this with pain scores. Intervention after intervention, when you do these for symptoms, there’s like a 20 to 30% response rate in the placebo group. Is that what you’re talking about, Connie?

Connie 07:24

Yeah. And then for sleeping pills, it’s even higher. There’s this one meta analysis that found that 63% of. Of the drug responses were seen even in the placebo groups. 63%. And then there’s another study where they look at the placebo versus, like, no pill. So. And they were finding with the placebo group, there was a higher response than with no treatment. So, I mean, there’s clearly, there’s something going on there.

Eric 07:53

I wish I had that because I took Ambien on a flight to Australia because I can’t sleep on a plane, and I still can’t sleep on a plane. Ambien help.

Emily 08:03

I just want to say a friend of mine took Ativan on a plane and woke up and got a DVT.

Alex 08:08

Oh, no.

Emily 08:10

You know, I guess things could be worse.

Eric 08:16

So, Connie, you actually did a study looking at this placebo response, but instead of, you know, starting somebody on an insomnia pill and putting them on a placebo, it was kind of the opposite. Can you tell us what you did in your study?

Connie 08:33

Yeah. So we, we thought, well, you know, if we want to get people off of the sleeping pill, then we need to try to tackle some of these. These mechanisms that are driving the placebo response. And so, so we had a randomized trial where one arm got a masked or blinded tapered. So they didn’t know exactly how much they were taking from night to night during their taper, along with did they.

Eric 08:59

Know they were being tapered at all?

Connie 09:01

They did. So when we started off, we said, you know, the overall goal is going to be to get you off. But they didn’t know the sort of the rate of tapering. And in addition to that, we also had specific cognitive exercises that were targeting sort of these expectations for the pillar. And the comparator arm was a traditional taper, where we gave them a bottle and a pill cutter and a taper schedule. And both arms got what would be considered first line treatment for chronic insomnia disorder, which is cognitive behavioral therapy for insomnia. So both arms got cbti, but one arm got this blinded taper and the other got this open taper.

Eric 09:45

And the CBTI was slightly different, right? In the group that got the blinded taper?

Connie 09:51

Yeah, we call it augmented cbti. So both groups got the standard multi component stimulus control, sleep restriction, et cetera. But the group in what we call the empty cap group, they got these additional exercises, and then also they had this special diary where we would ask them every night, night or the next morning. So how much relative to your baseline dose, percentage wise, do you think you were taking?

Eric 10:19

Were they able to guess it?

Alex 10:21

Yeah.

Connie 10:22

So, you know, some people were thinking they were taking more and others less. And we also include in this diary a list of questions for the next day. To see like, you know, how they were feeling and then also how they were sleeping. And during the sessions, therapists would try to, you know, kind of probe at their expectations towards the pill. So like, if somebody said, you know, I slept terribly, you know, and they thought like, well, you know, it’s because you cut my, my dose. You know, they’d be like, well, how do you know? How do you know? How do you know that what was in there?

Eric 10:55

And you did that for like nine weeks and the last week they, you actually told them what their dose was. Was that right?

Connie 11:02

Right. Yes, it was a. Basically like two months. And then at the end there’s what we called the unveiling. So then they would show their responses versus the actual. And they’re, you know, different patterns emerged as far as, you know, some people were kind of overestimators, others under, and others were maybe closer. But you know what, we really wanted them to because it’s like there’s the two month intervention, but then we were following up six months after.

Eric 11:29

So every time like eight months after the. So you do two months and then six months after the two months, right?

Connie 11:36

Yeah.

Eric 11:37

Okay.

Connie 11:38

So, you know, every time they were going to go and take their pill, let’s just say, like they were still taking at the end, we wanted them to kind of question, okay, like, so what is this going to do for me? And how do I know?

Eric 11:49

Yeah.

Emily 11:50

Did you have people drop out of the study because they thought they were being tapered when they weren’t?

Connie 11:55

We actually had a really great response, like follow up rate. So most people stuck with it. I mean, of course we did have some people who dropped out of an intervention, but didn’t necessarily drop out of the study. So we’re still able to follow them up. And yeah, I mean, I think for some people it was anxiety provoking to not know exactly what they were taking.

Amy 12:18

My question to Emily, which is how did you get people to sign up to begin with?

Connie 12:25

Right. Yeah. Actually it was because both arms were getting treatment for their insomnia. I think that was, it was really helpful.

Amy 12:33

Medication treatment or the CBTI treatment?

Connie 12:35

The cbti. So, so, so that was, that was helpful to, to keep people in the study or to sign up, rather.

Michelle 12:44

I’m curious about what proportion got their group correct. How many knew that they were being tapered and how many were wrong in that guess?

Connie 12:53

Right. So. So both, both, you know, arms, they were all being tapered off. But we’re actually still analyzing all of the different patterns in terms of like, which percentage Tended to be like overestimators versus underestimators. But I had a terrific MSTAR medical student working with me this past summer who was looking to see whether there were the type of medication, like the half life or was a Z versus a benzo. Which factors were predictors of being like overestimators or underestimators. And it turns out that none of the actual properties of the meds were predictors. So we’re still trying to figure that out.

Alex 13:34

MSTAR medical students in aging research apply now. I think the application medications are due soon, Eric.

Eric 13:39

Okay, I gotta find out what actually happened. So at six months. So really eight months from the start of this trial, six months after you stopped the intervention, were people able to stay off the benzos?

Connie 13:54

Yeah, so we found in the empty cap arm, the mass tapering arm, 73% were off. And that’s six months after they stopped.

Eric 14:01

Three out of four.

Connie 14:02

Yeah. Yes, it was great. And in the comparator arm, which was, you know, they were getting basically two months of CBTI with an open taper that was 59%. So we had a larger percentage in the empty cap arm who had discontinued. And it was significant. Both arms, their insomnia improved, which is great. I mean, they were off their meds and their insomnia. Yes.

Eric 14:32

I mean, those are pretty impressive numbers in both groups, highlighting that you can stop these medicines. CBT does work and your insomnia continues to improve.

Alex 14:45

And is it ethical to do a placebo taper in real world clinical practice? Ethical and practical. I guess those are the questions.

Eric 14:53

Because you need somebody compounding the placebos too, right?

Connie 14:57

Yeah. So our study, we actually had a local compounding pharmacy and. And I would write the prescription, which is that. That’s actually a challenge because the amount of writing that goes on, I was writing paper prescriptions, it barely fit with my chicken scratch. But yeah, so we had a local compounded pharmacy prepare the meds. And the participants, they were okay with not knowing. So I think that’s the key thing.

Emily 15:26

Right.

Connie 15:27

It wasn’t like we just sort of did something on the slide. So as long as you think you.

Eric 15:32

Could do this in clinical practice.

Connie 15:35

Absolutely. Yeah.

Alex 15:36

Yeah.

Connie 15:37

Yes. I mean, we can. We had the local compounding pharmacy, we have patients or participants who say that they’re interested. I mean, the Laura, Carolyn and big shout out to Dr. Carolyn, who she and I were sleep fellows together and she’s a psychiatrist in private practice. And she actually did the feasibility study up in Portland with a local compounding pharmacy through with her Private practice practice. We put it through irb, but, you know, so it just shows that you can actually do this.

Eric 16:06

And people knew from an ethical standpoint. Right. People knew that they were being tapered.

Emily 16:12

Yeah.

Eric 16:12

And it was a mask taper, but they didn’t know when they were being tapered or how much they were being tapered. Is that right?

Connie 16:18

Right.

Emily 16:19

And I think it would be ethical in the real world because it’s done already. It’s done for allergy testing, but a food allergen, they’ll come in and they encapsulate and you don’t know, but you’re told ahead of time. You don’t know if you’re getting placebo or your allergen. So I think it is already being done.

Alex 16:35

It’s a great point.

Eric 16:36

Okay, Connie, I got one last question about your article. The percent stopped was significant between the two groups, but the dose and the frequencies of taking these benzos weren’t significantly different. Why is that?

Connie 16:51

Well, I mean, I think at post one of those was significant, but I mean, there are so many people who came off of the Met. So if you look at the, you know, if you have zero versus zero.

Alex 17:02

Yeah, yeah. It doesn’t matter. All right, I love that.

Eric 17:05

So I guess one question then is this was a pretty intensive thing. So you had these nine weeks intensive CBTI therapy. You were writing these prescriptions. Is there another way for maybe more resource limited places? And this brings us to the question of patient directed education. Emily, you did a paper looking at patient directed consumer educational brochures around gabapentin deprescribing. Do you want to tell us a little bit about what you did?

Emily 17:40

Yeah. So I’m the director of the Canadian Medication Appropriateness and Deprescribing Network, Kaden, and we had previously made brochures that we mailed out to people telling them how to get off their sleeping pills if they were interested. And so we adapted that brochure for gabapentinoids because there’s a big problem of overprescribing of gabapentinoids in Canada and North America. I’m an internist, so when I hospitalize patients, 1 in 8 of my patients now takes pregabalin or gabapentin.

Eric 18:15

1 in 10 older adults in the U.S. 1 in 8 in the U.S. in the U.S. 1 in 10 older adults. All older adults.

Emily 18:23

Yeah. It’s one of the top most prescribed medications. Now I, you know, out of curiosity, sometimes I would ask my patients, do you know what this medication is and what it’s for a lot of times they didn’t know why they were taking it. A lot of times they couldn’t say whether it had helped them. So they didn’t know if they were taking it for pain. They couldn’t really say if it had improved their pain. So, you know, we thought that was interesting. So we took these brochures and said, you know, practically speaking, it is.

Again, we talked about, it’s hard to deprescribe in the clinical setting, acute care, that’s where the study took place. Like there’s a lot of competing work to be done. So we said, well, what if we just hand them brochure and just let them know, hey, did you know you’re on this medication? These are some of the things that gets prescribed for. But the evidence is not great that it helps people. Here’s some of the harms, if you’re interested. You’re in the hospital right now, we can write you a taper, just mention to your doctor, mention to the team that you got this brochure and here’s even a suggestion for how to do it.

We did that and we compared it to just what happens with usual care, which of course is not much. Just found that this was a really easy, low budget way of informing people about the medication, about the potential harms, and they were a captive audience in the hospital. Maybe even had come in with a complication. Right. Of the medication. Maybe a fall or delirium.

Eric 19:51

Yeah.

Emily 19:52

You know, some complication that might have even like, you know, kind of triggered that thought process about maybe it’s time for me to come off of the medication and just. Yeah, we found that it was a pretty easy way to increase deprescribing just by giving out the brochures.

Eric 20:07

So low cost. There was 11% difference between the two groups. Usual care versus it was like 10% versus 21%.

Emily 20:17

Yeah, exactly.

Eric 20:18

Yeah, but limitations. This was not a randomized control study, right?

Emily 20:23

No, it was not. No, it wasn’t a randomized control trial. We have staff that circulate between the hospitals and part of it is if the staff start to learn about it, there could be some contamination of your control. So we used patients who hadn’t seen the patients from before, and then patients from after got the intervention. So we did a controlled before and after study. So we kind of just looked at usual care up until we had recruited a certain number of patients, and then we switched over and started doing the intervention so that there’d be no contamination.

Eric 20:51

And your deprescribing was on hospital discharge. How many deprescribed or is it later on?

Emily 20:57

We looked at eight weeks. Yeah. So we called people up at eight weeks and they had to have either come off of it or have had a prescription in hand with a dose down to zero eventually. So someone was intending to get them completely off of it as opposed to like a dose reduction, for example.

Eric 21:13

Great. And then the adjusted analysis number needed to treat of like eight.

Emily 21:17

Yeah. So we did. We increased the absolute risk difference increase of 13%.

Alex 21:23

That is an amazing number needed to treat. Suggests we should be doing this, especially because it’s low cost, low risk. Why aren’t we? Or are we? Is this what’s happening in Canada? I don’t know, Emily, what’s going on?

Emily 21:36

Yeah, we have them available now on the ctu. They’re totally available online and we’ve started translating them into multiple languages. We’re starting to adopt them as well for different cultural communities. So it’s a matter of, you know, if, if you have a willing prescriber, if they’re willing to print it out, it’s. It’s there and you can do this.

Eric 21:57

So can our listeners download those right now?

Emily 22:00

Absolutely, yes. Yeah, they’re at the Canadian Deprescribing Network website. If you just Google Caden, we’ll have.

Eric 22:07

A link to that on our show notes.

Alex 22:09

Yeah, very well.

Eric 22:10

Wait, Alex, there’s more, there’s more, there’s more. Because if you act right now, you can also look at three different drugs and patient directed education on deprescribing PPIs, gabapentin and other hypoglycemic agents. Is that right, Amy?

Amy 22:30

That is, yeah, sort of. Right on the heels of Emily’s work, we used essentially the same brochures in discussion with Kaden. The PPI brochure, the hypoglycemia risk brochure, which was adapted slightly, which I’ll come back to, and explain a little bit more about the rationale behind it. And then we also had a gabapentin brochure. Although it was created at the same time and albeit similar, it was a bit more tailored to the VA and options that we had in the va, because as Emily was saying about the healthcare system in Canada, to get into some of the non pharmacologic ways of treating pain is a little bit more difficult. =

Within the VA, there’s a program called Whole Health, which is really about trying to engage patients in many different modalities for pain management and just self care. So acupuncture, Tai Chi, Yoga, mindfulness and physical therapy, chiropractor. And so those are all options that are available. And I can actually refer my patients directly to. So we wanted to look at gabapentin, but have it all these other options available. One distinction though is we only included, and I’ll come back to some of how we did it and other differences, but we only included those taking 1800 milligrams a day or higher. And so I noticed, Emily, in your study it looked like it was around 600 a day as the average dose on the gabapentin.

So we were looking at really high use and doses above 1800 milligrams per day that I have seen Patients of mine prescribed 4800mg per day. There’s no extra benefit with just a whole lot of extra risk. And so we were really just looking for the patients who are at doses beyond what they’re going to get any benefit from and trying to get them to a safer dose. And so we were interested in gabapentin from that perspective. And then the other that we were really interested in was for diabetes. And we were talking a lot at the top of the hour about a lot of the patient barriers, patient related barriers to deprescribing. But I think there’s a lot on us and I don’t necessarily think that we’re doing anything with ill intent because we want to do right by our patients.

And there’s something that feels better about doing something than doing nothing, especially if someone is on something and hasn’t had a side effect, then you’re like, oh, you’re fine with it. So we really wanted to address this risk for low blood sugar in older adults with diabetes who had an A1C below what we would consider like safe for that age. And taking two medicines for diabetes that are pretty decent risk for landing you in the emergency department with low blood sugar, you know, not to say like to not treat diabetes, but what were those two medicines? So insulin, which is the injections, and then sulfonylurea. So primarily in our study, glipizide, and.

Eric 25:22

It was a hemoglobin A1C of like less than 7, less than 7.

Amy 25:25

They also had to be have at least also than one of the following older than age 65 or cognitive impairment, because again, like we’re asking patients who their time to benefit with some of these medicines also is not necessarily there and the risk is much higher or renal problems, kidney problems. So these are like really, really, really, really well treated to the point that the risks are probably now, but they.

Eric 25:49

Didn’T have to have a hypoglycemic episode yet.

Amy 25:52

Exactly. So we want, the whole thing is we want this to be proactive, like, let’s get people to stop taking the risky medicines before they land in the emergency department. But honestly, a large part of the diabetes part we think is a little bit more on the clinician.

Connie 26:06

Right.

Amy 26:07

When you start someone on an insulin or diabetes meds, you really, really instill in the patient, like, oh, you have to take this every day. You probably take it for the rest of your life. There’s no exit ramp. And so we don’t prime patients to think about stopping in the future. And there are metrics about how clinicians get evaluated in terms of their diabetes that people misinterpret of what those metrics are. And so we just keep going. And so we really wanted to have a medicine group that was around that we thought was a little bit more driven by the clinician behaviors and a little bit less about the patients. And so that was the main adaption that we did for that one.

Eric 26:43

And this is in primary care, right?

Amy 26:45

Right. So we did primary care. You know, maybe egocentric, I’m a primary care doc. But we also, most, most routine, prescribing, most renewals, chronic meds are going to be happening in the outpatient setting. And what we really wanted to do was make this to not interfere with the clinicians. It was extremely low touch for the clinicians. It wasn’t a clinical reminder. There was no pop up. There was nothing that blocked you from proceeding in clinic. And essentially we mailed the brochures to using, you know, old fashioned postal mail. We think they arrived and the they were sent about two to three weeks in advance of an upcoming primary care visit.

And that was intentional. We didn’t want to just do a blanket, you know, send it to everybody on, you know, in the clinic who might be eligible, because that might inundate the clinic with phone calls or questions of like, why am I getting this brochure? Why is this, you know, medicine potentially bad for me? So we really wanted to time it in advance of that clinic visit so they could come in and raise the question with their clinicians. And similarly, so we looked for our outcomes at six months and essentially either stopping the medicine or just going down on the dose counted. And at six months, we found success across all the groups, about 1.2 times the odds for deprescribing in the group that got the intervention compared to those who did not.

There’s no difference across the different medications in terms of the effectiveness of the brochure and what I will say separate from the brochure though, if we were just to look at the medication groups and how they thought about or what their rates of deprescribing were. The biggest deprescribing rates, although it was the smallest group, was actually in that high dose gabapentin group.

Eric 28:30

Huh. And deprescribing as far as anything, you could stop it, but you could just go down on the dose.

Amy 28:37

You could go down on the dose. The only because of technical measurement issues, the insulin one was either start, you either had to stop it or continue. There was no dose reductions. Just because it’s really hard to capture in prescribing data how much people are.

Eric 28:50

Taking and how big of a magnitude is this? What percent of people actually had some deprescribing in the two groups?

Amy 29:00

So we actually had a really higher than expected baseline rates of deprescribing, about 25% and then it went up to about 29% in the intervention group. But that absolute difference of 4%, and we know at least in the PPI group, the proton pump inhibitor group, those are so over prescribed that even if we’re just talking about a small absolute magnitude difference here at the population level, it could save people from taking meds that just they’re not getting benefit from, they’re potentially getting risk and save, you know, their effort, our effort, and let us focus on the things that the meds that probably matter more for their continued health.

Alex 29:40

And did you consent people for this study?

Amy 29:43

We didn’t know, we did not consent. It was done as a pragmatic trial because we weren’t changing the medicines. All we were doing providing information and we did give a brief overview via email to the clinicians, basically saying, hey, your patients might come in with this. Here’s some information about deprescribing, here’s some information about these medicines. You make the clinical decisions. We cannot tell whether it’s appropriate or not completely from administrative data. We just want to educate the patients. And now you guys can make a more informed decision about whether you should or shouldn’t continue this medicine.

Alex 30:20

Any pushback from the clinicians or the patients?

Amy 30:24

No, we got very little pushback from clinicians to the extent that when we tried to interview them afterwards to see how it impacted their practice, very few clinicians, remember, even were aware that this had happened because it was so minimally invasive with their clinical practice. We did do a survey of patients Afterwards and just. But more to understand how they felt getting the brochure, whether they, that led to them talking about their medicines, basically. Were we able to activate patients to be more engaged in their healthcare and healthcare decisions?

Eric 31:00

I got a question for all three of you then. So on one hand you have Connie, high touch, high yield, I mean what 70 something percent deprescribing, like a huge deprescribing versus Amy and Emily kind of low touch. You’re just sending that as brochure. Smaller yield, like what’s the right approach? Or is it both thoughts on that? I’m going to open it up to the group.

Amy 31:29

Yes, both. I mean, why not?

Eric 31:33

But is there a group that you should target more for the high? Like you can’t do high touch for everything.

Emily 31:39

But I mean I think like your first wave is, you know, like Amy’s and my approach. Right. And then maybe you have other people who are more resistant or who need more oversight for their taper or who try, maybe they’ve tried it out and said, listen, I tried but that eight week taper, that one size fits all you have, it’s not for me, it was too fast, I need more support, you know, so then you need to go to the more high touch intervention, I think. But you could just start as a first wave, you know, using the brochure and see what you catch. But for sure there’s going to be people who need Connie’s approach.

Eric 32:14

Yeah, I love that. Connie, would you agree?

Connie 32:18

Yeah, I think there’s definitely a need for a variety of different approaches. It’s interesting because we tried, we had another study using a low touch approach. We mailed out the Empower Light brochures within the va but we also, it was a randomized control trial. So we mailed out just some general information about sleep with a link to a site that was just information about sleep. And the other, the Empower brochure was enhanced with a link to online CBTI that’s sort of free and anonymous and we unfortunately did not find a significant difference. Not sure why it was during the height of the pandemic. Could have been people weren’t seeing their providers. But I do think though that there’s still a role for like a low touch, you know, medium touch, high touch.

Emily 33:10

Well, and Justin Turner mailed out the brochures in the province of Manitoba for opioids.

Michelle 33:14

Right.

Emily 33:15

This study was published I think maybe six months ago and that, that did not work. So, you know, big population level study using the brochures and, and there Was. No, wasn’t really a significant difference in deprescribing of opioids. So, you know, the medication class might matter as well.

Eric 33:33

Yeah. Well, speaking of medication classes, like, we have some drugs, probably all of us are in agreement, worthwhile stopping. You got your benzos, you got your gabapentins, you got your PPIs. We have other drugs a lot of people are taking, like, antihypertensives that, you know, if you look at the SPRINT trial, older adults improves lots of outcomes, including mortality, potentially cognition. But is this an area that we should also be thinking about deprescribing, especially in fro older adults? And that brings us to our last study, Michelle. Michelle, you looked at this for nursing home patients deprescribing antihypertensives, is that right?

Michelle 34:15

That’s correct, yes. As you’ve said, we know blood pressure control is a cornerstone of cardiovascular prevention and even dementia prevention. In midlife and even in older age.

Eric 34:29

Everybody gets their blood pressure checked. Very first thing you do when you visit a doctor’s office. Weight and blood pressure.

Michelle 34:35

Absolutely. I would say it’s probably the most measured physiologic parameter and strangely still one of the least understood, especially in older adults with multiple chronic conditions.

Eric 34:48

Sprint actually excluded people in nursing homes, so we don’t have any good data on blood pressure control in nursing home patients. So frail or older adults.

Michelle 34:58

That’s correct. So SPRINT excluded nursing home residents, people with dementia, and of course, diabetics. So a huge proportion of the group that I’m interested in studying, and that is people who are older adults with multiple chronic conditions for lt who often reside in a nursing home.

Eric 35:19

Yeah. So you looked at this population. What did you do in your study?

Michelle 35:24

So we did a little different study design than the other articles we’ve heard about today. So we did a completely observational study. We were trying to mimic or emulate the trial that we would like to do. So, of course, we’d love to do a large randomized controlled trial where we randomize people to either maintain their antihypertensive regimen or or deprescribe. But this can be a difficult population to enroll in RCTs, primarily because of the vulnerability of the population. They have a lot of chronic conditions.

They might not be able to complete the trial protocol. So we leveraged observational data in which there’s already a lot of deprescribing happening. In our early studies in the same nursing home population, we found about 70% of residents had a deprescribing event of antihypertensives during their stay. And that was a stay of a median length of six months. So we’re seeing that there’s a lot of deprescribing happening of antihypertensives in this population, but very little RCT evidence to tell us about the benefits and harms of that deprescribing.

So in our study, we used observational data to observe residents who were deprescribed. And that’s a reduction in dose of 30% or more or a reduction in the number of antihypertensive medications. And then we used statistical methods to try to control for differences in the characteristics of people who deprescribed and did not deprescribe because of course it’s a non random event. So that’s where we have to turn to our friends of statistical models to try to balance the things that might lead someone to deprescribe. We used all our best methods available and we found that those residents who deprescribe their antihypertensives actually had a little bit of a slowing in their cognitive decline.

And we saw that that slowing of the cognitive decline was even stronger in those with dementia. The effect size was pretty mild. It wasn’t something that I think should be necessarily used as a cognitive preservation intervention. But we at least know that this is reassuring for providers and patients who are interested in deprescribing that we can say it looks like we’re not seeing any harm for your cognition and it might even improve cognitive function.

Eric 38:03

And there was another study, right? A very like a 14 or 16 week study looking at deprescribing antihypertensives in the nursing home. Was that a randomized control study? I think you mentioned it in your discussion section too.

Michelle 38:17

Yeah, there have been a couple studies. There have been the DANTE trial as well as the optimized trial, and both have shown really no harm in deprescribing antihypertensives. There was a newer trial which it was a little less conclusive. It’s called Danton, and it was just published mid last year, mid 2024. And unfortunately, that trial was stopped early because of some current concerns about increased adverse events in the deprescribing group. And that was also in nursing home residents.

So, you know, I think if anything, all of these, the small amount of evidence that does exist is just highlighting the uncertainty and the need for more randomized controlled trials and more high quality observational studies of depreciing to help providers and their patients make this decision.

Alex 39:10

Can we, because you’re an expert epidemiologist and have explained things to us in our division clearly in the past. I remember you explaining, I think inverse probability treatment, waiting at some point. This was a, this was like a target trial emulation. Is that the right, Am I emulated target trial? Is that the right words? And can you explain this to our readers? I mean our listeners and readers? Some people read emails. Yeah.

Michelle 39:37

So as I mentioned, in an ideal world we’d like to do a trial, but in situations where a trial is not feasible for multiple reasons or it just simply hasn’t happened yet, we can use observational data to emulate the trial we would like to do. So a couple things make this different than like maybe some of the observational studies that you’ll see in other settings. One is that we always want to be studying a change, right? So in this case we were studying the people at the moment of deprescribing and looking at those who deprescribed versus those who maintain their antipretensive regimen.

And the reason we want to be studying a change is because we need to control for the pre change characteristics that led to that change. In epidemiology, we call those confounders. We need to be able to look temporally at what happened before either the person de prescribed or stayed in their treatment group. Then again, we use statistical methods to try to control for those differences. That’s one feature of a target trial emulation is that you’re looking at longitudinal data and you’re looking at controlling for confounders before some change was made and then looking at events that happened after that change.

Alex 41:00

Now is that an advance over like great observational methods or is this like observational data dressed up to look like a randomized trial?

Michelle 41:10

You know, you could say that target trial emulation is just a well conducted epidemiologic study. Like there’s nothing magical that makes a target trial emulation. It’s just a collection of design principles and analytic methods that try to mimic a trial. In thinking about things like selection, the time of initiation or change in medication, looking at the length of follow up and looking at well defined outcomes, there’s nothing magic that puts it into the target trial realm other than this like set of principles of design and analysis.

Alex 41:54

Yeah, and one of the, in defense of emulated target trials, you’re using real world observational data. If you had a randomized trial, it’s likely that there are some patients who would not agree to participate or who you might exclude from that trial who you would include in an observational data set of patients with multimorbidity dementia residing in nursing homes.

Eric 42:20

But it also comes at the cost, right, that fundamentally there’s a difference between the group of people that we decide to stop these medicines and the group that we decide to continue them on.

Alex 42:32

And you don’t think you can balance.

Eric 42:34

All those things really.

Alex 42:35

You can balance for. Right. The Donald Rumsfeld. I heard that Stephen Colbert talked about this last night. You can balance for the knowns, the known knowns, but you can’t balance for the known unknowns that you can’t measure or the unknown unknowns. And if it was a real randomized trial, it would be sorted. The reason Stephen Colbert was talking about it, by the way, is because he said there’s another category called unknown unknowns, unknown knowns, which is when we know something because we had the same president previously and then we unknow that. Michelle, thoughts about randomized trials, not your.

Michelle 43:13

Thoughts about $I don’t want to tackle politics, but, you know, we’re all taught that randomized controlled trials are the, you know, the pinnacle of evidence because of this key feature of randomization, balancing, confounding. And yes, that’s a wonderful feature that is critical for causal inference. But the huge limitation, as you alluded to, is that randomized controlled trials, even the best designed RCTs, don’t represent real world patients. You know, if you look at Sprint, for example, you know, Sprint represented a fraction of the community dwelling older adults who are actually indicated antihypertensive medication.

And so even though Sprint was an, you know, an exceptionally well conducted trial, it does the results might not apply to community dwelling elders and especially nursing home residents because they didn’t include any nursing home residents. And we know that the calculus of benefits and harms depends on characteristics of the patients. So I think RCTs have been oversold to us a little bit because of this nice feature of randomization. But what we’re trading off for that is selection bias. And I’m not a clinician, but from what I hear from my clinical colleagues is they’re concerned about treating patients to guide to guideline level standards when the guidelines were derived from RCTs that don’t look like their patients. And that was really what motivated me getting into this research is wanting to help my clinical colleagues make better decisions when they didn’t have evidence in people that looked like the patients that they were prescribing for.

Eric 44:55

I love it. Michelle, I’m wondering if I could turn to all four of you now, kind of magic wand. If you had a magic wand, after listening to everybody else’s articles and doing your own, what would you use that magic wand on around the issue of deep prescribing in older adults? Connie, I’m going to turn to you first. We’ll go in order of how we did your articles.

Connie 45:17

Gosh, it’s so hard. I mean, I feel like there’s so many different factors. Right. So I think the targeting the prescribers, really trying to reach out to the patients, doing both at the same time. Does my wand cover like the full spectrum?

Eric 45:33

It could cover big or small. So what would you do if you were focused on the prescribers? How would you reach out to them with the wand?

Connie 45:40

I mean, for me, I’m really kind of intrigued by this placebo effect. So I would want to see where we can tackle that more. Right. And try to see how we can leverage mass tapering or trying to target expectancies. That would be my thing.

Eric 45:57

Oh, that’s great. All right, Emily.

Emily 46:01

I think I’d go for the public, you know, big multi million dollar campaign just to make people aware that medication isn’t the answer. That, you know, every medication has side effects that if you can, for some of the conditions we’ve talked about today, that, you know, you’d want to do what you can to avoid medications if it’s safe to do so and like really just increase the public’s knowledge on this topic. I think it’s at a love it.

Eric 46:28

For every drug company ad, there has to be another drug company sponsored pharma deep prescribing ad talking about why you should stop this drug.

Emily 46:37

Yeah, big public awareness campaign.

Alex 46:39

It’s great.

Eric 46:40

Amy.

Amy 46:42

I think I would mine actually attack approaches, both the patients and the clinicians. I would want to change sort of people’s psychological comfort with this idea that if you stop something or reduce something, not only might you have no additional harm like Michelle was showing, but that you can also actually have benefits that this isn’t going to be a we’re taking something away and you’re going to do worse. Like we’re taking something away and there’s a lot of benefits that go with it and it’s worth trying that Wonderful.

Eric 47:13

Michelle.

Michelle 47:15

I’m going to go really big with my magic wand. You know, I think kind of like we were talking about at the beginning of the show, I think a lot, a lot of resistance to deprescribing comes from the fact that our society is fairly ageist and we don’t really want to talk about or understand the changes that happen in bodies when they age. So I think if we could just embrace aging as a change that happens to all of us, should we be fortunate to go through go through it. We would be more open to deprescribing that accompanies.

Eric 47:42

That was great. Love the magic wand. How about a little bit more pink?

Speaker 7 47:57

I’m lying here on the floor where you left me I think I took too much I’m crying here what have you done? I thought it would be fun I can’t stay on your life support There’s a shortage in the switch can’t stay on your morphine Cause it’s making me itch I said I tried to call the nurse again but she’s being a little rich I think I’ll get outta here Where I can run just fast as I can through the middle of nowhere to the middle of my frustrated fears and I swear you’re just like a pill Instead of making me better you’re making me ill Just making me ill.

Eric 48:54

Thank you all for being on this podcast. Learned a lot. Love the work that you’re doing.

Alex 48:59

So creative too. Thank you.

Michelle 49:01

Thanks for having us.

Connie 49:02

Thank you.

Amy 49:02

Had a great time.

Eric 49:03

And to all of our listeners, thank you for your continued support.

***** Claim your CME credit for this episode! *****

Claim your CME credit for EP340 “Deprescribing Super Special III”
https://ww2.highmarksce.com/ucsf/index.cfm?do=ip.claimCreditApp&eventID=14880

Note:
If you have not already registered for the annual CME subscription (cost is $100 for a year’s worth of CME podcasts), you can register here https://cme-reg.configio.com/pd/3315?code=6PhHcL752r

For more info on the CME credit, go to https://geripal.org/cme/

Disclosures:
Moderators Drs. Widera and Smith have no relationships to disclose.  Guests Constance Fung, Emily McDonald, Amy Linsky, and Michelle Odden have no relationships to disclose.

Accreditation
In support of improving patient care, UCSF Office of CME is jointly accredited by the Accreditation Council for Continuing Medical Education (ACCME), the Accreditation Council for Pharmacy Education (ACPE), and the American Nurses Credentialing Center (ANCC), to provide continuing education for the healthcare team.

Designation
University of California, San Francisco, designates this enduring material for a maximum of 0.75 AMA PRA Category 1 credit(s)™. Physicians should claim only the credit commensurate with the extent of their participation in the activity.

MOC
Successful completion of this CME activity, which includes participation in the evaluation component, enables the participant to earn up to 0.75 MOC points per podcast in the American Board of Internal Medicine’s (ABIM) Maintenance of Certification (MOC) program. It is the CME activity provider’s responsibility to submit participant completion information to ACCME for the purpose of granting ABIM MOC credit.

ABIM MOC credit will be offered to subscribers in November, 2025.  Subscribers will claim MOC credit by completing an evaluation with self-reflection questions. For any MOC questions, please email moc@ucsf.edu.