Beyond fentanyl and midazolam: ketamine sedation in everyday IR

Episode Transcript

Recorded live from the Cook booth at SIR and featuring leading experts in the field of interventional radiology discussing a wide range of IR-related topics, this is the Cook@ SIR Podcast Series.

Good morning. This is Dan Sze. I’m an interventional radiologist at Stanford University. We’re coming to you live today from the Cook Broadcast Booth at the SIR Annual Scientific Meeting, 2026, in Toronto, Canada. And today we’re super excited to have a conversation with Dr. Amy Deipolyi, who is an associate professor and the section chief of interventional radiology at the Charleston division of the University of West Virginia. So good morning, Amy. And tell us first a little bit more about yourself. How did you get here, what are your interests, and where are you going?

Thank you so much. And thanks for being here with me on the podcast today, Dan. So, yep, my name’s Amy. I grew up and trained and had my first jobs up in the Northeast. My first jobs were in New York City. And I found this job opportunity in West Virginia. I work in the state capital at a tertiary care medical center that never had an interventional radiology program before I got there. So I moved there about five years ago and started a program up from scratch. It had a pretty strong vascular surgery presence. So really what I brought was interventional oncology, portal intervention, and have been doing that since. So now we do everything you can get anywhere in the country, Y-90, ablation, TIPS, all of it done right there in the capital. And we’re a group of three now. I’m not alone anymore.

Wow. Okay. So you’ve tripled.

We have— we’re hiring another one. Yeah. Have a team about 30 nurses, techs, APPs, and delivering high-quality care to the people of West Virginia. We basically are a catchment area for the bottom half of the state.

About how many people is that? What’s the population?

Ooh. Not that many millions.

Okay.

The capital itself has about, I think, 40,000 as a population, but there’s a lot of— it’s a big catchment area. I have to look that up.

Okay.

Stumped me on the first question.

Yeah. Well, you may or may not know. Actually, I grew up in Pittsburgh, Pennsylvania, which is not far from West Virginia. It’s like an hour drive. And in high school, we used to make that drive, because the drinking age was lower in West Virginia, and they actually sold near beer, which is a lower alcohol content. This I heard from my friends, of course, I never actually partook myself. So today we’re going to talk about some of the really cool work that you’ve been doing on procedural patient management and in particular on sedation. So all IRs are comfortable with Versed and fentanyl. So you’ve been doing research on ketamine, and ketamine’s been in the news, been in the lay press. Out where I live in California, the tech bros are microdosing on it. So tell us a little bit about ketamine, the history of ketamine, and why aren’t we all using ketamine as the default already?

That’s a great question. So ketamine is not a new drug, as you know. It’s been around for decades, but it’s been resurrected recently because it has many properties, beneficial properties— antidepressant, used in PTSD and in pain management. And what’s really cool about ketamine is that you can achieve a deep sedation— an analgesia— but it preserves your breathing reflexes. So you don’t stop breathing. If you have secretions, you still cough, protect the airway. So it’s kind of an ideal agent in that way. And it does it by being a dissociative. So it works on the NMDA receptor, unlike your opiates, which work directly on the pain pathway, but then have the side effects of the respiratory depression. So it essentially dissociates you from the experience of pain. Now, with that, you get these psychomimetic properties, the hallucinations. And I think because of the hallucinatory effects of the ketamine, people have shied away from it historically, but it’s come back.

We’re looking at it with lower doses or in combination with other agents that reduce those hallucinatory effects to really leverage the positives that you get from ketamine. IRs have been using fentanyl-midazolam since the 1990s. Our field was born in the ’60s, and over the next few decades after that, we kind of followed what GI doctors were doing and that was using an opiate with a benzodiazepine for procedural sedation. And in those years before the ’90s, we basically went to shorter- and shorter-acting agents. And that’s how we landed on fentanyl-midazolam, but we haven’t updated our techniques since the ’90s. So it’s interesting too, ketamine’s used by other specialists. So ER doctors are using it across the country all the time to set bones, sew lacerations. They’re aware of it. It’s just that I think we’re just so busy and so familiar with fentanyl-midazolam that we haven’t looked to update our methods.

So— and you mentioned that fentanyl and midazolam are short-acting and are quickly metabolized. So what about ketamine?

It has fast-on, fast-off, similar to fentanyl-midazolam. The only difference is you don’t have a reversal agent. So I think that’s another thing where people are hesitant, because they’re so used to having that one-to-one reversal agent, but because it is fast-on, fast-off, if you experience a side effect or something, it’ll be over shortly. And because the side effect is not usually respiratory depression, it isn’t as scary, if you will. It’s not something that requires a code.

Tell us more about the hallucination, because I think that’s one of the things that really made people shy away from using ketamine for many decades. Some of my patients pay good money to have hallucinations. So what’s wrong with having some hallucinations?

That’s actually a really good question. And I don’t think objectively it is wrong necessarily. I mean, we’ll have some people— you can kind of tell people who’ve partied before, because you’ll give them the drug, and then they’ll say things like, “Oh my God, I’m tripping.” But they’re clearly having a good time. So I think it isn’t necessarily good or bad, but it just depends on how that experience is interpreted. Now, you have other agents; you’re not giving ketamine alone. You’re going to give midazolam upfront.

For an adult, usually about two milligrams unless you’re worried they’re frail, they’re old, you might do one. You wait a couple minutes, and then you do the ketamine. And by reducing that anxiety, I think you’re kind of like getting them to a good trip. And so we had been doing a midazolam-ketamine combo for thousands of cases, and now we’ve started adding a little bit of fentanyl on top to— in my mind, I’m getting them to a good experience, or guaranteeing it more, because in our data, we’ve shown that about 11% of people will report perceptual disturbances, but that doesn’t necessarily mean that they’re having a bad time.

Yeah. And by perceptual disturbances, how different are these hallucinations from dreaming? Do people realize that, “Okay, this is not reality,” and that can be upsetting to some people, I guess?

Yeah. They’ll say things like, “I can’t see that well,” or, “I’m seeing double,” something like that, or like, “I’m seeing funny things.” I haven’t unfortunately experienced it myself, so it’s hard to tell you exactly.

Yeah.

And I kind of want to, although I don’t want to have to need a procedure, I guess.

Yeah.

But it is curious.

So you mentioned that you’ve done this on a lot of patients now, and you’ve published a retrospective case series and a prospective registry, and you’ve just completed a randomized controlled study that is in the process of being published. So what is your current evidence that you’ve generated?

Yes. So just as a step back, I think this is a very understudied subject in our field. I think the vast majority of our work is in device development, new procedures, clinical efficacy. We really haven’t focused on the patient experience and patient-reported outcomes. There’s a big gap. And I think these days, patients are expecting more from us. They come in, “You’re going to put me asleep, right, doc?” And you’re like, “No.” But now I can kind of say, “Yes, you are going to probably be asleep.” So what happened was, especially in West Virginia, where there happens to be a lot of obesity and a lot of opiate exposure, that fentanyl-midazolam combination was really not cutting it for us in our practice. We encounter the patient that starts screaming in pain in the middle of your abscess drainage, and you can’t stop, and it’s uncomfortable fairly frequently, a couple of times a week. It was upsetting. And that’s what inspired us to get credentialed in deep sedation, to be able to do ketamine.

And that was because of their preexisting tolerance to opiates or—

Yeah.

Okay.

I think so. So we got credentialed, and we saw how effective it was, and I knew the credentialing was coming in a couple of months. So I leveraged that timing, because I knew we would just switch over to do that prospective registry that was published in JVAR. So we— I had— we collected intra-procedural and post-procedural and baseline pain scores on the 10-point scale before and after and vital signs and things like that. The nurses helped me document those things for— There was like 200 fentanyl-midazolam cases, and then when we switched, we did 100 ketamine-midazolam cases, and document what kind of procedure it was, etc. After that, the nurses revolted. They were like, “We can’t collect your data anymore.” They did a huge effort, 300 cases of data collection. That was our prospective registry, and that showed that the— not only the intra-procedure pain scores, but the post-procedure pain scores were lower in the patients that received ketamine. And that is reflected in the surgical literature. So if you get general anesthesia and they include ketamine in your regimen, you will use fewer opiates post-procedurally. So there is an enduring—

Dissociation of some sort—

Well, it has a—

Or analgesia—

—an analgesic effect.

I see.

It’s an anti-inflammatory, actually. So that was the registry, but of course that’s not randomized. So that’s— it’s observational. Even though it was prospective, it’s still not— you’re getting a higher level of evidence. So that was our baseline data that we used to apply for an SIR pilot grant to support a randomized study where we picked certain procedures: lung biopsy, bone biopsy, and abscess drainage. Lung biopsy was chosen because it’s our most common biopsy, and there’s a higher complication rate with pneumothorax.

So I figured that would be a good one to look at, even though it’s not necessarily that painful, but then bone biopsy and abscess drainage being painful. So we opened that study, and we collected almost 300 patients, randomized one-to-one, to receive either fentanyl-midazolam or ketamine-midazolam, and used a validated survey, a portion of the questions from that, to ask the patients about their experience. So that was a study that we did and is currently in submission. In that study, we found, again, that both the intra-procedure pain scores and the post-procedure ones were lower for all three of the procedures, and even at the end of recovery, they were still lower. And then on the questionnaire, on every reported outcome from the patient, ketamine was superior. So they felt less fear. They were more comfortable. They reported that waking up was better. The only thing that was worse for ketamine was that more patients endorsed having hallucinations.

Although, as we discussed, that’s kind of not necessarily negative or positive. So that one is in the works and is going to come out. And then in the meantime, we looked back, because now we’ve done thousands of patients. So we looked over like six months, collected almost 1,000 patients of just safety data. Because I think part of the barrier for a lot of the IR doctors is they want to use ketamine, but they’re having to go to the sedation committee and ask for help. And then I wanted to provide our field with— “Look, they did 1,000 patients, and there were no respiratory problems. They didn’t have to do any reversal agents or call codes, etc.” We have good-quality just safety data.

In your practice, do you share recovery room with the operating room?

No.

So you have your own techs and nurses, you have your own recovery room?

Yes.

And so these people also need to be trained at what to look for in recovery from ketamine versus recovery from opiates?

Correct.

Okay. And even though it seems that many of the effects of ketamine do wear off quickly, but this analgesic effect is prolonged. So are there any other effects that we need to know about that even after a patient’s deemed recovered enough to leave the recovery room, what else could maybe go wrong afterwards? Are we— is there a danger that we’re releasing patients too early from the recovery room?

It hasn’t really affected our discharge times. People— we’re not keeping them longer. Within that recovery space from the procedure, sometimes, not commonly, but you will encounter people who will start to have emergence phenomena in terms of getting very anxious, and it’s treated— we give another milligram of midazolam, and they’ve all— that’s resolved them all. So I think it’s less about seeing over sedation in the sense that they’re not breathing or something like that and more about making sure that you’re addressing that anxiety head on and not letting it get too much so they start to get too perturbed.

So you mentioned that you had to go to a sedation committee to get privileges. So I’m not even sure what I have at my hospital, and I’m sure the sedation privileges are managed by the anesthesiology department. So if I wanted to take this to my home institution and say, “Okay, I want to start using ketamine,” how do I do that?

That’s a good question. So every hospital have some sedation committee, and almost always it’s chaired by an anesthesiologist, but you’ll be surprised. Our sedation committee, it has ER doctors, dentists, oral maxillofacial-type surgeons, other professions that— non-anesthesia doctors that do sedation. So we are, I think, underrepresented. We don’t realize, and we’re probably doing more procedural sedation than almost anybody else. So I think— that was one of the early things I did was I volunteered and got on the committee. So I was like, “We need to be represented,” because they’re going to discuss some of your complications there, and you need to have a voice. So that’s— number one is just identify— Actually, you know what, take a step back. Use your leader nurse, whatever— nurse manager, IR head nurse—and you have to partner with that person and explain to them, show them the data.

We also published a review article in the Journal of Radiology Nursing that talks about a lot of the steps and barriers and nursing aspect, and they will help you find the policies, because there’s some hospital policy written about sedation and about— there’s going to be a policy about how you can achieve deep-sedation privileges. So that’s step one is just getting your hands on the policy. Then you go to the committee, you find out they meet monthly, you get on their agenda, and you present published data and what you want to do. And usually the policies have to be adjusted, because they’ll say something like, “These following types of doctors are allowed to do this.”

So you have to get yourself added and your procedural area added to the locations where it’s possible and allowed. So this takes a few months. And then for us, we had to be proctored on three cases. That’s it. So I allied with the ER doctors who are very— I think that’s the best proxy, because they’re the ones doing quick procedures with sedation. So they fully understand where we’re coming from, and they just attended three cases, and then sign off, and then we got our privileges. But again, it took some amount of months, which is why we were able to do that prospective registry, knowing that it was coming in the future.

And it’s interesting; you mentioned that you need the collaboration of nursing. So there are some issues there. First of all, in your data, you also asked your nurses and techs about whether they preferred cases where patients received opiates versus ketamine.

Yeah.

And there was a clear signal there, right?

Yes.

And so that— the nurses said, “Okay, if I’m going to undergo a procedure myself, I want ketamine.”

Yes.

I think that speaks pretty loudly, that they’re the observers of what’s going on, and they could see a difference.

Yes.

Are the nurses the one that actually administer the ketamine?

That’s a good question. So actually we listed out in that review article. It’s state by state; there are laws governing who’s allowed to physically push drugs. Now, in West Virginia, there was a law that said that a non-CRNA nurse is not allowed to push anesthetics. There’s still a law that says that, but all of these drugs— I think we’re thinking about it wrong, because lidocaine is an anesthetic. That’s a local anesthetic, but fentanyl and midazolam are anesthetics. So I think we need to think about a state that we want to achieve for the patient and various medicines we have to get there. And we have to convince others to look at it with this framework as well. So the West Virginia Board of Nursing interpreted that law to mean that my nurse could not push ketamine, even though they could push fentanyl-midazolam; made no sense.

So I had to go to the nurse board, because in the beginning, I’d have to— they would push the midazolam, I would have to push the ketamine, potentially scrub out to give PRN boluses. It was kind of annoying. So went to the nurse board, had them change their interpretation and publish a memo, then went back to the sedation committee and the hospital leadership and had them change the policy for the hospital. And it’s coming at a good time, because there’s a lot of interest in having ketamine pushed by nursing on the floor for pain. Whether that’s your trauma victims, anybody on the floor, we want to reduce how much opiates we’re giving them, so ketamine is a natural alternative to that. So found partnership with the trauma surgeons, the ER doctors, and got the hospital to basically change it. In other places, not an issue. So it depends on the state where you’re practicing.

But you’ve mentioned out of the thousand patients that you’ve treated, you haven’t had to call a code for oversedation or anesthetic complications. And I think dentists use this too, right?

Yeah. Well, dentists in the community seem to be using a lot of laughing gas like nitrous, and I’m not familiar with that as a drug. I don’t know how—

Sure. Yeah. Yeah, we believe that [laughs].

So I don’t know. It is, to me, when you’re talking to the sedation committee— because a lot of people say, “Oh, my committee’s not going to let me do this.” And it’s like, there are dentists in the community doing maybe ketamine, definitely nitrous with their secretary—

Yeah.

—without all of the machinery that you have in the hospital. Like, if you did have a code, you could call for help.

Right.

I’m not practicing OBL; I’m practicing at the hospital. But after my experience with it now, man, I’d feel more confident doing ketamine in an OBL—

Yeah.

—without having that code back-up.

Right.

Because we saw— in our prospective, randomized study, we saw that over 10% of the fentanyl group had desat below 90% during the case—

Yeah. Yeah.

—where then we treated it with oxygen. There wasn’t any major issue.

Yeah.

However, it was almost never in ketamine.

Yeah. And someone has to pay attention to that.

Yeah.

So there’s— you have to have more eyes and ears in the room to look for that.

Yeah.

Okay. Yeah. I would like to think that if something really bad happened, that I could run a code better than a dentist could.

I guess that’s to be seen.

Yeah [laughs]. I hope not, but what are the requirements for the physician? So—

Oh, in terms of credentials?

ACLS?

Yes. ACLS. And I’ll say we got PALS too, because now we’re doing kids.

Yeah.

So I’m doing— if I’ve got an appendicitis drain, I used to have to get anesthesia, wait however many hours to get them to come down. We actually had to go to a different hospital, because we don’t have pediatric anesthesia—

Wow.

—at our main place.

Yeah.

So now we just got PALS, and the nurses got PALS, and we’re able to do kids now, which has been super helpful. But, yeah, so ACLS— and essentially all the same thing that you do for moderate sedation. They usually have some course about sedation that you have to take, and then the proctoring of those three cases, and then show that you maintain the privilege because you do a certain number a year. So it isn’t very onerous on top of what you’re already doing for moderate sedation. And I would argue that having such a vast experience with moderate positions you perfectly with this drug, which is safer.

Yeah. Yeah.

And I am not advocating for us to go with propofol or something where we’d have to worry about the airway more.

Right.

I think ketamine is a special case.

Right. You did mention that there are some elderly or frail people that you would, for instance, cut back on the amount of Versed you give. And I think there are some patients that will have the opposite reaction to Versed that we want anyway, so that we actually stay away from that class of drugs altogether. Are there other patients that you think are high risk for ketamine that we should know about?

Well, it’s contraindicated if the patient has schizophrenia or is pregnant, so you’re not going to be using them in those populations. I’ve noted patients on dialysis or with really bad kidney function or really bad liver function, they don’t clear the drugs as well. I think all the drugs, but the ketamine— it’ll— they’ll be more sedated longer, but we still use it. They’re just—

They just take longer.

They just take longer to recover. Yeah.

Great.

Have I convinced you?

Well, I would love to try it. I’d have to do some research on my own to figure out who I need to convince. One crazy thing about my hospital is that it does not require the IR attending to be ACLS certified.

Interesting.

Yeah. So I did it on my own for many, many years. I got to check even whether I’m still current on that. It does seem though that in some cases— if it really does prove to be safer, it does seem that in some ways you might be able to concentrate more on your procedure.

I think that’s the main thing or a couple of the main things, is that now we’re spoiled with the ketamine. They don’t talk, they don’t move. Now I’m fairly friendly. I’ll sometimes chat with people if they’re not fully asleep, but sometimes— I just want to focus on my case—

Yeah.

—and they’re talking. So the ketamine— that— essentially eliminates it. The other thing that’s really nice is the procedural recall is basically zero. We studied that in the randomized study. So they remember a lot more about the fentanyl. So especially in a training place when you’re trying to talk shop, it’s awesome. They’re not— they can’t— they don’t hear you.

Yeah.

They don’t remember. And if you accidentally drop an f-bomb, it’s not getting reported.

Yeah. Yeah [laughs].

So whenever there’s—

We would never do that [laughs].

If there’s a— Exactly, just theoretically.

Yeah.

So if there’s a patient that for whatever reason we can’t use the ketamine, it’s extremely annoying to us now. We’re like, “No! It won’t work!” Oh, the other thing— watch out for patients with really bad hypertension, because ketamine will bring up your systolic by 20— 10 to 20 points.

Ah! Okay.

So even— we have— everybody is hypertensive on blood thinners in my state. So even if in there the high 100s, we’ll give some hydralazine ahead—

I see.

—just so that we can use the ketamine. That’s how strongly we want to use it. And then another thing we’re doing now is we’re doing a study about clinician perception of sedation. So this is also an area that we haven’t really studied, because it is interesting, that feeling of how great we feel with the ketamine. I want to capture that. And so I’ve adapted this tool, and we’re collecting data from the nurses and us and seeing what correlates. I’m trying to validate it for IR, basically. So this could be part of our research program. And do our perceptions match how the patients feel?

So IR clinicians, not the referring clinicians.

No, they’re not there. Us in the room.

Okay.

So that’s another aspect, but it—

And that’s the most important thing, because it’s all about me [laughs].

Well, studies show that if you are distracted by the patient moaning and in pain, you do worse—

Yeah.

—at least in the surgical literature, you have more mistakes. You’re more— you’re— so I think it is important to think about our perspective.

Absolutely. Absolutely. And actually one of the things that I’m not proud of and I’m not happy about is that probably the majority of the times that my patient gets in trouble, whether it’s a respiratory or cardiac or whatever issue, I’m the first one to notice. I’m looking at the EKG tracing or listening to the O2 sat, want it to go down—

Boop, boop, boop.

—or something, and I seem to be more sensitive to that than other people in the room, but that detracts from what I’m trying to do from a procedural point of view.

That’s very interesting, because I— in my lab, I decided that the monitor, it just goes directly to the nurse desk, and I don’t see it at all.

Yeah.

I just rely on them fully to report to me when there’s an issue.

Yeah. Well, that’s the way it should be.

Yeah, maybe you should do that.

Well, I do that, but I still want to be in tune with what’s going on with the patient.

Well, certainly, you want to be able to detect if there’s an issue coming.

Yeah. All right. So where can I get some of this stuff? Is it on Hims & Hers or something like that? You call in and—

We’ll have to ask Elon [laughs].

Yeah. Give him your credit card number and then [laughs] a package arrives in the mail.

I mean, the good news is it’s generic. It’s cheap. Every hospital has a bunch of it, and I’m happy to help you in any way—

So it is, it’s such an old drug that it’s all generic.

Yeah. I actually pulled all the costs recently, just to see. So it’s like for a vial of fentanyl was like two bucks, for midazolam a couple bucks, for the ketamine we paid like five or seven dollars a vial. It’s very minimal.

Okay. Which means with the upcharge, at Stanford, it would be I think $8,300 [laughs].

Exactly [laughs]. Yeah. But no, it’s good. And I’m— again, I’m happy to help you try to get— argue to your sedation committee, go picket in front of their office.

Yeah. I think that’s one of the hurdles that we have to get over is that the perception that it requires more physician attention when you’re using ketamine. But what you’re saying is the exact opposite, is that patients have less pain. They’re more dissociative, that there’s more of an analgesic effect during and after the procedure. So it actually saves cognitive space for the performing IR physician.

Yeah. And I’d say one kind of learning point for us is the patient is really asleep. So you have to make sure your techs are very good about securing— we had patients that their arm would fall out of the CT table and get scratched.

Yeah. Yeah.

So they have to— they’re immobilized and totally— So that is how unconscious they are.

Yeah.

And it’s wonderful.

But I suppose that’s part of the required training too, because we don’t want people to wake up with a—

Scratch.

—palsy or like a brachial plexus injury or something like that because they were positioned in a way and were unable to move—

Exactly.

—because they were so asleep.

It is.

So do you still have to put a strap around the patient, around the table?

Oh, yeah. We buckle them in, we secure the arms, just because you don’t want them falling off. Well, I feel like I convinced you.

I’m convinced, but as the former JVI editor, I’m a believer in evidence.

Right.

And you have produced that evidence, it’s unequivocal, and again, if it’s not extra work for me during the procedure, if it’s actually less work for me, that’s the win.

Right.

Right? I want— I’m not good at multitasking, but if I focus on one task, I’m pretty good at that one task, and I don’t want to be the anesthesiologist in addition to being the interventional radiologist—

Right.

—in the room.

Yeah. And I can just say in closing too, another benefit for us is that— the pediatric, it’s extending what we can do without anesthesia. So if you’re in a place where it’s kind of hard to get them, now all of our biliaries, for example, are with ketamine, all of my hepatic ablations, just because I get more throughput through my room and equal outcomes—

Right.

—in terms of pain control.

You’re eliminating MAFAT.

Exactly. Which doesn’t stand for anything.

Yeah. If you don’t know what that is—

Look it up.

—yeah, you’ve got to look it up. We’re not going to disclose that publicly.

And I think ultimately you just go to the committee and you’re like, “Look, here’s the data. Not only do patients all prefer it, and the clinicians prefer it, but it’s safer.” What’s the argument?

Yeah.

There is no counterargument to that.

Yeah.

So.

Okay. So wrapping up here, so what should the listener to this podcast take home as the final message?

Ah, look up my papers, print them out, and go straight to your sedation committee, and give it to them, and get on the ketamine train, because this is the new thing. We haven’t upgraded our sedation practice in 30 years, and we’re going to update it big with a very old drug, I think.

All right. Let’s do it.

Yes.

Okay. Thank you. So thank you to Dr. Amy Deipolyi, associate professor and section chief at the Charleston branch of the University of West Virginia for this fascinating conversation that could have a great impact on all of our patients.

Thank you so much, Dan.