skip to Main Content

Hot off the press is a brand spanking new updated 2023 AGS Beers Criteria for Potentially Inappropriate Medication Use in Older Adults.  The Beers Criteria is one of the most frequently cited reference tools in geriatrics, detailing potentially inappropriate medications to prescribe to older people. 

We’ve invited two members who helped update the criteria including Todd Semla and Mike Steinman.   We discuss a little history of the Beers criteria, including the original Beers Criteria that was published by the late Dr. Mark Beers, and how it has evolved over the last three decades.  We also discuss specifics about how to use and not use the Beers Criteria, how medications are selected for inclusion in the criteria, and specifics about certain medications. 

And of course, take a deep dive by downloading the JAGS paper on the updated Beers Criteria or any of the great links from AGS including the:




Eric: Welcome to the GeriPal podcast. This is Eric Widera.

Alex: This is Alex Smith.

Eric: And Alex, I think we’re having Beers today.

Alex: We’re having Beers. Beers. The Beers. [laughter]

Eric: Who do we have with us today?

Alex: That’s what it was named after, right? We’ll learn about that. Okay. We’re delighted to welcome Mike Steinman, who’s a geriatrician professor of medicine at UCSF in the division of geriatrics, prior guest on this podcast. He’s co-PI of the USD Prescribing Research Network. Welcome back to the GeriPal podcast, Mike.

Mike: Thanks for having me.

Alex: And we’re delighted to welcome Todd Semla, who is a clinical pharmacist and associate professor at the Feinberg School of Medicine at Northwestern University. Welcome to the GeriPal podcast, Todd.

Todd: Thank you. It’s good to be here.

Eric: So we’re going to be talking about, it’s probably a joke that both Mike and Todd are so very tired of when we’re talking about the Beers List. But we’re going to be talking about the Beers List and potentially inappropriate drug prescribing in older adults. But before we do that, who has a song request for Alex?

Alex: I think that’s Mike.

Eric: Mike?

Mike: Both Todd and I were both talking about it and we thought that, Don’t Let Me Be Misunderstood by The Animals would be an apt song for this podcast.

Alex: Yeah, I think Mike seconded it, right?

Todd: No, I think I did. I had some Warren Zevon request.

Alex: Oh, that’s right. Right, right. Yeah. But Todd has this on LP, right?

Todd: Yes, I do.

Alex: Yeah, this is a great choice. I’m a little sick, so see what happens here. Could be exciting.


Eric: All right. Why this song?

Mike: I would say because the Beers criteria can be a really useful resource, but Todd and I and many other folks have observed that they can be misused and misinterpreted. So as we prepare to release the updated version of Beers criteria, we’re really excited to get the sort of new recommendations out there, but we are equally excited to make sure that people use them in the ways they’re intended and minimize use in the ways in which they’re not intended and could cause inadvertent harm.

Eric: So we’re going to talk both of that. But maybe you could just briefly, what is the history of the Beers? Is it Beers criteria, Beers list? How should I reference this to the podcast?

Todd: Well, your earlier comment about us being tired of the jokes about the Beers, we went so far as to have AGS copyright it. So it’s officially the AGS Beers Criteria.

Eric: AGS Beers Criteria. What’s the history behind it?

Todd: So it was started back in 1991 was the first edition was published by a geriatrician named Mark Beers and he was at UCLA. So he got a bunch of his colleagues together and they came up with a list of medications that they thought were problematic and whose risks outweighed the benefit. But they were specifically looking at nursing home population. And then when they updated it in 1997, they expanded it to include all older adults. And then there was a version in 2003 which Mark participated in and he since passed. And then in 2010, the AGS took over because it was sort of wafting out there and people were wondering, when are you going to update it? Who’s going to update it? And we were trying to get funding from different places and then AGS kind of stepped in and so they’ve been the stewards of it since 2010.

Eric: And when was the last update?

Mike: 2019.

Eric: 2019.

So is the plan updated every four years?

Todd: Well, we try every three years. But sometimes-

Eric: Sometimes pandemics happen.

Todd: Well, pandemics and the amount of literature that the panel has to wade through seems to be growing exponential. It doesn’t mean we’re finding any more diamonds, but there’s certainly a lot of coal to look through.

Eric: And Mike, you were talking about not being misunderstood. What is the intention of the Beers criteria, the AGS Beers Criteria?

Mike: Thank you. So the intention of the AGS Beers Criteria is to provide information primarily to clinicians, although to other audience as well, but medications that we consider to be potentially inappropriate in older adults. And by that we mean that in many or most older adults, these drugs have either lack of benefit or a degree of harm relative to benefit that is inferior to readily available alternatives. So it’s not to say that these drugs should never be used. There are some older adults in whom these are the correct drugs, either because of their particular clinical circumstances or because shared decision making and patient values and preferences.

It’s not to say these drugs should never be used. But we also recognize that for a lot of older adults, these drugs can cause more harm than benefit either just by themselves or relative to these older alternatives. So we want to alert clinicians as well as patients and policy makers to the presence of these drugs. And therefore, for most older adults try to steer them towards safer and more effective alternatives while not being going to the extreme where we are overly restricting access to these drugs for people in whom they are appropriate.

Eric: So are there any examples that you’d be willing to share about people or systems using it inappropriately or potentially thinking about using it inappropriately, using inappropriately list inappropriately?

Mike: Potentially inappropriate.

Eric: Potentially. What’s an example of that?

Todd: Well, I think things we’ve heard from, we’ve received letters from over the years is perhaps there’s a PBM or something that’s being overly restrictive or overly interpreting a criteria leaving out one of the caveats we may have on it, or just saying avoid means no and no in all our circumstances and making it extremely onerous to try to make an appeal through that.

Alex: Yeah, so using it say as a quality metric would be another example or using it in an end of life or hospice setting. Would those be examples?

Mike: Yeah. The explicitly mean the criteria explicitly are designed to exclude people in hospice or at the end of life. And it says that very directly that is not the intended audience for this. But that said, this is a situation in which people might not read that caveat and would apply it in that situation, which would be wrong. Or as Todd said, getting wrapped up in endless reams of paperwork, trying to clear prior authorizations that really overly restrict access to drugs on the list even for people in whom they are reasonable choices.

Alex: A and is there, to your knowledge, a Beers criteria or equivalent for people who are at the end nearing the end of life or have a limited life expectancy, say six months or less?

Mike: There are some lists which are out there. There’s nothing that’s really sort of gained traction as a more, I’d say broadly accepted or universal guideline. But people have come up with lists about medications which are potentially inappropriate for people at the end of life. Those include drugs often, drugs which are preventive drugs with a long lifetime to benefit.

Alex: So there’s an opportunity there for the AAPM, some name, list drugs to potentially avoid. Okay. Eric, you have a question?

Eric: Well, I was going to say, I think ePrognosis has a little bit of a time to benefit table that folks can check out.

Alex: On our own website.

Eric: How do you want physicians to use this? We talk about ways not to use it. How do you want me to use this in my daily practice? Either Todd or Mike.

Todd: Well, my short answer is we expect you to memorize it and then eat it.

Eric: ChatGPT does all my work for me now, so just make sure it knows.

Todd: We haven’t signed that contract yet. So I think the best thing to do is to look through it and think about how you do prescribe and look for drugs on there that you maybe are prescribing regularly, that you should think twice about prescribing to those particular patients. Or if you have patients who are on those medications, is it time to reconsider whether there is any benefit or the harms outweigh the benefit? So I think that those are good ways to start to use it.

Mike: And the analogy I like to use is to think of it like a stop sign, where if you’re thinking about prescribing one of these drugs or you have a patient who comes to you and they’re already taking one of these drugs that you kind of take a close look and you say, “Stop. Is this the right drug for this patient?” You kind of go through that mental process talking to the patient, understanding their actual reaction to the drug, looking around, looking left, looking right at your stop sign for potential alternatives. And if the decision at the end of the day, this is the right drug for the patient, then after stopping then you go. But it does kind of provide that caution, it kind of slows things down to make sure you’re really paying attention. That of course, should apply for any drug, not just for the Beers criteria. But because the Beers criteria are higher risk for being drugs that you would want to avoid, it doesn’t merit that extra pause and really thinking about it before you then proceed.

Alex: I have a question for both of you, because both of you’re interested in deprescribing. As you’re reviewing the literature and you’re trying to decide how to revise these lists, I’m sure you’re noticing deficiencies in certain areas. And one of them I would guess is that there are many more studies, trials of initiating drugs than there are deprescribing drugs. Is that your sense as well? And then what do you do when, like you just said, Eric should think about using these lists when a patient is either you’re considering starting it or they’re on the drug already and you’re considering stopping it. Any thoughts about the level of evidence for deprescribing on somebody who’s already on the drug?

Mike: I can start. So feel free to chime in, of course.

You’re right. There’s not a whole lot of data about deprescribing. The vast majority of evidence comes from either randomized control trials where you have a group who starts it and a group who doesn’t, or observational studies in which you find an exception cohort, people who start the drug and you compare them to a group of people who didn’t, then follow them forward in time.

So the data on deprescribing is limited. And interestingly, if you look closely at the criteria, we encourage people to look closely because there are important nuances in there. There are some criteria in which that criteria says avoid initiating the drug. And the reason why is because there are, to the extent to which there are deprescribing studies out there, some of them randomized, but many more observational or others, there is some evidence that there can be a differential effect between starting the drug and stopping it.

And it’s not 100% clear why, and because these studies are imperfect, it’s unclear if that’s a true difference or it’s just sort of an artifact of research randomness. But there can be some differences between starting and stopping, particularly if someone’s been taking a drug for a long time and they really seem to be tolerating it well, then that might provide a different risk benefit calculus than someone who you’re thinking about newly starting the drug and then they’re exposed to all of that kind of upfront risk.

So most of the criteria of the drug just say, avoid this drug for most patients and don’t make a distinction between continuing versus starting. But there are some criteria in which we do make that distinction where we specify avoid initiation and then maybe consider deprescribing for the patient already on it.

Alex: Yeah. There are some drugs for which when you stop the drug, if you just stop it without tapering, there are potentially harmful effects. You don’t go back to a place, it’s not equivalent to the person who hasn’t yet started it.

Mike: Yeah, that’s exactly right. That’s even independent. That’s even assuming you do an appropriate taper of a drug that does need a taper. So you could definitely can do direct harm for many of these drugs if you stop them abruptly. Benzodiazepines are one to some of the best known drugs on the Beers criteria, for good reason, they’re overused and they cause all sorts of harms. But if a patient’s been taking a benzodiazepine every day for the last 10 years and then you tell them tomorrow to stop it, they’re going to be in a world of pain and that’s not going to be good for anyone. So clearly being attentive to how to de-prescribe these drugs safely is important.

Alex: Great.

Todd: And I think there’s two really good examples because this is the first update where we really talk about deprescribing and brought it into the Beers criteria in our discussion and cited some resources for people. But the two examples are the use of aspirin as a preventative measure where it perceivably had been used with caution because of lack of evidence. And then the US Preventative Health Services Task Force came out with their recommendation. So we moved it over to an avoid, and recommended potentially deprescribing it in patients for whom there is no perceived benefit and potentially harms.

And the other example was in postmenopausal estrogen use, specifically the transdermal and oral formulations where it always had been an avoid in women over 65. And now the recommendation is to avoid new starts in women over 65 and women who’ve been continued on it to consider a deprescribing decision based on shared decision making and those types of things.

Eric: And in the article that’s being published in JAGS, you have multiple tables. First table is potentially inappropriate drugs in older adults. It gives the drug, a rationale and a recommendation with quality of evidence and strength of recommendation. Is the way I’m supposed to be using this table is I see the drug, I look at the rationale and then let’s say for Megace, it’s just avoid in older adults because of the rationale, which is doesn’t really work and it has a lot of side effects? Regular insulin, sliding scale, same thing, it’s just an avoid. Medications like nitrofurantoin, there’s some risk, but the rationale isn’t just avoid, I think it was avoid with a creatinine clearance of less than 30.

Todd: 30. Yeah.

Eric: Is that how I should be using this table?

Todd: I think that’s correct, the way you would use it. Yeah.

Eric: Okay.

Todd: You really want to avoid nitrofurantoin, for example, in patients less than 30. Because we really don’t have any evidence that it works in patients who have creatinine clearance below 30.

Alex: Did you say Megace?

Eric: Yeah. Megace? You questioning my-

Mike: Generic names only.

Eric: Megestrol.

Alex: Sorry, couldn’t resist.

Eric: What are some of the other big-

Mike: Does make the important point that some criteria just avoid the drug. And as we’re talking before, it doesn’t mean you should avoid it in 100% of people, you should avoid it in most people. But there might be some exceptions. But these other situations like the one we’re just talking about, nitrofurantoin, it’s a great drug. People with good creatinine clearance, it’s a great drug for treatment of UTIs, and we should be using it. But in that case, that criteria warns people that if your creatine is less than 30, or if you’re thinking about using it for long-term suppression, those situations in which A, it may not work and B, may cause disproportionate harm relative to alternatives. So in those clinical situations, you should consider avoiding that drug. But it doesn’t mean avoid it in all situations. In many situations it might be a great choice.

Eric: So NSAIDs is a good example of this. I feel like geriatricians have a, maybe it’s more of a hate/hate relationship with NSAIDs right now. What’s the recommendation for NSAIDs in that table? Well, I guess I got it right here. Avoid in chronic use, and avoiding short term use if they’re taking other things like corticosteroids, anticoagulants, antiplatelets. But it’s not a complete avoid.

Todd: No. The big harm that we’re concerned about is GI bleeding. Because that seems to be what older adults seem to be most prone to. So there’s mechanisms around that. You’re giving someone your proton pump inhibitor for coverage or an H2 blocker, if that’s the best alternative. Avoid using it with other drugs that are also going to increase GI bleeding, like corticosteroids, for example. And then they also come up in the renal dosing table to avoid in patients who have creatinine clearances below 30.

Eric: Okay.

Mike: Yeah, exactly. And what the criteria does not cover is if someone, older adult, exerts themselves disproportionately once a week and then is a lot of pain, if you want to take a single dose of an NSAID once a week, that’s probably reasonable. That’s not what we’re saying. The criteria is really referring to chronic daily use because those are the people at really as highest risk of GI bleed.

Eric: Any other big updates for that table for this time around?

Todd: Well, can you read my mind, Mike?

Mike: Go ahead.

Todd: Well, so I think probably the biggest update is what we did with anticoagulation with the DOACs and warfarin. So warfarin has moved into the avoid category unless somebody is, you have no other alternatives, but specifically for non-valvular afib and VTEs, things like where you have all other alternatives.

Eric: Why did warfarin move into that table?

Todd: I’ll let Mike take that.

Mike: Sure. So we’ve been using Warfarin for such a long time, but imagine warfarin was just introduced today and that the DOACs were introduced today. And you looked at those two groups of drugs side by side, and you said, which one is better for most people with non-valvular Afib? It would be unambiguous that DOACs are better in terms of they have equal or to superior in terms of preventing the outcomes we care about, things like for VTE would be pulmonary embolism or AFib to be for preventing stroke. And they carry a lower bleeding risk, particularly for intracranial hemorrhage. Some of the DOACs actually can confer a higher GI bleeding risk than warfarin. But globally, the risk of really serious bleeding is lower with the DOACs, and you get similar to better efficacy, and you don’t have all the monitoring stuff to deal with.

So on one hand this is a slam dunk and this is what consistent with all the major anticoagulation guidelines say. The caveat that we put in there is two things. First of all, the recommendation that we have is to avoid initiating warfarin for initial therapy, which is distinction from someone who’s already been taking it and they come to you on warfarin and say, “Should I change?” And that’s because there’s data to suggest, and it’s imperfect data, but there’s data suggests people who have been on warfarin for a long time, they’re tolerating it well. They have a good amount of time in the therapeutic range, at least 70%, that those people don’t necessarily get a whole lot of benefit to switching over to DOACs. It’s still unclear. It’s not a bad thing to switch to a DOAC, but they don’t seem to get a whole lot of benefit in terms of the harm calculus.

But if you’re just a new patient comes with a new onset Afib, but they’re newly considering anticoagulation, then clearly the harm benefit skews towards preferring DOACs. So we make that distinction for this drug. The other thing that we put in here, and we were very mindful that there’s, for example, these direct acting oral anticoagulants, they are hella expensive, especially if they’re not a formulary. And the last thing we want to do is say, you can’t take warfarin anymore and I have to pay $500 per month for your DOAC if it’s not covered by a formulary.

So we are very mindful of not putting people in causing that sort of harm, or patients who really want to stay on warfarin and they’re very attached to it. Ruining the patient doctor relationship because you get into a fight about what anti-coagulants should be on. That’s not what we’re aiming for. So we have language in this criteria that really tries to call out these challenges and decision making, recognizes that for some patients, warfarin might end up being the right choice, but we’re really trying to encourage people to not rely on warfarin as first line therapy because of the disproportionate harms relative to the alternatives.

Eric: So I guess it was table two in the JAGS article is about potentially inappropriate drugs. Table three is about potentially inappropriate meds in older adults due to drug disease or drug syndrome interactions. If they have heart failure, avoid things like NSAIDs. One thing that I thought was interesting because I was trying to go through this list of things that I do, under delirium, they had things like anticholinergics, benzos, steroids, which all seem reasonable. I saw at the end it had opioids, which is interesting because when I think about delirium, I think certainly opioids can cause delirium, but pain can cause delirium too. How should I think about that with the AGS Beers Criteria?

Todd: Well, as we say in the rationale, the emerging data highlight an association between opioid administration and delirium. And then we go on to point out that conundrum where you’ve got someone with pain and they may have delirium, are you going to worsen it by adding an opioid? And I think the point here is to recognize that opioids can lead to delirium. You have to take that into consideration clinically. If the delirium isn’t clearing and you’re treating the person’s pain, is it because of the opioid or is it some other underlying factor that hasn’t been addressed?

Eric: Yeah.

Mike: Eric, can I bring us back to table two for a sec? Because I think there’s just a few things I think that we are worth highlighting in terms of new changes. So the biggest changes that Todd started to explain were warfarin, getting it on the list. But another drug that’s now on our main list of drugs for avoid is rivaroxaban, the treatment venous thromboembolism and atrial fib. This is a big change because this is a very commonly used drug and it’s one of the most commonly used DOACs. If we’re saying don’t use warfarin. And when I say don’t put that in big quotes around that. Don’t use rivaroxaban. That we recognize that does potentially limit choices. And so this was a controversial, and controversial is the wrong word, but this merited a lot of discussion in the panel that way in thinking through what the implications were.

And again, as we said with warfarin, the point is not to restrict all uses of rivaroxaban. For people who can only take a day drug, that’s the main go-to, and that’s a very reasonable choice, or for people whom that’s the only drug on their formula, now they have to pay this huge amount of money out of pocket for something else. These are part of clinical decision making. And there’s no direct head-to-head trials between the different DOACs.

So we don’t have direct randomized controlled trial data to say one is better than the other. But there is a preponderance of observational evidence to suggest that the risk of GI bleeding is substantially higher on rivaroxaban compared with DOACs, with the most common comparator showing that difference being a apixaban.

And so there’s this push-pull about not wanting people to be overly restrictive in therapy and get people into sticky situations. But at the same time, we do want to call out the fact that if we’re thinking about the purpose of the criteria, which is can we identify drugs in which there is a safer and or more effective alternative? The preponderance or evidence right now shows that if you’re thinking about DOACs, rivaroxaban is probably not your safest choice.

Alex: You’ve got a box calling this out, which is terrific because I could see where this could lead to confusion. So just to be clear, warfarin is in table two, potentially avoid. Rivaroxaban is also in table two, potentially avoid. And dabigatran is in table four, I believe, used with caution. So then it kind of raises a question, how much did the makers of a apixaban pay for the AGS Beers criteria at this route?

Mike: Only half. [laughter]

Todd: Not a penny.

Alex: They did not pay, right? No?

Mike: They did not pay.

Alex: Want to be clear, that’s a joke.

Todd: There is a firewall between the panel and the pharmaceutical industry.

Mike: Yeah.

Alex: My point is it kind of points you towards a apixaban.

Eric: Well, I guess the question is about pointing. Let’s say insomnia meds, sleep medication. So under table two, avoid includes, which is probably the worst name of any named kind of categories of drug, non-benzodiazepine benzodiazepine receptor agonist hypnotics, Z drugs. These are the zolpidems of the world, the Ambiens. And in that, it says to avoid because they basically have similar side effects as benzodiazepines. So does this mean that we should be using the other because the other ones are safer and more effective? It orexins and the melatonin receptor agonists?

Todd: Well, we haven’t come out with a statement to that effect. There’s always questions, are we going to update the alternatives list that we did for the 2015? That’s under discussion. But when we look at those drugs, I think one of the things we have to understand is when you look at the lag time for falling asleep and the total sleep time improvements, it’s pretty negligible. So while they may not convey the same potential harms, one has to consider whether they actually have efficacy that’s worth the money spent or any other risks.

Mike: And the other thing is one of the goals of the criteria is not just to say don’t use drug A, so instead you should use drug B. It’s like we’re basically saying be cautious about using drug A and think about other pharmacologic and non-pharmacologic alternatives that are going to get you to goal. So in the case of helping people with sleep, we know from studies that the best thing to help people with sleep better than any drug is something like cognitive behavioral therapy for insomnia. Safer, works better. And so, hard to get, don’t want to downplay the limitations on access to therapy, cost and coverage issues. They’re real, they’re there. But the point is not you should just be subbing out Beers drugs for other drugs as you should be considering reducing use of those drugs and thinking, what else can I do to help the patient, which might often include non-pharmacologic alternatives.

Eric: How do you decide which drugs to include on this list? Is there a lot of debate? What does that process look like?

Todd: It’s a real sausage making process. And I think it’s the most difficult part of the whole process is our literature search. Because you really can’t do an online search and say, okay, what drugs cause harm in older people, and come up with a new literature. So we cast a very broad net in terms of search terms. And as a result, we get back tens of thousands of articles that have to be gone through either by title or abstract to throw out the things that are of no interest to us or that are redundant. The review articles, the case studies, opinion pieces, things like that. So we’re trying to find articles that maybe support.

And we do search the existing criteria so we can look for new drugs for those to whether support what we’ve said or are going to change what we said. And then we’re also looking for new drugs. And that’s where the panel comes into play because they will suggest topics, new drugs that we need to look at. They may provide articles that they’ve seen in the interim that we want to follow up on and see if there’s any more data out there, other studies that would show the same thing.

Eric: A good example of a new drug I saw Dextromethorphan and Quinidine. Those are decades old, but the combination in a very expensive pill becomes a new drug. How did that one get on there?

Mike: That one is on there in terms of drugs to avoid in people with heart failure because there is a explicit warning on the label that we shouldn’t use it for people with heart failure. It is not on there otherwise. And that’s something that we discussed. It was sort of passed on for this criteria. It may sort of come on subsequently. I can’t say, it may or may not. A lot of other drugs may or may not.

But what we do encourage people, so we use our panel expertise to say, “Hey, are there new drugs which are out there, all drugs which are now have increasing evidence of harm or lack of benefit that we should think about adding to the criteria.” And so we just got to kind of crowdsource the panel and we encourage people to send in, for lack of a better word, nominations, like drugs we should look at, drugs you think maybe should be on the criteria, maybe worth consideration. And we encourage that because we love to get those ideas. So if you and of your listeners have things that you want us to think about, please send it our way. Can’t guarantee what the outcome’s going to be, but we’d be happy to dig in a little bit.

Todd: And the combination, the dextromethorphan quinidine combination has been in the criteria in the previous edition. And it still is under use with caution.

Mike: I’m sorry, I misspoke. Sorry.

Todd: No, no, that’s fine. But that was I think what Eric was going for. But we talk about not using it in behavioral symptoms with dementia, but it’s okay for the labeled use, we’re not going to challenge that. But those off-label uses where the risk of harms, particularly things like falls, were far greater than any benefit that were shown.

Mike: And it’s one in those in the use with caution table, table four. So it’s not a straight up like a void, it’s more just be careful.

Eric: Okay. I got another question. Should I use the AGS Beers criteria or the STOPP guidelines or some other? What’s the benefit of this over some others, STOPP in particular? S-T-O-P-P.

Mike: Yeah, they’re both good. They’re slightly different. They have converged over the years. So basically the STOPP criteria are a Beers like criteria that developed out of a group in Europe. They have their own process. They’re thoughtfully constructed and there’s a lot of overlap between the two. So if you look at the Beers and you look at the STOPP and you think of a Venn diagram, the circles are largely overlapping. There are some different.

So historically, the STOPP criteria have been, in some ways, less operationalizable for… Very useful for individual clinicians, but maybe less operationalizable for quality improvement because it includes stuff like, I can’t remember if this is the exact phrasing, don’t use a drug without an indication, source’s not indicated. Or be careful in this certain situation that’s sort of got complex decision making around it.

And those things are true and absolutely useful and important to think about clinically. It becomes a little harder to operationalize that in terms of a more explicit list. I’d say over time, the Beers criteria and the STOPP criteria, if anything, come together a little bit. Over time, the BEERS criteria has added more clinical nuance in part, because we observed some of the misuse of the criteria. We are really trying to highlight that clinical decision making is required and there are exceptions to these avoid statements and we just want to highlight that. And so they’re moving together. But they’re complementary. We encourage people to look at the STOPP criteria as well. And there’s a lot of good stuff in there too. But like I said, a lot of content is overlapping.

Todd: What Mike was trying to say is that the STOPP/START criteria more implicit than the Beers criteria, AGS Beers criteria, are more explicit.

Mike: Right.

Eric: Well that’s interesting because STOPP is paired with START, I think it’s 20 something meds to consider starting. Is that ever something that you think about for the AGS Beers criteria?

Todd: Kind of goes back to the alternative.

Eric: Overuse but potentially underuse.

Mike: I already have lots of gray hairs, Eric. [laughter]

Todd: I think we’d be looking at an update every six years.

Alex: This is a huge amount of work. I hear you. And thank you for this tremendous amount of work for you and the AGS, which I understand was involved in helping with the literature search. So that is certainly to be acknowledged. But I understand where Eric’s going with this. There are limitations to a guide that is primarily around medications to potentially avoid rather than what to do as the alternative. And it sounds like there was this alternatives list from 2015, and you’re considering maybe. Maybe there’s an opportunity for somebody to take on beefing that up to make a start equivalent for drugs that are commonly used in the US and not in Europe, for example.

Mike: Exactly. If someone has a bunch of money burning your hole in the pocket and you would love to help to subsidize that sort of effort. [laughter]

Todd: Yeah. I was one of the authors on that paper, and tried to do an update in 2019, but for various reasons it fell through. But one of the harder parts of the alternatives list is actually finding evidence to support your alternatives. We have sometimes a desert in what we’re looking for in the AGS Beers criteria, but finding alternative recommendations and evidence to support them also can be pretty barren. And you end up looking at the Scottish Medical Society for their guidelines and the treatment of insomnia, for somebody else who will support the steps for sleep hygiene or something like that.

Eric: Coming close to my last question is there any evidence that the Beers criteria changes any outcomes? I know these medications are associated with bad things, but is there any evidence or is anybody looking at whether they’re not actually using the Beers criteria changes any clinical outcomes or is that possible?

Mike: So the observational studies looking at people taking Beers criteria and those who don’t, and looking at their outcomes are really hard to interpret because there’s so much potential there. Probably the closest thing to get at the question you’re asking is say you do an educational intervention where you teach a bunch of people about the Beers criteria and see if their patients do better than people who didn’t get the educational intervention or something integrated into electronic medical record that sort of prompts people when a Beers criteria drug is being flagged. I say there’s mixed evidence. It’s hardly overwhelming that this is the best thing since sliced bread, in terms of those outcomes. Some studies show no benefit at all, some studies show some benefit. It’s a little hard to, I think come to an inclusive recommendation.

That said, but that kind of intervention, a lot of people don’t end up stopping the meds. If you get an electronic alert, how many times clinicians will just cancel out that alert and keep going. So the real effect, if you could really do a randomized trial like that was sufficiently powered and methodologically rigorous, that hasn’t really been done to really see if you really are vigorous stopping these meds. There’s good data for de-prescribing more in the context of poly pharmacies or studies on nursing homes is where the best data is. If you go through and you really try to de-prescribe, the focus on Beers criteria meds, but not only Beers criteria meds, there actually does appear to be a benefit for mortality and falls in meta-analyses. But you should take it with a grain of salt because the studies have limitations. So there’s a suggestion of benefit, but it ain’t rock solid.

Todd: There are some studies that’ll show decreased readmission if they’re stopped in hospital. But there’s also studies that show they don’t because people end up being put back on the medications when they get back to their primary care or wherever they’re getting cared for out of the hospital within a certain period of time. So it’s like Mike said, it’s a mixed bag and the ideal study hasn’t bent on yet.

Alex: I have two more questions before Eric gets to his magic wand. First is we had a podcast recently about gabapentin, and I know we’d mentioned Mike’s prior study in Annals of Internal Medicine on off-label use of gabapentin. Where does gabapentin fall out on the Beers criteria?

Todd: Well, I’m trying to remember where it is.

Eric: I think it was under the renal.

Todd: Okay. Yeah, that’s what I thought it might be.

Mike: It’s in renal and it’s in drug-drug interactions.

Basically says don’t so gabapentinoids, so gabapentin or pregabalin with opioids, unless you’re sort of cross titrating because there’s increased risk of really bad things like respiratory suppression and death among people who get both. And then we also say that you should avoid the use, if possible, of three or more CNS active medications. So basically try to avoid CNS active polypharmacy. And gabapentinoids are one of the long list of CNS active drugs, also includes antidepressants, benzodiazepine, skeletal muscle relaxants, yada, yada, yada.

Alex: I guess my point though is medication landscape shifts rapidly, and in response, partially we think, in response to the opioid epidemic and concern about prescribing opioids and maybe in search of other medications to prescribe for people with dementia and behavioral disturbances, there’s been a rise in prescribing of gabapentin. And I’m wondering if there’s a way that the AGS Beers criteria could be nimble to respond to potentially inappropriate shifts, dramatic shifts in some cases, with gabapentin in prescribing patterns?

Eric: Or another good example is like for individuals with dementia. We’re seeing this shift from antipsychotics and opioid use and we’re seeing a lot more of the antiseizure meds used in them.

Todd: Have to have us look at it from a different perspective, I think in terms of, again, looking at the efficacy side of it more than the harms because we’ve already got them in there for harms, at least gabapentin and then the other anti-epileptic in the drug-drug interaction section. So we would have to take a look at the evidence and decide how the efficacy weighs with harms in terms whether they would go in Beers, as Beers is currently structured, or if they would go into use with caution kind of schedule.

Mike: I believe that’s something we didn’t really explicitly consider for this update. But that’s exactly the kind of suggestion that we welcome so that we can put it on the list to think about for subsequent updates. And again, it does reinforce the point that one of the things that isn’t there about basically don’t use antipsychotics for the management of behavioral symptoms of dementia unless other stuff has failed, or non-pharmacologic intervention stuff failed and the patient’s threat to self for others. So that comports with a lot of other guidelines. What it doesn’t mean is, okay, stop using antipsychotics and go ahead and go crazy with gabapentin … [laughter]

Eric: EPA for everybody.

Alex: We don’t want that to be misunderstood. But wait, before we get to the song, here’s another. My final question is about, this again, about shifting and rapidly changing landscape. What about new drugs that come on the market? Aducanumab, lecanemab, how do you consider those? Do they appear at the nth hour, right as you’re about to publish, and what do you do in those cases? Thoughts about that?

Todd: Well, we had, early on, or maybe in the middle of our process here, we had a discussion about Aducanumab and what to do with it. But I think that problem sort of resolved itself in terms of it’s kind of going away, but with the second coming and the third-

Eric: Everybody thinks Aducanumab is going away and it keeps on rising from the grave. We’ll see about that.

Todd: Well, anyway. That’s food for thought for three or four years from now.

Eric: All right. My last question is, where can I find the Beers criteria?

Todd: Well, it’ll be published in JAGS, and it’ll be available on the AGS website. And there’ll be materials both for clinicians as well as for patients for the public through the Health on Aging site as well, with a couple of different documents on that. And then there’ll be a pocket card that people can download.

Eric: Great. And we’ll have links to those on our show notes.

Mike: And an app too.

Eric: And an app. There’s an app for that.

Alex: Great.

Eric: Well, Alex, you want to give us more of the… What’s the title again?

Alex: Don’t Let Me Be Misunderstood, by The Animals. Here we go.


Eric: Todd and Mike, thanks for joining us for the GeriPal podcast.

Todd: Thanks for having us. Thank you.

Mike: Thank you.

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

Back To Top