Prescribing Lessons (4.17.2026) Save
Transcription
It's the 17th of April, 2026. This is the RheumNow podcast. Hi, I'm Jack Cush, executive editor with RheumNow.com. This week, we're going to discuss lessons in prescribing. It seems like to be a good rheumatologist, you seem to have to always sort of hone your craft, fine-tune how you go about things. If you're doing it the way you used to do it 10 years ago, maybe you need to change because rheumatology seems to always throw us new information, new curves, new updates, if not, you know, through guidelines, through the reports like we'll cover today.
So, I think one of the interesting things that we put up this week was this report about the use of artificial intelligence scribes or ambient dictation that could occur and is often now included in a lot of EHR packages — not ER, EHR packages — where you don't have to sit at the keyboard while you're talking to the patient and hammer things out or write your notes afterwards. You can just turn on the ambient dictation and the EHR will take notes for you that you have to review later on and whatnot.
So, this was sort of a systematic review in a medical center and looked at how impactful AI scribes were on reducing documentation and if it could do that, wouldn't it therefore be good to lessen clinician burnout? Because clinician burnout — the number one thing we point to is the administrative work that goes into that visit that you do over and over and over again. And what this showed was that while ambient dictation may seem to reduce burnout, the effect on clinician workflows were pretty small when you looked at the overall system. However, if you looked at those who used the EHR and ambient dictation — it was primarily being used by primary care doctors, by women more so than men, and by those who used it in more than half of their patient encounters — they're the ones who got the big benefit. The rest of you who are dabbling are shrugging and rolling your eyeballs and saying it didn't really do me. But if you're one of the people that use a lot of it, they experience twice the reduction in EHR total time and three times the reduction in documentation time. But yet the number of people who were adopting that level of technology is less than one-third.
So I think it's time to get on board and use these new tools. You know, I prided myself in being the best at writing an electronic note in front of the patient using all kinds of tricks and tools, and I wrote really good detailed notes. And what did it get me? Nothing. It got me patients who thought I wasn't paying attention. It got me one-and-a-half-page notes that no one read. I think you should minimize your time with documentation. Meet the rules as dictated by your job, but spend more time talking to patients — and if you have a tool like ambient dictation, use it, because it's going to give you more face-to-face time with patients.
A few things on pain showed up this week. The NHIS survey, which goes on periodically in the United States — this one looked at chronic pain in almost 200,000 Americans and showed those people who had chronic pain versus those who didn't had double the rates of smoking. Now, is that not a double whammy? It's not bad enough you got chronic pain, but now you're smoking on top of that. And they run together. Smoking, while it has decreased over time in the US and other cultures, always had a more significant association in the individuals who had chronic pain. E-cigarettes, however, have increased over time, but yes, they too have had an association with chronic pain. Bad news.
You know, another disheartening thing about pain and use of pain meds is that yeah, you're not supposed to use narcotics. That's like a no-no and it's hard to get anyone who will use narcotics these days. So we're using a lot of adjunctive therapies and that includes the gabapentin-like medicines. A worldwide study of gabapentin use from 2008 to 2018 showed that a) it's used worldwide, and b) it's going up worldwide as far as incremental use year-to-year. It's gone from roughly four doses per day to 18 doses per day in 10,000 population. The highest gabapentin use was amongst higher income countries, so much so that it was six-fold higher compared to low- and medium-income countries. And again, the number of pill equivalents was like 40 per day versus six per day in high- versus low-income countries. But then there's all this data that says gabapentin is dangerous and associated with bad outcomes, and doesn't work all that well. What are we going to do about pain?
I do use gabapentin, and I do think most people use it at wimpy doses — 100 milligrams once or twice a day, you might as well just flick that right out the window. That's like totally useless. But you know, how high will you go and do you believe all the scary literature on gabapentin? Hard to know what to do.
Uh, another study
looked at how good we are at doing the things we're supposed to do when we start to prescribe medium to high-dose steroids. So this is a UK CPRD — it's a Practice Research Data Link — looked at 40,000 people with polymyalgia rheumatica who were started on steroids, and they looked at how many were co-prescribed bisphosphonates, because once you go on steroids with PMR you're going to be on it for four years on average, right? Not the one year that you tell everyone.
And how many were co-prescribed either an H2 receptor antagonist or a PPI? The numbers were 67% received a co-prescribed bisphosphonate and almost 80% received a PPI. This was less likely to happen in males and those who were living in deprived areas. And if you didn't get co-prescriptions of bisphosphonates, there were more fractures in the first 12 months — the numbers low: 2.3% versus 1.4%. It was more. But this is like the data where people show up with a new fracture and they don't get a DEXA scan. What's the deal with that? Or if they do have it, they're not treated for osteopenia or osteoporosis and they're not started on calcium and vitamin D. There are these little things that probably need to be reflexively done when steroids are used. And I think most of us are good at this, but it wouldn't be bad to check your checklist to see if that's on there. By the way, it didn't seem to make any difference as far as GI outcomes for those who are prescribed PPIs or H2 blockers.
I saw an interesting article this week on maybe a new mechanism about how methotrexate works. You know, it's kind of voodoo and we need Bruce Cronstein and other brilliant rheumatologists to really go through the biology of it and talk about adenosine and how adenosine is anti-inflammatory — sorry, is inflammatory — and how methotrexate counteracts that. But anyway, this new mechanism looks at methotrexate effects being mediated through monocytes, and you know monocytes and monocyte counts have been linked to efficacy of the drugs that we use in RA, and maybe that is related to methotrexate. We do know that methotrexate induces — results in — polyglutamation, and that is seen not just in RBCs but also in monocytes, that peaks at day five post-methotrexate. But to get a more substantial transcriptional effect on the genes that might drive inflammation, again mediated through monocyte-associated genes, peaks at around week four, and these transcriptional changes do seem to predict methotrexate responses. I like this new postulate. We'll have to see if it pans out in the hands of other researchers.
Leeds, known for their great studies on early RA and early interventions in pre-clinical RA, published a study about 63 patients with palindromic rheumatism who progressed to develop either persistent inflammatory arthritis or met criteria for rheumatoid arthritis. These people who did progress — from P and I, again, I've said before, PR is just another version of clinically suspect arthralgia, pre-clinical RA; it's just a more defined presentation — but these 63 who progressed from palindromic to RA-like disease did so at a mean/median of 13 months. And the point of the paper was the joints that you get swelling and inflammation in, mostly in the hands and fingers, are the same ones that you're going to get when RA manifests as chronic symmetric polyarthritis. So it's kind of like a joint-priming kind of event. And it turns out that if you were the person who developed RA but had a palindromic onset, you're going to have significantly more swollen joints at the time of diagnosis compared to others. Now, is that a real thing or is that like a selection bias? Like, once you're labeled as palindromic, you need to have more persistent swollen joints to be labeled as RA. Hard to say. But I like this research coming out of Dr. Emery's research group.
New England Journal article this week — there were two. One on Kawasaki's disease, saying that there's no role for prednisolone in the early management of Kawasaki's disease over standard-of-care therapies. And that's been said before. This was a study out of China. And then there was a study on ITP and the use of ianalumab. That's the BAFF inhibitor — it's a B cell depleter. It was a randomized controlled trial, placebo-controlled, double-blind, 152 patients with ITP. Everyone receives the oral thrombopoietin receptor agonist, which is a standard of care — it's called eltrombopag — it is a standard of care for ITP, and then they were either given placebo or 3 or 9 milligrams of ianalumab. They looked at outcomes at 4 months and 6 months, and better outcomes were seen with ianalumab, especially at the 9 milligram dose, with the time to failure being significantly longer with ianalumab. And the failure was like getting a platelet count less than 30,000. And when
they looked at six-month outcomes and six-month response rates were better again with that 9 mg dose versus placebo 62 versus 39%. I don't know about you, but I've always recommended ITP patients be treated with rituximab. I think B-cell-depleting therapies like ublituximab are the way to go as part of the regimen. I think it's better — in my opinion it's better at actually inducing remission so that those patients won't need chronic long-term therapy and anticoagulant sort of therapy.
I like this report on deprescribing of PPIs. The article went on to say that there's a lot of inappropriate prolonged use of PPIs. And how do you get people off of PPIs when they don't need to be on it? You could do that with usual care — instructions just for the patient, instructions for the GP and the patient in a GP-and-patient-facing intervention. And it turns out that the GP-patient-facing intervention was twice as successful as usual care or just educating the GP only about you should do this. And why do I put this up and why am I going on about PPI deprescribing? We really aren't great at deprescribing, are we? Especially when it comes to steroids — meaning you put people on steroids because they're doing lousy and then you put them on advanced therapies and then they get better. How good are you at withdrawing those therapies? Most of us aren't very good. And that is not just for RA. That's been seen for PsA and for spondylitis. So deprescribing should be put in the guidelines, as it was for the JIA guidelines that were published at the last ACR — they made that a big component.
I think this is a Lancet global burden of disease study that looked at gout and its incidence — that it has increased by 158% from 1990 to 2023, and that most of that increase was related to aging of the population. There has been some talk about gout seeming to be occurring younger. Well, this paper says it's more from an aging population and not from younger onset of gout. They didn't go into the influence of obesity, but obesity is affecting the young and the old, and aging seems to be the big factor. So while the mean onset age did increase from 58 to 62 years, the percentage of young-onset gout — below 50 I think the number was — dropped from 16% to 9.8%. Figure that out.
TRACE was a study of what happens when you escalate secukinumab in patients with active axial spondyloarthritis. The TRACE study was a 90-patient active SpA, bio-naive study where they were randomized to receive 150 milligrams of secukinumab, and then at week 16, if you were in ASAS remission you stayed on that dose, but if you weren't you increased the secukinumab to 300 milligrams. By week 24 there wasn't much increase in efficacy. So escalating secukinumab really didn't do a lot, but for the 55 or so — one in five of them did achieve an ASAS remission. So it had a minor effect but it wasn't significant as far as clinical outcomes or as far as MRI inflammation as measured by SPARCC scores. Most of the MRI changes in this one-year study occurred in the first four weeks and not much after that.
This is reminiscent of the ESCALATE study published I think last year that looked at whether the use of treat-to-target dose escalation with secukinumab in SpA was going to be effective, and it didn't really show that it was any more effective than usual care with other drugs. So will you do better if you escalate from 150 to 300 in SpA? I would do it. I have done it. I think though that these two studies are questioning its validity and whether that's going to be good enough to pay for it.
There was a study this week — novel in my view in that it's something I don't do — and that is plasma exchange in patients with idiopathic inflammatory myositis. Systematic review, 35 studies, 473 myositis patients. 361 of them were treated with plasma exchange. There's got to be a publication bias here — like who else would do this and report it? And many of the reports did suggest improvement, but it was kind of weak. And the seven studies that were randomized and comparative showed no mortality benefit, especially with regard to severe ILD outcomes and ILD lung mortality.
I don't use and have not had to use plasma exchange in inflammatory myositis, and I've treated patients just as bad as you. But I think the problem here is that there's no standard playbook when it comes to polymyositis, dermatomyositis, and — I don't know — myositis. I hope it's not inclusion body myositis. We all use steroids to start and they're dangerous, really dangerous in myositis in my opinion. More dangerous in myositis than lupus, my opinion. I've had more osteonecrosis in myositis. I've had more severe, deadly infections in myositis patients.
all occurring in the first 6 months by the way. Um so you know we all use steroids and then we jump in with what drugs that have not been proven to work and are not FDA approved because there are no clinical trials. I mean the only drug that's that's FDA approved is IVIG and to me that's a third line drug. My first line is first line DMARD is um is methotrexate followed by leflunomide and then azathioprine and then I've used a whole bunch of other things including calcineurin inhibitors and cytotoxics and whatnot but I've not used plasma exchange and I wouldn't based on this report.
Two more reports. Another one uh coming out of the uh American Academy of Dermatology meeting that just completed — a study uh called the AMETHYST study was presented by rheumatologist/dermatologist uh Dr. Joseph Moro of UT Southwestern. He gave us a little video that you might want to look at where he describes what happens when they use this novel monoclonal antibody that targets BDCA2. Uh it's a marker that you find on plasmacytoid dendritic cells. Um there was a previous phase 2 trial and this is a I think another phase 2 trial the AMETHYST study but this one's being done in cutaneous lupus who that was active despite prior therapy with antimalarials. To get into the study, you had to have active either CLE or chronic CLE, most of those being DLE. These were not patients with acute LE. And you either received placebo or um litifilimab and um everybody received uh um background standard of care that remained stable. The outcomes at week 24 um was an investigator global assessment of CLE um and almost total clearing was seen in 15% versus 3% favoring uh the litifilimab. If you look at another better measure in my opinion CLASI-50 — that you use for cutaneous lupus — at week 24 it was 41% versus 21%. Um this is encouraging data and we like to see more of this going forward.
Um lastly uh I found a novel report about uh what physicians are doing with regard to um artificial intelligence. This was a commissioned series of surveys by the AMA that started in 2023, 2024 and now repeated in 2026. From 2023 to 2026, AI adoption by US physicians has doubled from 48% to 81% of you are using AI in some way, shape, or form. Most of you are using it to summarize research. About 40% versus 13% in 2023. That would include looking at, you know, newer therapies, how to use them, where to use them, guidelines, etc. Um, only a minority are using AI for diagnostic assist. That's 17%. Um, the majority of you think it's okay for your patients to use AI for general health and medication questions, but more than half of you are sort of stringently against patients using it to interpret radiology or pathology results.
Um, the majority of us, three-quarters of us expect that AI is going to automate clinical and administrative tasks that will therefore lessen burnout. But um, and I think that most of us are playing with this and trying to use it to do things faster with more expediency. Um, again, I think AI works when you're the expert in the topic that you're using it in because then you can spot the problems where AI might be making up numbers or making up uh conclusions because that is certainly possible with AI.
Um, but I want to bring to your attention what um, uh, Dr. Janice Yazdany brought up at the ACR 2025. She and Jeff Curtis had a session on AI in medical practice and she brought up this issue of skill erosion — that when you start using all these AI tools, you get away from skills that you would normally use — and that um it turns out that 70% of physicians are worried about um skill erosion and that that could lead to a change in uh how we practice uh and the standards that we have rigidly adhered to thus far. The article, the review, the survey does recommend that there's a big call for trust and training on the use of AI in medicine. Anyway, hope you enjoyed these reports as much as I did. Um, go to the website, check out these citations and more, and we'll talk next week. Take care of yourself.
So, I think one of the interesting things that we put up this week was this report about the use of artificial intelligence scribes or ambient dictation that could occur and is often now included in a lot of EHR packages — not ER, EHR packages — where you don't have to sit at the keyboard while you're talking to the patient and hammer things out or write your notes afterwards. You can just turn on the ambient dictation and the EHR will take notes for you that you have to review later on and whatnot.
So, this was sort of a systematic review in a medical center and looked at how impactful AI scribes were on reducing documentation and if it could do that, wouldn't it therefore be good to lessen clinician burnout? Because clinician burnout — the number one thing we point to is the administrative work that goes into that visit that you do over and over and over again. And what this showed was that while ambient dictation may seem to reduce burnout, the effect on clinician workflows were pretty small when you looked at the overall system. However, if you looked at those who used the EHR and ambient dictation — it was primarily being used by primary care doctors, by women more so than men, and by those who used it in more than half of their patient encounters — they're the ones who got the big benefit. The rest of you who are dabbling are shrugging and rolling your eyeballs and saying it didn't really do me. But if you're one of the people that use a lot of it, they experience twice the reduction in EHR total time and three times the reduction in documentation time. But yet the number of people who were adopting that level of technology is less than one-third.
So I think it's time to get on board and use these new tools. You know, I prided myself in being the best at writing an electronic note in front of the patient using all kinds of tricks and tools, and I wrote really good detailed notes. And what did it get me? Nothing. It got me patients who thought I wasn't paying attention. It got me one-and-a-half-page notes that no one read. I think you should minimize your time with documentation. Meet the rules as dictated by your job, but spend more time talking to patients — and if you have a tool like ambient dictation, use it, because it's going to give you more face-to-face time with patients.
A few things on pain showed up this week. The NHIS survey, which goes on periodically in the United States — this one looked at chronic pain in almost 200,000 Americans and showed those people who had chronic pain versus those who didn't had double the rates of smoking. Now, is that not a double whammy? It's not bad enough you got chronic pain, but now you're smoking on top of that. And they run together. Smoking, while it has decreased over time in the US and other cultures, always had a more significant association in the individuals who had chronic pain. E-cigarettes, however, have increased over time, but yes, they too have had an association with chronic pain. Bad news.
You know, another disheartening thing about pain and use of pain meds is that yeah, you're not supposed to use narcotics. That's like a no-no and it's hard to get anyone who will use narcotics these days. So we're using a lot of adjunctive therapies and that includes the gabapentin-like medicines. A worldwide study of gabapentin use from 2008 to 2018 showed that a) it's used worldwide, and b) it's going up worldwide as far as incremental use year-to-year. It's gone from roughly four doses per day to 18 doses per day in 10,000 population. The highest gabapentin use was amongst higher income countries, so much so that it was six-fold higher compared to low- and medium-income countries. And again, the number of pill equivalents was like 40 per day versus six per day in high- versus low-income countries. But then there's all this data that says gabapentin is dangerous and associated with bad outcomes, and doesn't work all that well. What are we going to do about pain?
I do use gabapentin, and I do think most people use it at wimpy doses — 100 milligrams once or twice a day, you might as well just flick that right out the window. That's like totally useless. But you know, how high will you go and do you believe all the scary literature on gabapentin? Hard to know what to do.
Uh, another study
looked at how good we are at doing the things we're supposed to do when we start to prescribe medium to high-dose steroids. So this is a UK CPRD — it's a Practice Research Data Link — looked at 40,000 people with polymyalgia rheumatica who were started on steroids, and they looked at how many were co-prescribed bisphosphonates, because once you go on steroids with PMR you're going to be on it for four years on average, right? Not the one year that you tell everyone.
And how many were co-prescribed either an H2 receptor antagonist or a PPI? The numbers were 67% received a co-prescribed bisphosphonate and almost 80% received a PPI. This was less likely to happen in males and those who were living in deprived areas. And if you didn't get co-prescriptions of bisphosphonates, there were more fractures in the first 12 months — the numbers low: 2.3% versus 1.4%. It was more. But this is like the data where people show up with a new fracture and they don't get a DEXA scan. What's the deal with that? Or if they do have it, they're not treated for osteopenia or osteoporosis and they're not started on calcium and vitamin D. There are these little things that probably need to be reflexively done when steroids are used. And I think most of us are good at this, but it wouldn't be bad to check your checklist to see if that's on there. By the way, it didn't seem to make any difference as far as GI outcomes for those who are prescribed PPIs or H2 blockers.
I saw an interesting article this week on maybe a new mechanism about how methotrexate works. You know, it's kind of voodoo and we need Bruce Cronstein and other brilliant rheumatologists to really go through the biology of it and talk about adenosine and how adenosine is anti-inflammatory — sorry, is inflammatory — and how methotrexate counteracts that. But anyway, this new mechanism looks at methotrexate effects being mediated through monocytes, and you know monocytes and monocyte counts have been linked to efficacy of the drugs that we use in RA, and maybe that is related to methotrexate. We do know that methotrexate induces — results in — polyglutamation, and that is seen not just in RBCs but also in monocytes, that peaks at day five post-methotrexate. But to get a more substantial transcriptional effect on the genes that might drive inflammation, again mediated through monocyte-associated genes, peaks at around week four, and these transcriptional changes do seem to predict methotrexate responses. I like this new postulate. We'll have to see if it pans out in the hands of other researchers.
Leeds, known for their great studies on early RA and early interventions in pre-clinical RA, published a study about 63 patients with palindromic rheumatism who progressed to develop either persistent inflammatory arthritis or met criteria for rheumatoid arthritis. These people who did progress — from P and I, again, I've said before, PR is just another version of clinically suspect arthralgia, pre-clinical RA; it's just a more defined presentation — but these 63 who progressed from palindromic to RA-like disease did so at a mean/median of 13 months. And the point of the paper was the joints that you get swelling and inflammation in, mostly in the hands and fingers, are the same ones that you're going to get when RA manifests as chronic symmetric polyarthritis. So it's kind of like a joint-priming kind of event. And it turns out that if you were the person who developed RA but had a palindromic onset, you're going to have significantly more swollen joints at the time of diagnosis compared to others. Now, is that a real thing or is that like a selection bias? Like, once you're labeled as palindromic, you need to have more persistent swollen joints to be labeled as RA. Hard to say. But I like this research coming out of Dr. Emery's research group.
New England Journal article this week — there were two. One on Kawasaki's disease, saying that there's no role for prednisolone in the early management of Kawasaki's disease over standard-of-care therapies. And that's been said before. This was a study out of China. And then there was a study on ITP and the use of ianalumab. That's the BAFF inhibitor — it's a B cell depleter. It was a randomized controlled trial, placebo-controlled, double-blind, 152 patients with ITP. Everyone receives the oral thrombopoietin receptor agonist, which is a standard of care — it's called eltrombopag — it is a standard of care for ITP, and then they were either given placebo or 3 or 9 milligrams of ianalumab. They looked at outcomes at 4 months and 6 months, and better outcomes were seen with ianalumab, especially at the 9 milligram dose, with the time to failure being significantly longer with ianalumab. And the failure was like getting a platelet count less than 30,000. And when
they looked at six-month outcomes and six-month response rates were better again with that 9 mg dose versus placebo 62 versus 39%. I don't know about you, but I've always recommended ITP patients be treated with rituximab. I think B-cell-depleting therapies like ublituximab are the way to go as part of the regimen. I think it's better — in my opinion it's better at actually inducing remission so that those patients won't need chronic long-term therapy and anticoagulant sort of therapy.
I like this report on deprescribing of PPIs. The article went on to say that there's a lot of inappropriate prolonged use of PPIs. And how do you get people off of PPIs when they don't need to be on it? You could do that with usual care — instructions just for the patient, instructions for the GP and the patient in a GP-and-patient-facing intervention. And it turns out that the GP-patient-facing intervention was twice as successful as usual care or just educating the GP only about you should do this. And why do I put this up and why am I going on about PPI deprescribing? We really aren't great at deprescribing, are we? Especially when it comes to steroids — meaning you put people on steroids because they're doing lousy and then you put them on advanced therapies and then they get better. How good are you at withdrawing those therapies? Most of us aren't very good. And that is not just for RA. That's been seen for PsA and for spondylitis. So deprescribing should be put in the guidelines, as it was for the JIA guidelines that were published at the last ACR — they made that a big component.
I think this is a Lancet global burden of disease study that looked at gout and its incidence — that it has increased by 158% from 1990 to 2023, and that most of that increase was related to aging of the population. There has been some talk about gout seeming to be occurring younger. Well, this paper says it's more from an aging population and not from younger onset of gout. They didn't go into the influence of obesity, but obesity is affecting the young and the old, and aging seems to be the big factor. So while the mean onset age did increase from 58 to 62 years, the percentage of young-onset gout — below 50 I think the number was — dropped from 16% to 9.8%. Figure that out.
TRACE was a study of what happens when you escalate secukinumab in patients with active axial spondyloarthritis. The TRACE study was a 90-patient active SpA, bio-naive study where they were randomized to receive 150 milligrams of secukinumab, and then at week 16, if you were in ASAS remission you stayed on that dose, but if you weren't you increased the secukinumab to 300 milligrams. By week 24 there wasn't much increase in efficacy. So escalating secukinumab really didn't do a lot, but for the 55 or so — one in five of them did achieve an ASAS remission. So it had a minor effect but it wasn't significant as far as clinical outcomes or as far as MRI inflammation as measured by SPARCC scores. Most of the MRI changes in this one-year study occurred in the first four weeks and not much after that.
This is reminiscent of the ESCALATE study published I think last year that looked at whether the use of treat-to-target dose escalation with secukinumab in SpA was going to be effective, and it didn't really show that it was any more effective than usual care with other drugs. So will you do better if you escalate from 150 to 300 in SpA? I would do it. I have done it. I think though that these two studies are questioning its validity and whether that's going to be good enough to pay for it.
There was a study this week — novel in my view in that it's something I don't do — and that is plasma exchange in patients with idiopathic inflammatory myositis. Systematic review, 35 studies, 473 myositis patients. 361 of them were treated with plasma exchange. There's got to be a publication bias here — like who else would do this and report it? And many of the reports did suggest improvement, but it was kind of weak. And the seven studies that were randomized and comparative showed no mortality benefit, especially with regard to severe ILD outcomes and ILD lung mortality.
I don't use and have not had to use plasma exchange in inflammatory myositis, and I've treated patients just as bad as you. But I think the problem here is that there's no standard playbook when it comes to polymyositis, dermatomyositis, and — I don't know — myositis. I hope it's not inclusion body myositis. We all use steroids to start and they're dangerous, really dangerous in myositis in my opinion. More dangerous in myositis than lupus, my opinion. I've had more osteonecrosis in myositis. I've had more severe, deadly infections in myositis patients.
all occurring in the first 6 months by the way. Um so you know we all use steroids and then we jump in with what drugs that have not been proven to work and are not FDA approved because there are no clinical trials. I mean the only drug that's that's FDA approved is IVIG and to me that's a third line drug. My first line is first line DMARD is um is methotrexate followed by leflunomide and then azathioprine and then I've used a whole bunch of other things including calcineurin inhibitors and cytotoxics and whatnot but I've not used plasma exchange and I wouldn't based on this report.
Two more reports. Another one uh coming out of the uh American Academy of Dermatology meeting that just completed — a study uh called the AMETHYST study was presented by rheumatologist/dermatologist uh Dr. Joseph Moro of UT Southwestern. He gave us a little video that you might want to look at where he describes what happens when they use this novel monoclonal antibody that targets BDCA2. Uh it's a marker that you find on plasmacytoid dendritic cells. Um there was a previous phase 2 trial and this is a I think another phase 2 trial the AMETHYST study but this one's being done in cutaneous lupus who that was active despite prior therapy with antimalarials. To get into the study, you had to have active either CLE or chronic CLE, most of those being DLE. These were not patients with acute LE. And you either received placebo or um litifilimab and um everybody received uh um background standard of care that remained stable. The outcomes at week 24 um was an investigator global assessment of CLE um and almost total clearing was seen in 15% versus 3% favoring uh the litifilimab. If you look at another better measure in my opinion CLASI-50 — that you use for cutaneous lupus — at week 24 it was 41% versus 21%. Um this is encouraging data and we like to see more of this going forward.
Um lastly uh I found a novel report about uh what physicians are doing with regard to um artificial intelligence. This was a commissioned series of surveys by the AMA that started in 2023, 2024 and now repeated in 2026. From 2023 to 2026, AI adoption by US physicians has doubled from 48% to 81% of you are using AI in some way, shape, or form. Most of you are using it to summarize research. About 40% versus 13% in 2023. That would include looking at, you know, newer therapies, how to use them, where to use them, guidelines, etc. Um, only a minority are using AI for diagnostic assist. That's 17%. Um, the majority of you think it's okay for your patients to use AI for general health and medication questions, but more than half of you are sort of stringently against patients using it to interpret radiology or pathology results.
Um, the majority of us, three-quarters of us expect that AI is going to automate clinical and administrative tasks that will therefore lessen burnout. But um, and I think that most of us are playing with this and trying to use it to do things faster with more expediency. Um, again, I think AI works when you're the expert in the topic that you're using it in because then you can spot the problems where AI might be making up numbers or making up uh conclusions because that is certainly possible with AI.
Um, but I want to bring to your attention what um, uh, Dr. Janice Yazdany brought up at the ACR 2025. She and Jeff Curtis had a session on AI in medical practice and she brought up this issue of skill erosion — that when you start using all these AI tools, you get away from skills that you would normally use — and that um it turns out that 70% of physicians are worried about um skill erosion and that that could lead to a change in uh how we practice uh and the standards that we have rigidly adhered to thus far. The article, the review, the survey does recommend that there's a big call for trust and training on the use of AI in medicine. Anyway, hope you enjoyed these reports as much as I did. Um, go to the website, check out these citations and more, and we'll talk next week. Take care of yourself.
If you are a health practitioner, you may Login/Register to comment.
Due to the nature of these comment forums, only health practitioners are allowed to comment at this time.